Portland loves its small businesses. How well does it help them?
About half the jobs in the Portland area are at companies with fewer than 500 employees. We looked at what those businesses do, what they pay, what gets in their way and whether the help on offer is working.
Start with the big pictureof jobs are at companies with fewer than 500 employees. Those companies account for a smaller share of pay and sales.
Portland–Vancouver–Hillsboro metro area. Counts only businesses that have employees. Sales are not the same thing as GDP. Source ↗
The big picture
Counting businesses is easy. Knowing how they’re doing is harder.
A restaurant is a small business. So is a therapist’s practice, a subcontractor, a home-care provider, a design studio, a repair shop and a consultant working alone. Some owners want to grow. Others want a steady income and control over their own time.
We set out to answer a plain question: is Portland a good place to make a living from a small business? Public data answers part of it. We can see how many jobs small companies provide, what industries they are in and how the years since 2019 have gone.
We can see very little about what owners take home, or whether the help the city offers has solved anyone’s problem. That gap is also an opening. Portland could make it easier to get from a workable idea to a workable living, and then check whether it did.
Half of Portland-area jobs are at companies with fewer than 500 employees. Of the eight metro areas we compared, only Sacramento’s share is higher.
Finding 2We found no local data on whether business owners are making a living.Sales, payroll and survival each show part of the picture. None of them shows what an owner keeps after expenses.
Finding 3We can see more of the assistance. Completed outcomes remain the missing link.OSB’s October 20 dashboard reports 940 unique businesses with action logs and feedback from 50 survey responses. It does not show how often the help fixed the problem.
The city, the county and the metro area are three different places.
Our questions are about the city of Portland. The most detailed data on company size covers the whole metro area, so we show which place each number describes.
jobs at small businesses in 2024, as reported by Prosper Portland
The closest match to city policy. The report is not clear about how it defines a small business.jobs at businesses with employees in 2022
Includes places outside Portland and leaves out the parts of Portland that sit in other counties.jobs at businesses with employees in 2022
The whole regional job market, including Vancouver, WA, and Hillsboro. Most of our size comparisons use it.These three numbers measure different places and different things, so they should not be compared with each other. Every chart below names the place and year it covers.
SourcesInsights and Indicators 2025 ↗2022 SUSB county enterprise size workbook ↗2022 SUSB MSA 3-digit NAICS dataset ↗
What counts as small
How small is small?
It depends on where you draw the line. Count only companies with fewer than 20 employees, and small business is about one job in five. Count every company with fewer than 500, and it is one job in two.
We use two cutoffs, and neither is official.
The Census Bureau publishes figures for companies with fewer than 20 employees and for companies with fewer than 500, so we show both. Neither is a legal definition. Grant and loan programs set their own rules, often by industry, revenue, ownership or location.
The Census counts the whole company. A shop with ten workers that belongs to a national chain is counted with the chain, as a large company. The Census also uses the word “receipts” for what a business takes in. We call that sales.
Small companies have half the jobs and a smaller share of the money.
Portland metro area, 2022. Each number is a share of the total for all businesses with employees.
of all jobs in the Portland area
are at companies with fewer than 500 employees, counting every location the company has.
Each square is one percent of the total. Businesses with no employees are not in this chart.
See the numbers in a table
| Company size (employees) | Jobs | Share of jobs | Share of pay | Share of sales |
|---|---|---|---|---|
| Under 5 | 56,535 | 5.2% | 4.0% | 4.0% |
| 5–9 | 63,433 | 5.8% | 3.9% | 3.5% |
| 10–19 | 83,601 | 7.7% | 5.4% | 4.4% |
| 20–99 | 191,448 | 17.7% | 14.3% | 12.0% |
| 100–499 | 149,838 | 13.8% | 13.4% | 11.5% |
| 500+ | 539,680 | 49.8% | 59.0% | 64.6% |
Jobs are counted in mid-March. Pay and sales cover the full year. The Census Bureau slightly alters some figures to protect individual businesses, so we round them here. The download keeps the Census flags.
Sources2022 SUSB MSA 3-digit NAICS dataset ↗SUSB methodology ↗Download the data ↓
About half of all jobs are at companies with 500 or more employees.
Portland metro area, 2022. Jobs at businesses with employees, by the size of the whole company.
The six groups add up to 1,084,535 jobs. The under-20 and under-500 cutoffs used elsewhere overlap, so they should not be added together.
Source2022 SUSB MSA 3-digit NAICS dataset ↗Download the data ↓
Why the money is split less evenly than the jobs
Companies with fewer than 500 employees have 50.2% of the jobs, 41.0% of the pay and 35.4% of the sales. One reason is that large and small companies tend to be in different lines of work. Industries differ in what they pay, how many hours people work and how much equipment and material a business has to buy.
These totals don’t show that any one small business is badly run, or that a large company does more good. They do show why “most of our businesses are small” is a weak argument for a subsidy. A count of businesses treats a one-person practice and a large employer as one each. Jobs, what owners earn, what the business provides and what the subsidy costs all belong in the decision.
A number we can’t give you
What share of Portland’s economy comes from small business?
We could not find a reliable answer for the city. GDP measures the value a business adds, which is different from its sales or its headcount. The federal Bureau of Economic Analysis published experimental small-business estimates and no longer produces them regularly. The bureau’s notice ↗
Download our memo on estimating itHow someone could estimate the GDP share
Start with total economic output for the area. Work industry by industry, estimating how much of each industry’s output comes from small companies, using pay, production and income data that line up with each other. Estimate businesses with no employees separately. Set aside government, housing and anything that can’t be assigned, so the leftover is not counted as “big business.” Then test how far the answer moves when the assumptions change.
The shortcuts don’t work. Applying the national small-business share to Portland’s GDP only repeats the national answer. Using the local share of jobs assumes small and large companies are equally productive, which is the thing we are trying to find out.
An estimate like this could be useful if it came with a range and was clearly labeled as a model.
What the businesses do
Restaurants and shops are only the part you can see.
Health care employs more people here than any other industry. Construction, professional services and people who work for themselves deserve as much attention as restaurants and retail.
Health care, retail and manufacturing employ the most people.
Portland metro area, 2022. Every business with employees, grouped by industry.
Try switching from jobs to companies. Health care has many small practices alongside a few very large health systems, so it looks different each way. A company that works in two industries is counted in both, so the company counts can’t be added up.
See the numbers in a table
| Industry | All jobs | Jobs at companies under 500 | Companies | Total pay |
|---|---|---|---|---|
| Health care | 165,698 | 71,150 | 7,083 | $10,626,422,000 |
| Retail | 121,817 | 43,691 | 4,855 | $4,962,872,000 |
| Manufacturing | 113,841 | 50,188 | 2,941 | $9,759,784,000 |
| Hotels and restaurants | 100,485 | 69,212 | 5,217 | $2,842,986,000 |
| Professional services | 85,883 | 54,625 | 8,870 | $8,740,718,000 |
| Construction | 82,002 | 62,157 | 8,366 | $6,614,793,000 |
| Administrative services | 63,727 | 34,945 | 3,501 | $3,048,521,000 |
| Wholesale | 60,441 | 30,981 | 2,946 | $5,379,331,000 |
| Transportation | 47,592 | 15,201 | 2,029 | $2,638,813,000 |
| Corporate offices | 47,333 | 4,615 | 359 | $6,998,318,000 |
| Finance | 46,368 | 13,664 | 2,217 | $5,033,764,000 |
| Other services | 44,510 | 37,844 | 5,710 | $2,004,240,000 |
| Information | 29,344 | 11,018 | 1,111 | $3,435,723,000 |
| Education | 26,392 | 15,429 | 1,101 | $1,040,249,000 |
| Real estate | 23,358 | 15,061 | 3,482 | $1,455,539,000 |
| Arts and recreation | 17,212 | 12,905 | 1,132 | $737,920,000 |
| Utilities | 5,771 | 133 | 42 | $713,444,000 |
| Agriculture | 2,274 | 1,776 | 243 | $143,621,000 |
| Mining | 422 | 195 | 27 | $36,493,000 |
Jobs are counted in mid-March. Pay is the total for the year.
Source2022 SUSB MSA 3-digit NAICS dataset ↗Download the data ↓
Care and other essential services
A clinic, a childcare provider or a home-care business is valuable because people can get to it and count on it. If it grows while families can’t afford it or staff keep leaving, the growth has missed the point.
Manufacturing and the trades
A fabricator or contractor needs space, equipment, power, skilled workers, customers and cash to cover costs until the customer pays. A marketing workshop helps with one of those.
Professional and independent work
A professional practice can sell to clients far outside the region and needs very little space. What runs short is reputation, contracts, specialized skill and the owner’s time.
Businesses with no employees are most common in professional services, transportation and the arts.
Multnomah County, 2023. The ten largest industries by the measure you choose.
See the numbers in a table
| Industry | Businesses with no employees | Total sales |
|---|---|---|
| Professional services | 15,372 | $754,842,000 |
| Transportation | 10,637 | $443,507,000 |
| Arts and recreation | 9,131 | $231,804,000 |
| Health care | 7,373 | $385,458,000 |
| Real estate | 7,043 | $994,451,000 |
| Other services | 5,620 | $224,701,000 |
| Administrative services | 4,333 | $136,566,000 |
| Retail | 4,171 | $294,203,000 |
| Construction | 3,931 | $309,021,000 |
| Education | 3,361 | $59,234,000 |
| Information | 2,001 | $75,753,000 |
| Hotels and restaurants | 1,610 | $80,494,000 |
| Manufacturing | 1,139 | $72,596,000 |
| Finance | 999 | $89,135,000 |
| Wholesale | 812 | $96,296,000 |
| Agriculture | 266 | $15,521,000 |
| Utilities | 16 | $4,677,000 |
| Mining | 14 | $3,657,000 |
The Census places each business by its mailing address, which may not be where the work happens. The count includes side businesses. Sales are before expenses, so they are not what the owner earned.
Sources2023 Nonemployer Statistics county file ↗Nonemployer Statistics methodology ↗Download the data ↓
A business with no employees is still someone’s job.
The owner does the selling, the work, the bookkeeping and the customer service, sometimes on top of another job. Helping these businesses can raise someone’s income and independence even if they never hire anyone.
More of these businesses is not always good news. It can mean people are piecing together income after losing steady work, and the counts can’t tell us which.
We also can’t add these 77,829 businesses to the metro numbers above to get a total for “Portland’s businesses.” The places, the years and the things being counted are all different.
How we compare
Small companies matter more here. Is that a strength?
Among nine metro areas, only Sacramento has a bigger share of its jobs at companies with fewer than 500 employees. When we give every metro the same mix of industries, Portland comes out on top. What that says about the economy is a judgment call.
Only Sacramento has a bigger share of jobs at small companies.
Nine U.S. metro areas, 2022. Share at companies below the size you choose.
Across all 387 U.S. metro areas, 47.5% of jobs are at companies this size. The chart compares nine metros. It does not rank them from best to worst.
See the numbers in a table
| Metro area | Actual share | Share shown |
|---|---|---|
| Sacramento | 50.4% | 50.4% |
| Portland | 50.2% | 50.2% |
| Austin | 48.1% | 48.1% |
| Minneapolis | 45.8% | 45.8% |
| Denver | 45.2% | 45.2% |
| Pittsburgh | 44.9% | 44.9% |
| Seattle | 43.9% | 43.9% |
| Indianapolis | 42.7% | 42.7% |
| Salt Lake City | 42.0% | 42.0% |
The industry adjustment covers 19 broad industries and leaves out a small “unclassified” group, which is 0.006% of Portland jobs.
Source2022 SUSB MSA 3-digit NAICS dataset ↗Download the data ↓
Many people here can own something.
A high share may mean it is possible to start small in Portland. It may reflect networks of specialized suppliers, professional practices and services that neighborhoods use. Those benefits are real even when a business never becomes a national brand.
There may be too few big employers or good jobs.
The same share can be high because the region has fewer large employers, because businesses here struggle to grow, or because people start businesses when they can’t find good paid work. The number alone can’t tell us which.
Portland is a little above the figure for all U.S. metro areas.
Every U.S. metro area, 2022. Share of jobs at companies with fewer than 500 employees.
Each dot is one metro area. The dots are stacked only to keep them from overlapping. Smaller metros often have higher shares. This view shows 387 metro areas and leaves out small towns and rural areas.
See all 387 metro areas in a table
| Metro area | Jobs | Share at companies under 500 |
|---|---|---|
| New York-Newark-Jersey City, NY-NJ | 8,206,342 | 50.2% |
| Los Angeles-Long Beach-Anaheim, CA | 5,460,130 | 52.6% |
| Chicago-Naperville-Elgin, IL-IN | 4,175,455 | 44.6% |
| Dallas-Fort Worth-Arlington, TX | 3,439,459 | 42.1% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 2,775,367 | 49.1% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 2,765,882 | 46.0% |
| Houston-Pasadena-The Woodlands, TX | 2,730,105 | 46.3% |
| Boston-Cambridge-Newton, MA-NH | 2,599,200 | 44.4% |
| Atlanta-Sandy Springs-Roswell, GA | 2,567,019 | 43.1% |
| San Francisco-Oakland-Fremont, CA | 2,360,860 | 46.7% |
| Miami-Fort Lauderdale-West Palm Beach, FL | 2,333,976 | 53.6% |
| Phoenix-Mesa-Chandler, AZ | 1,998,365 | 42.6% |
| Seattle-Tacoma-Bellevue, WA | 1,845,497 | 43.9% |
| Minneapolis-St. Paul-Bloomington, MN-WI | 1,840,739 | 45.8% |
| Detroit-Warren-Dearborn, MI | 1,781,381 | 47.2% |
| Denver-Aurora-Centennial, CO | 1,369,510 | 45.2% |
| San Diego-Chula Vista-Carlsbad, CA | 1,338,161 | 51.1% |
| Riverside-San Bernardino-Ontario, CA | 1,335,043 | 47.2% |
| St. Louis, MO-IL | 1,244,433 | 44.4% |
| Tampa-St. Petersburg-Clearwater, FL | 1,230,894 | 45.0% |
| Orlando-Kissimmee-Sanford, FL | 1,192,927 | 38.6% |
| Baltimore-Columbia-Towson, MD | 1,188,401 | 48.7% |
| Charlotte-Concord-Gastonia, NC-SC | 1,178,479 | 42.1% |
| San Jose-Sunnyvale-Santa Clara, CA | 1,147,117 | 37.0% |
| Pittsburgh, PA | 1,108,776 | 44.9% |
| Portland-Vancouver-Hillsboro, OR-WA | 1,084,535 | 50.2% |
| Cincinnati, OH-KY-IN | 1,003,391 | 41.8% |
| Kansas City, MO-KS | 990,521 | 44.1% |
| Austin-Round Rock-San Marcos, TX | 985,636 | 48.1% |
| Nashville-Davidson--Murfreesboro--Franklin, TN | 953,416 | 41.4% |
| Indianapolis-Carmel-Greenwood, IN | 951,953 | 42.7% |
| Cleveland, OH | 946,377 | 46.3% |
| San Antonio-New Braunfels, TX | 946,286 | 40.9% |
| Columbus, OH | 933,569 | 40.5% |
| Las Vegas-Henderson-North Las Vegas, NV | 917,211 | 44.6% |
| Sacramento-Roseville-Folsom, CA | 809,745 | 50.4% |
| Milwaukee-Waukesha, WI | 763,119 | 47.7% |
| Salt Lake City-Murray, UT | 676,512 | 42.0% |
| Jacksonville, FL | 646,582 | 42.4% |
| Providence-Warwick, RI-MA | 636,408 | 51.7% |
| Virginia Beach-Chesapeake-Norfolk, VA-NC | 633,601 | 47.1% |
| Louisville/Jefferson County, KY-IN | 630,363 | 39.8% |
| Raleigh-Cary, NC | 590,073 | 47.4% |
| Richmond, VA | 564,882 | 43.4% |
| Memphis, TN-MS-AR | 544,567 | 42.0% |
| Oklahoma City, OK | 531,803 | 51.5% |
| Grand Rapids-Wyoming-Kentwood, MI | 528,648 | 50.6% |
| Hartford-West Hartford-East Hartford, CT | 520,790 | 46.0% |
| Buffalo-Cheektowaga, NY | 475,093 | 49.7% |
| Birmingham, AL | 470,623 | 44.8% |
| Omaha, NE-IA | 436,354 | 45.4% |
| Rochester, NY | 433,281 | 46.4% |
| New Orleans-Metairie, LA | 400,119 | 51.5% |
| Bridgeport-Stamford-Danbury, CT | 398,349 | 51.4% |
| Tulsa, OK | 390,711 | 50.9% |
| Greenville-Anderson-Greer, SC | 373,076 | 44.1% |
| Knoxville, TN | 358,015 | 43.7% |
| Allentown-Bethlehem-Easton, PA-NJ | 352,044 | 40.8% |
| Albany-Schenectady-Troy, NY | 351,094 | 45.3% |
| Des Moines-West Des Moines, IA | 350,871 | 40.9% |
| Madison, WI | 348,900 | 44.7% |
| Baton Rouge, LA | 340,218 | 49.8% |
| Tucson, AZ | 331,786 | 45.9% |
| Urban Honolulu, HI | 330,476 | 51.3% |
| Greensboro-High Point, NC | 329,264 | 44.9% |
| Fresno, CA | 324,552 | 53.2% |
| Dayton-Kettering-Beavercreek, OH | 320,376 | 43.8% |
| Worcester, MA | 311,959 | 47.3% |
| Boise City, ID | 310,862 | 52.9% |
| Albuquerque, NM | 308,226 | 48.9% |
| Charleston-North Charleston, SC | 307,235 | 53.0% |
| Columbia, SC | 303,225 | 44.0% |
| Little Rock-North Little Rock-Conway, AR | 302,544 | 47.7% |
| Akron, OH | 297,184 | 47.6% |
| Durham-Chapel Hill, NC | 294,892 | 35.9% |
| Harrisburg-Carlisle, PA | 292,594 | 38.6% |
| North Port-Bradenton-Sarasota, FL | 272,404 | 58.1% |
| Provo-Orem-Lehi, UT | 272,086 | 50.5% |
| Oxnard-Thousand Oaks-Ventura, CA | 269,050 | 58.1% |
| Toledo, OH | 261,747 | 45.2% |
| Colorado Springs, CO | 260,935 | 49.1% |
| Wichita, KS | 260,626 | 49.9% |
| New Haven, CT | 259,228 | 46.4% |
| Syracuse, NY | 256,247 | 46.4% |
| Cape Coral-Fort Myers, FL | 255,271 | 55.8% |
| Portland-South Portland, ME | 252,562 | 52.4% |
| Winston-Salem, NC | 249,687 | 42.5% |
| El Paso, TX | 248,813 | 49.9% |
| Lancaster, PA | 245,184 | 54.9% |
| Lexington-Fayette, KY | 233,341 | 43.8% |
| Chattanooga, TN-GA | 232,075 | 44.0% |
| Scranton--Wilkes-Barre, PA | 232,014 | 44.8% |
| Reno, NV | 231,341 | 50.3% |
| Fayetteville-Springdale-Rogers, AR | 230,601 | 41.1% |
| Jackson, MS | 227,258 | 45.7% |
| Lakeland-Winter Haven, FL | 223,169 | 40.0% |
| McAllen-Edinburg-Mission, TX | 221,390 | 50.9% |
| Kiryas Joel-Poughkeepsie-Newburgh, NY | 217,581 | 53.7% |
| Stockton-Lodi, CA | 214,027 | 47.7% |
| Spokane-Spokane Valley, WA | 212,637 | 53.0% |
| Huntsville, AL | 209,045 | 43.2% |
| Fort Wayne, IN | 208,459 | 47.8% |
| Bakersfield-Delano, CA | 206,892 | 54.3% |
| Trenton-Princeton, NJ | 200,283 | 41.4% |
| Springfield, MO | 196,974 | 46.4% |
| Augusta-Richmond County, GA-SC | 195,277 | 42.5% |
| Ogden, UT | 195,082 | 53.4% |
| Palm Bay-Melbourne-Titusville, FL | 194,219 | 50.3% |
| Deltona-Daytona Beach-Ormond Beach, FL | 184,924 | 52.7% |
| Manchester-Nashua, NH | 184,188 | 50.2% |
| Santa Rosa-Petaluma, CA | 176,295 | 63.1% |
| York-Hanover, PA | 175,009 | 44.4% |
| Savannah, GA | 171,121 | 45.0% |
| Lansing-East Lansing, MI | 169,328 | 48.3% |
| Springfield, MA | 168,312 | 49.6% |
| Asheville, NC | 166,060 | 52.1% |
| Davenport-Moline-Rock Island, IA-IL | 161,771 | 45.4% |
| Anchorage, AK | 160,998 | 50.9% |
| Lafayette, LA | 159,834 | 63.1% |
| Reading, PA | 159,778 | 45.2% |
| Boulder, CO | 157,001 | 57.6% |
| Wilmington, NC | 156,846 | 55.8% |
| Santa Maria-Santa Barbara, CA | 156,607 | 62.6% |
| Corpus Christi, TX | 155,803 | 46.8% |
| Sioux Falls, SD-MN | 154,840 | 48.8% |
| Green Bay, WI | 154,736 | 47.7% |
| Peoria, IL | 153,533 | 45.0% |
| Lincoln, NE | 152,800 | 52.5% |
| Mobile, AL | 152,767 | 46.3% |
| Spartanburg, SC | 151,851 | 37.6% |
| Ann Arbor, MI | 151,846 | 47.3% |
| Modesto, CA | 151,505 | 51.9% |
| Pensacola-Ferry Pass-Brent, FL | 145,458 | 45.5% |
| Naples-Marco Island, FL | 145,185 | 63.8% |
| Waterbury-Shelton, CT | 144,240 | 54.7% |
| Canton-Massillon, OH | 144,239 | 52.2% |
| Elkhart-Goshen, IN | 143,423 | 38.4% |
| Hickory-Lenoir-Morganton, NC | 141,070 | 45.5% |
| Roanoke, VA | 138,814 | 44.2% |
| Youngstown-Warren, OH | 138,368 | 58.6% |
| Shreveport-Bossier City, LA | 135,884 | 53.7% |
| Evansville, IN | 135,727 | 42.9% |
| Fort Collins-Loveland, CO | 135,516 | 58.2% |
| Port St. Lucie, FL | 135,457 | 56.6% |
| Atlantic City-Hammonton, NJ | 134,222 | 50.8% |
| Salem, OR | 132,193 | 61.4% |
| Montgomery, AL | 131,323 | 49.3% |
| Lubbock, TX | 130,372 | 54.7% |
| Gulfport-Biloxi, MS | 130,044 | 43.1% |
| South Bend-Mishawaka, IN-MI | 128,199 | 44.2% |
| Eugene-Springfield, OR | 127,436 | 63.4% |
| Beaumont-Port Arthur, TX | 127,346 | 51.0% |
| Cedar Rapids, IA | 126,825 | 45.1% |
| Rockford, IL | 126,257 | 45.9% |
| Fargo, ND-MN | 124,634 | 50.7% |
| Flint, MI | 120,612 | 52.4% |
| Vallejo, CA | 119,655 | 48.3% |
| Tallahassee, FL | 119,574 | 54.5% |
| Rochester, MN | 118,610 | 35.7% |
| Myrtle Beach-Conway-North Myrtle Beach, SC | 118,467 | 58.9% |
| Salinas, CA | 116,582 | 57.7% |
| Brownsville-Harlingen, TX | 114,735 | 49.2% |
| Appleton, WI | 114,283 | 49.6% |
| Kalamazoo-Portage, MI | 113,731 | 47.4% |
| Huntington-Ashland, WV-KY-OH | 113,441 | 49.4% |
| Waco, TX | 113,393 | 45.0% |
| Erie, PA | 112,074 | 49.8% |
| Killeen-Temple, TX | 111,440 | 43.9% |
| Gainesville, FL | 106,762 | 49.6% |
| Duluth, MN-WI | 104,858 | 53.8% |
| Visalia, CA | 103,600 | 51.9% |
| Fayetteville, NC | 101,831 | 45.3% |
| Kennewick-Richland, WA | 101,547 | 48.6% |
| Norwich-New London-Willimantic, CT | 101,175 | 45.2% |
| Burlington-South Burlington, VT | 101,120 | 59.6% |
| Longview, TX | 100,592 | 52.1% |
| Greeley, CO | 99,469 | 53.9% |
| San Luis Obispo-Paso Robles, CA | 99,466 | 66.3% |
| Amarillo, TX | 98,649 | 52.3% |
| Columbus, GA-AL | 98,196 | 41.8% |
| St. Cloud, MN | 98,182 | 49.1% |
| Tyler, TX | 98,083 | 49.2% |
| Ocala, FL | 97,716 | 51.6% |
| Kingsport-Bristol, TN-VA | 97,395 | 43.8% |
| Lynchburg, VA | 97,081 | 46.3% |
| Hagerstown-Martinsburg, MD-WV | 95,750 | 40.4% |
| Utica-Rome, NY | 94,440 | 48.4% |
| Midland, TX | 94,277 | 59.1% |
| Oshkosh-Neenah, WI | 90,158 | 39.9% |
| Columbia, MO | 89,736 | 45.0% |
| Crestview-Fort Walton Beach-Destin, FL | 89,676 | 56.1% |
| Macon-Bibb County, GA | 89,072 | 42.4% |
| Bend, OR | 87,450 | 66.5% |
| Tuscaloosa, AL | 87,005 | 45.8% |
| Gainesville, GA | 86,885 | 42.8% |
| Charlottesville, VA | 85,356 | 56.4% |
| Slidell-Mandeville-Covington, LA | 85,333 | 57.4% |
| Topeka, KS | 84,424 | 48.8% |
| College Station-Bryan, TX | 82,971 | 53.2% |
| Florence, SC | 82,329 | 41.9% |
| Fort Smith, AR-OK | 80,874 | 47.3% |
| Springfield, IL | 80,458 | 47.3% |
| Laredo, TX | 79,662 | 58.5% |
| Olympia-Lacey-Tumwater, WA | 79,575 | 57.3% |
| Lake Charles, LA | 79,203 | 54.0% |
| Lafayette-West Lafayette, IN | 78,998 | 44.4% |
| Charleston, WV | 78,935 | 44.1% |
| Binghamton, NY | 78,629 | 47.3% |
| Santa Cruz-Watsonville, CA | 78,616 | 70.1% |
| Bellingham, WA | 78,565 | 68.0% |
| Billings, MT | 78,459 | 58.6% |
| Clarksville, TN-KY | 77,809 | 46.6% |
| Medford, OR | 77,479 | 59.8% |
| Champaign-Urbana, IL | 77,346 | 48.1% |
| Saginaw, MI | 77,307 | 44.6% |
| Barnstable Town, MA | 76,094 | 67.8% |
| Iowa City, IA | 75,684 | 41.0% |
| La Crosse-Onalaska, WI-MN | 75,342 | 48.4% |
| Bloomington, IL | 74,674 | 37.0% |
| Joplin, MO-KS | 74,638 | 42.4% |
| Eau Claire, WI | 74,114 | 52.6% |
| Waterloo-Cedar Falls, IA | 74,069 | 47.6% |
| Jackson, TN | 72,883 | 44.2% |
| Monroe, LA | 72,468 | 59.9% |
| Yakima, WA | 71,783 | 59.8% |
| Athens-Clarke County, GA | 71,337 | 57.5% |
| Daphne-Fairhope-Foley, AL | 69,115 | 61.3% |
| Sioux City, IA-NE-SD | 68,780 | 47.1% |
| Johnson City, TN | 68,034 | 44.7% |
| Racine-Mount Pleasant, WI | 67,644 | 53.1% |
| Hilton Head Island-Bluffton-Port Royal, SC | 67,170 | 66.7% |
| Idaho Falls, ID | 67,129 | 54.4% |
| St. George, UT | 67,024 | 69.2% |
| Houma-Bayou Cane-Thibodaux, LA | 66,214 | 61.3% |
| Prescott Valley-Prescott, AZ | 65,697 | 61.2% |
| Wausau, WI | 65,241 | 46.5% |
| Napa, CA | 65,151 | 65.9% |
| Bowling Green, KY | 64,581 | 42.0% |
| Kenosha, WI | 64,321 | 42.4% |
| Greenville, NC | 64,149 | 45.2% |
| Panama City-Panama City Beach, FL | 63,787 | 53.7% |
| Coeur d'Alene, ID | 63,347 | 65.1% |
| Abilene, TX | 62,813 | 49.8% |
| Chico, CA | 62,659 | 59.7% |
| Bremerton-Silverdale-Port Orchard, WA | 62,479 | 64.8% |
| Odessa, TX | 61,338 | 56.1% |
| Janesville-Beloit, WI | 61,182 | 46.2% |
| Rapid City, SD | 59,946 | 63.6% |
| Dalton, GA | 59,912 | 42.4% |
| Kahului-Wailuku, HI | 59,759 | 57.6% |
| Bangor, ME | 59,149 | 56.1% |
| Traverse City, MI | 58,957 | 63.8% |
| Dover, DE | 58,874 | 55.9% |
| Logan, UT-ID | 58,562 | 52.5% |
| Bismarck, ND | 58,256 | 53.9% |
| Burlington, NC | 57,917 | 46.9% |
| Winchester, VA-WV | 57,878 | 44.3% |
| Bloomington, IN | 57,316 | 48.2% |
| Bozeman, MT | 56,303 | 74.6% |
| Grand Junction, CO | 55,960 | 58.0% |
| Morgantown, WV | 55,620 | 38.1% |
| Terre Haute, IN | 55,293 | 46.8% |
| Blacksburg-Christiansburg-Radford, VA | 55,143 | 45.7% |
| Jefferson City, MO | 54,963 | 52.4% |
| Auburn-Opelika, AL | 54,752 | 53.0% |
| Las Cruces, NM | 54,657 | 57.1% |
| Dubuque, IA | 54,547 | 51.2% |
| Harrisonburg, VA | 54,233 | 44.8% |
| Hattiesburg, MS | 53,883 | 48.1% |
| Redding, CA | 53,483 | 61.2% |
| Dothan, AL | 53,415 | 52.1% |
| Altoona, PA | 53,254 | 48.6% |
| Missoula, MT | 53,147 | 68.1% |
| Muskegon-Norton Shores, MI | 52,813 | 51.2% |
| Sheboygan, WI | 52,523 | 43.4% |
| Chambersburg, PA | 52,464 | 45.7% |
| Niles, MI | 52,032 | 55.0% |
| Wheeling, WV-OH | 51,369 | 47.7% |
| Lexington Park, MD | 51,360 | 61.5% |
| Yuma, AZ | 51,328 | 50.0% |
| Pueblo, CO | 51,115 | 52.2% |
| Pittsfield, MA | 51,090 | 56.8% |
| Decatur, AL | 50,545 | 48.1% |
| Rocky Mount, NC | 50,447 | 41.3% |
| Jackson, MI | 50,395 | 52.3% |
| Flagstaff, AZ | 50,372 | 57.7% |
| Jonesboro, AR | 50,226 | 48.6% |
| Sebastian-Vero Beach-West Vero Corridor, FL | 50,149 | 59.4% |
| Warner Robins, GA | 49,355 | 49.9% |
| Florence-Muscle Shoals, AL | 48,863 | 48.3% |
| Lebanon, PA | 48,443 | 51.3% |
| Merced, CA | 48,315 | 51.0% |
| Mankato, MN | 48,314 | 51.6% |
| Amherst Town-Northampton, MA | 48,146 | 54.2% |
| Ithaca, NY | 48,098 | 37.6% |
| St. Joseph, MO-KS | 47,855 | 42.1% |
| Lake Havasu City-Kingman, AZ | 47,559 | 57.5% |
| Battle Creek, MI | 47,549 | 40.6% |
| Santa Fe, NM | 47,378 | 64.5% |
| Alexandria, LA | 47,118 | 57.2% |
| Wichita Falls, TX | 46,664 | 53.4% |
| Texarkana, TX-AR | 46,422 | 54.2% |
| Vineland, NJ | 46,006 | 53.5% |
| Columbus, IN | 45,508 | 37.5% |
| Williamsport, PA | 45,401 | 50.5% |
| State College, PA | 44,379 | 59.0% |
| Owensboro, KY | 44,049 | 45.1% |
| Albany, GA | 44,039 | 49.6% |
| Johnstown, PA | 43,953 | 56.7% |
| Valdosta, GA | 43,952 | 52.6% |
| Kingston, NY | 43,813 | 69.4% |
| Decatur, IL | 43,742 | 39.2% |
| Mount Vernon-Anacortes, WA | 43,347 | 59.7% |
| Lewiston-Auburn, ME | 42,989 | 53.6% |
| Lima, OH | 42,814 | 46.9% |
| Cleveland, TN | 42,200 | 43.0% |
| Sandusky, OH | 42,057 | 50.1% |
| Fond du Lac, WI | 41,989 | 52.5% |
| Twin Falls, ID | 41,927 | 56.0% |
| Punta Gorda, FL | 41,883 | 55.5% |
| Springfield, OH | 41,763 | 45.1% |
| Staunton-Stuarts Draft, VA | 41,755 | 46.9% |
| San Angelo, TX | 41,468 | 53.6% |
| Wenatchee-East Wenatchee, WA | 41,248 | 64.2% |
| Sherman-Denison, TX | 41,183 | 48.9% |
| Morristown, TN | 40,949 | 40.2% |
| Paducah, KY-IL | 40,754 | 53.6% |
| Grand Forks, ND-MN | 40,360 | 56.1% |
| Cape Girardeau, MO-IL | 40,333 | 53.4% |
| Muncie, IN | 40,300 | 45.1% |
| Ames, IA | 39,689 | 56.8% |
| Glens Falls, NY | 39,627 | 57.8% |
| Rome, GA | 39,305 | 39.1% |
| Salisbury, MD | 39,294 | 56.6% |
| Hammond, LA | 39,114 | 57.8% |
| Mansfield, OH | 39,077 | 52.9% |
| Albany, OR | 38,716 | 53.6% |
| Elizabethtown, KY | 38,079 | 45.5% |
| Manhattan, KS | 37,977 | 62.4% |
| Jacksonville, NC | 37,904 | 54.7% |
| Lawrence, KS | 37,284 | 62.0% |
| Yuba City, CA | 36,945 | 58.2% |
| Monroe, MI | 36,499 | 51.7% |
| Grand Island, NE | 36,009 | 46.7% |
| Cheyenne, WY | 35,586 | 56.9% |
| Hot Springs, AR | 35,281 | 56.3% |
| Pinehurst-Southern Pines, NC | 35,146 | 52.0% |
| Kankakee, IL | 35,013 | 43.0% |
| Midland, MI | 34,863 | 40.1% |
| Farmington, NM | 34,387 | 58.8% |
| Anniston-Oxford, AL | 34,206 | 45.5% |
| El Centro, CA | 33,992 | 54.4% |
| Lawton, OK | 33,819 | 49.7% |
| Michigan City-La Porte, IN | 33,755 | 52.7% |
| Brunswick-St. Simons, GA | 33,489 | 57.6% |
| Longview-Kelso, WA | 33,354 | 58.1% |
| Weirton-Steubenville, WV-OH | 32,907 | 45.1% |
| Beckley, WV | 32,640 | 49.4% |
| Goldsboro, NC | 32,595 | 50.4% |
| Victoria, TX | 32,229 | 53.5% |
| Casper, WY | 31,750 | 63.5% |
| Wildwood-The Villages, FL | 31,573 | 50.6% |
| Homosassa Springs, FL | 31,186 | 55.8% |
| Sumter, SC | 30,656 | 45.1% |
| Great Falls, MT | 30,571 | 56.0% |
| Kokomo, IN | 29,731 | 38.7% |
| Helena, MT | 29,514 | 63.6% |
| Watertown-Fort Drum, NY | 28,821 | 59.0% |
| Parkersburg-Vienna, WV | 28,729 | 48.1% |
| Elmira, NY | 28,516 | 53.3% |
| Bay City, MI | 28,314 | 61.4% |
| Pocatello, ID | 27,944 | 55.2% |
| Gettysburg, PA | 27,639 | 59.8% |
| Gadsden, AL | 27,628 | 56.1% |
| Corvallis, OR | 26,893 | 60.6% |
| Hanford-Corcoran, CA | 26,796 | 44.9% |
| Minot, ND | 26,609 | 61.2% |
| Sierra Vista-Douglas, AZ | 26,000 | 58.0% |
| Fairbanks-College, AK | 25,941 | 67.9% |
| Grants Pass, OR | 23,906 | 64.5% |
| Carson City, NV | 22,997 | 58.2% |
| Lewiston, ID-WA | 22,926 | 53.0% |
| Sebring, FL | 22,520 | 55.1% |
| Walla Walla, WA | 21,382 | 48.9% |
| Enid, OK | 20,387 | 58.0% |
| Hinesville, GA | 14,855 | 37.6% |
| Eagle Pass, TX | 12,294 | 48.5% |
Across all 387 metro areas in the Census file, 47.5% of jobs are at companies with fewer than 500 employees. That figure covers metro areas only, so it is different from a national total.
Source2022 SUSB MSA 3-digit NAICS dataset ↗Download the data ↓
A business that outgrows “small” is a success.
If a Portland company grows past 500 employees, small-business policy has not failed. Large employers buy from small suppliers, pay wages that customers spend locally and train people who later start companies of their own. Small companies help large ones adapt and reach specialized markets. Each can also hurt the other, through market power or slow payment.
The eight other metros are different kinds of places. Seattle has very large technology employers. Sacramento has a large public sector. Austin and Salt Lake City have grown along their own paths. Adjusting for industry mix removes one of those differences and leaves the rest. Government jobs are not in this data at all.
Since 2019
There is no single Portland recovery.
Jobs at small businesses inside the city were up a little between 2019 and 2024. Private-sector jobs across Multnomah County were down 7% between 2019 and 2025. The two numbers cover different places, businesses and years, so they don’t contradict each other.
Small-business jobs in the city grew 3.3% from 2019 to 2024.
City of Portland. Jobs at small businesses in 2019 and 2024, as reported by Prosper Portland.
Both bars start at zero and use the same scale. Two yearly totals can’t show which businesses survived, or how small businesses compare with large ones inside the city.
From Prosper Portland’s Insights & Indicators 2025, Figure 1.06, PDF page 17. The report describes these businesses as having 1 to 20 employees in one place and 1 to 19 in another, and it is unclear whether it counts whole companies or individual locations. We have not seen the city data behind the figure.
Thirteen of 18 industries had fewer jobs in 2025 than in 2019.
Multnomah County, private-sector jobs. Each circle is an industry. Farther right means it added jobs. Higher means it is a bigger part of the economy here than nationally.
67,288 jobs in 2019, 75,064 in 2025
- Share of jobs here vs. U.S.
- 1.03×
- Average yearly pay
- $73,905
Bigger circles have more jobs. Green industries pay more than the county’s private-sector average. Rust industries pay less.
See the numbers in a table
| Industry | 2019 jobs | 2025 jobs | Change | Share of jobs vs. U.S. | Average pay, 2025 |
|---|---|---|---|---|---|
| Agriculture | 1,563 | 1,640 | +4.9% | 0.42× | $57,142 |
| Utilities | 1,331 | 1,538 | +15.6% | 0.8× | $152,496 |
| Construction | 25,393 | 23,706 | −6.6% | 0.91× | $108,813 |
| Manufacturing | 35,565 | 28,155 | −20.8% | 0.71× | $78,155 |
| Wholesale | 23,419 | 20,846 | −11.0% | 1.09× | $96,572 |
| Retail | 42,841 | 37,024 | −13.6% | 0.76× | $48,312 |
| Transportation | 28,476 | 32,116 | +12.8% | 1.53× | $68,623 |
| Information | 12,354 | 10,926 | −11.6% | 1.21× | $170,633 |
| Finance | 19,530 | 14,794 | −24.2% | 0.74× | $135,598 |
| Real estate | 10,504 | 9,511 | −9.5% | 1.24× | $76,498 |
| Professional services | 40,624 | 39,370 | −3.1% | 1.15× | $126,524 |
| Corporate offices | 19,284 | 15,591 | −19.2% | 1.88× | $137,460 |
| Administrative services | 24,913 | 23,700 | −4.9% | 0.83× | $59,049 |
| Education | 12,181 | 10,579 | −13.2% | 1.03× | $58,584 |
| Health care | 67,288 | 75,064 | +11.6% | 1.03× | $73,905 |
| Arts and recreation | 8,084 | 8,423 | +4.2% | 0.99× | $65,632 |
| Hotels and restaurants | 51,021 | 43,456 | −14.8% | 0.97× | $35,191 |
| Other services | 22,396 | 19,146 | −14.5% | 1.29× | $60,943 |
The vertical scale compares an industry’s share of county jobs with its share of U.S. jobs. Above 1.0 means the industry is a bigger part of the economy here. It says nothing about productivity. Pay is the yearly average across all jobs in the industry and is not adjusted for inflation. Businesses of every size are included. The chart shows the 18 industries with more than 1,000 jobs in the county.
Sources2019 QCEW county annual bulk ↗2025 QCEW county annual bulk ↗QCEW overview ↗Download the data ↓
Health care
Health care added jobs while most industries lost them. That could reflect unmet need, an aging population or large systems buying up practices. It doesn’t tell us whether owning a health care business got any easier.
Manufacturing
The county lost about one manufacturing job in five. We would need data on individual plants to know how much came from closures, automation, moves out of the county or falling demand.
Hotels and restaurants
The businesses most tied to Portland’s image had fewer jobs than in 2019. New openings matter less than whether customers have come back and whether the margins work.
Our calculations from Bureau of Labor Statistics county data. These figures cover businesses of every size and do not measure the effect of any policy.
Making a living
Can people make a decent living?
This is the question that matters most, and it is where public data is weakest. Some datasets show what a business takes in. Almost none show what the owner keeps, how many unpaid hours went into it or how secure the household is.
What counts is what’s left.
Sales have to cover materials, rent, wages, insurance, software and everything else before the owner is paid. A business with impressive sales can leave very little for the owner’s household. A small professional practice can leave more from far less.
Then there is time and risk. How many unpaid hours did the owner put in? Did a partner’s job provide the health insurance? Did the owner personally guarantee the loan? Did the family cover months of losses? Owning a business can bring freedom and flexibility. It can also push risk and unpaid work onto a family.
Research finds that owners want different things, and many don’t intend to grow much. A good program helps people choose a path that works, which sometimes means deciding not to open. Hurst & Pugsley ↗
The same sales can leave very different take-home pay.
A calculator with made-up numbers. Move the sliders to see what is left for the owner.
$18 for each hour the owner works
The math is monthly sales minus expenses, divided by the owner’s hours in a month. Expenses here don’t include paying the owner. Loan payments, replacing equipment, unpaid help from family and co-owners would all change the answer.
This does not estimate a typical Portland business’s margin, tax bill or hourly pay.
The largest companies pay the most per job.
Portland metro area, 2022. Total yearly pay divided by the number of jobs in mid-March, by company size.
This is a company-size group’s total pay divided by its jobs. It is not the typical worker’s wage. Part-time hours, the kinds of jobs and the industry all affect it, and benefits are left out.
The gap doesn’t prove that company size causes higher pay. Testing that would mean comparing similar jobs in the same industry, with hours, benefits and worker experience taken into account.
Source2022 SUSB MSA 3-digit NAICS dataset ↗Download the data ↓
Don’t forget the people who work there.
A small workplace can mean close relationships, flexibility and a chance to learn the whole business. It can also mean unpredictable schedules, thin benefits, nowhere to advance and nobody to cover when someone is sick. Being local, or being small, doesn’t settle whether the jobs are good.
Local data can’t compare benefits, schedules, injuries or wage violations by company size. That is a reason to go and ask workers. When the public helps a business, it should say plainly what it expects on job quality, in proportion to the help.
A fair evaluation looks at both sides: the owner’s income and hours, and the employees’ pay, hours, stability and benefits. If a program raises sales by leaning on more unpaid owner hours or worse schedules for workers, the headline number overstates the win.
Openings and closings
Openings are only half the story.
A healthy economy needs new businesses, room for good ones to grow and room for people to change course. It also needs to see the closures and layoffs that a positive net number hides.
Each year about one business location in ten is new, and nearly as many close.
Business locations with employees, 2019 to 2023. The chart starts with the Portland metro area. Choose another metro to compare.
See the numbers in a table
| Year | Opened | Share opened | Closed | Share closed |
|---|---|---|---|---|
| 2019 | 6,283 | 9.8% | 5,831 | 9.1% |
| 2020 | 6,196 | 9.6% | 6,154 | 9.6% |
| 2021 | 6,737 | 10.5% | 6,714 | 10.4% |
| 2022 | 7,698 | 11.8% | 6,193 | 9.5% |
| 2023 | 6,911 | 10.5% | 6,588 | 10.0% |
A location is one place of business, so a second branch of an existing company counts as an opening. Each rate divides the year’s openings or closings by the average number of locations in that year and the year before. The rates do not show how long new businesses survive, and some closings are moves.
Sources2023 BDS MSA all-years CSV ↗BDS methodology ↗Download the data ↓
Portland-area employers added 148,949 jobs in 2023 and cut 129,091.
Portland metro area, 2023. Companies of every size.
Jobs added come from new and growing locations. Jobs cut come from closed and shrinking ones. A net gain of about 20,000 jobs sat on top of far more hiring and far more loss than that.
See job changes by company age
| Company age (years) | Jobs | Net job change |
|---|---|---|
| 0 | 18,724 | 18,724 |
| 1 to 5 | 97,901 | 2,241 |
| 6 to 10 | 93,820 | 434 |
| 11+ | 448,276 | -10,174 |
| Founded before the records begin | 446,178 | 8,990 |
The net gain was 19,858 jobs. In the table by company age, the oldest group was founded before the records begin, so its exact age is unknown.
Sources2023 BDS MSA all-years CSV ↗2023 BDS MSA by firm age coarse CSV ↗Download the data ↓
Young and small are two different things.
In 2023 the Portland metro area had 4,472 brand-new companies with employees, accounting for 18,724 jobs. That is a sharper measure of new business than registrations or new branch locations. It still leaves out people who work for themselves.
Research shows that the link between a company’s size and its job growth changes once you account for how old the company is. A twenty-year-old practice with three people and a two-year-old company hiring its tenth employee are both small. They need different things and are likely to contribute differently. Haltiwanger, Jarmin & Miranda ↗
What gets in the way
The right help depends on the business.
Finding customers, covering cash flow and getting through a buildout are three different problems. Before offering more advice, find out which one is actually stopping the business.
The steps to opening depend on the kind of business.
Three simplified examples. Real requirements depend on what the business does and where it is.
Test the idea
Are there enough customers at this price, and enough cash to last until they come?
Check the space
Before signing a lease, confirm the use is legal there, the space is accessible and the utilities can handle it.
Build and get licensed
Line up drawings, permits and construction, plus health approvals if you serve food.
Open and hire
Pass inspection, set up payroll and insurance, and build a schedule staff can rely on.
Stay open
Keep customers coming back, pay for repairs and still pay the owner.
A website can list the requirements. It can’t make the wrong space legal, or pay the bills while an unfinished buildout drags on.
What a delay costs
An example of the bills an owner keeps paying before opening day. Lost profit and construction costs would be on top of this.
These are simplified paths. They are not complete licensing checklists, and they do not measure how long permits take. The delay calculator shows what waiting costs whatever the cause. Not every delay is the government’s doing.
SourcesOSB impact report article ↗Permit Review Dashboard description ↗State Office of Small Business Assistance ↗
Customers
A grant writer cannot create a customer.
In the Federal Reserve’s 2025 survey of small businesses, reaching customers and growing sales was the leading day-to-day challenge. For a neighborhood business, foot traffic, safe access, nearby places people want to go and repeat customers may matter more than another business-plan template.
That survey is national and does not rank Portland’s problems. Local interviews should separate “I need help with marketing” from “the customers who used to buy this are gone.” Federal Reserve ↗
Money
A loan can’t fix a business that loses money.
Working capital can cover a late payment or a slow season. It can’t keep rescuing a business whose ordinary costs are higher than its sales. A loan also has to be repaid, and a personal guarantee puts the owner’s household on the hook.
The financing should fit the problem: a long-term loan for equipment, a credit line backed by reliable invoices, a grant when the public gets something in return, or honest advice to stop. The biggest loan is not always the best result.
Fewer than half of the businesses that applied for financing got all of it.
United States, 2025 Small Business Credit Survey. Businesses with employees that applied for financing.
These figures leave out owners who never applied because they expected to be turned down. Being approved also doesn’t mean the amount, the interest rate or the repayment terms were right for the business.
The survey had 6,525 responses from businesses with employees. It is weighted to resemble businesses nationally but is not a random sample, and it is not specific to Portland.
Source2026 Report on Employer Firms: 2025 Small Business Credit Survey ↗
Permits should be timed from start to finish.
Portland’s permit dashboard reports how many applications come in, how many permits go out, what the projects are worth and how long reviews take. That is useful. But an owner lives through the whole path from finding a space to opening the doors: the meetings before applying, the redesigns, the time spent answering the city’s questions, approvals, construction, inspections and final permission to occupy.
A fair measurement would follow every application, including the ones that were abandoned or withdrawn. It would separate the city’s review time from the applicant’s time and the builder’s, and compare similar kinds of projects. An average that covers only the permits that were issued leaves out everyone who gave up. City dashboard and definitions ↗
Taxes cost time as well as money.
A business can owe very little and still spend hours working out which returns to file. Simpler filing, or exemptions that line up across governments, can help without changing any tax rate. Taxes also pay for services that businesses depend on, so a tax cut should be weighed against what it would stop paying for.
Listening to local businesses
What businesses say is getting in the way
Explore safely summarized feedback by one topic at a time. Voluntary responses describe participants; they do not represent every Portland business.
Patterns are not available yet
There are not enough responses to publish a safe summary for this view. Small groups are withheld to protect privacy.
Portland neighborhood reference
Business location and reported issue location are mapped separately. A business neighborhood is never substituted for an issue location. Neutral official boundary reference; no response counts are shown when a neighborhood view is suppressed. Some association areas overlap or are not covered.
Official neighborhood boundaries ↗The help on offer
Portland offers a lot of help. Is it working?
The city’s Office of Small Business, community groups, lenders and public programs all offer services. Two questions come next. Can people who start from different places actually reach that help? And does it change how things turn out?
unique businesses with action logs in the dashboard supplied October 20, 2026; includes outside-city and unclassified records
October dashboard, p. 1 ↗clients in fiscal year 2024–25. This is a different program and a different period.
Table 4.14 ↗of its clients identified as Black, Indigenous or people of color. 69% were women or gender expansive.
Fiscal year 2024–25 ↗These counts can’t be added together. Some businesses use more than one program, and nobody has counted the eligible businesses that use none.
New evidence · October 20, 2026
A clearer picture of the help, and the next questions to answer.
Prosper Portland shared an aggregate dashboard from OSB’s tracking system. It reports 940 unique businesses with action logs, what clients asked for, how staff responded and feedback from a small survey. This makes the work more visible. It still does not tell us how often a problem was solved, how long that took or what changed for the business.
Data provided by Dan Ruan and Mitch Daugherty, Prosper Portland’s Office of Small Business. Analysis by Portland Civic Lab. Read the supplied dashboard ↗
Read this as a dated snapshot. The engagement chart runs from February 2025 through October 2026, including months before OSB’s reported May 2025 launch. Exact date-filter settings are not shown, and October was still in progress. The earlier year-one report remains below; subtracting its 759 businesses from 940 would not establish how many new businesses were served.
Four counts describe different parts of the work.
OSB dashboard supplied October 20, 2026. These are administrative records, not stages in a measured success funnel.
Engagements created
An engagement record can be different from a business.
With a linked action log
A recorded follow-up, with no measure of completion or speed.
Unique businesses with logs
Provider-reported reach; includes records outside the city.
Action logs
Multiple entries can relate to the same business.
October 2026 is a partial month. Its shorter bar does not establish a fall in demand.
Read the monthly values and the follow-up calculation
| Month | Engagements |
|---|---|
| 2025-02 | 5 |
| 2025-03 | 19 |
| 2025-04 | 61 |
| 2025-05 | 69 |
| 2025-06 | 73 |
| 2025-07 | 53 |
| 2025-08 | 86 |
| 2025-09 | 72 |
| 2025-10 | 81 |
| 2025-11 | 57 |
| 2025-12 | 54 |
| 2026-01 | 53 |
| 2026-02 | 55 |
| 2026-03 | 65 |
| 2026-04 | 62 |
| 2026-05 | 56 |
| 2026-06 | 62 |
| 2026-07 | 38 |
| 2026-08 | 57 |
| 2026-09 | 60 |
| 2026-10 (partial) | 32 |
1,003 ÷ 1,170 = 85.7% with a linked action log, assuming the cards use the same filters. The remaining difference is 167 engagement records. It does not prove that 167 owners were ignored: timing, eligibility, duplicate contacts and logging practices are not explained.
The dashboard also reports 48 businesses with five or more action logs and 22 project-based businesses. Their overlap and the definition of a project are not provided. They cannot be added to 940.
Page 1 reports engagements and businesses; page 2 reports action logs. The 21 monthly engagement counts sum to 1,170. Date filters, business deduplication and the meaning of an engagement need confirmation. A logged action is not necessarily a resolved problem.
SourcePortland Office of Small Business Dashboard, October 20, 2026 ↗Download the data ↓
Funding and business assistance lead the recorded requests.
Counts in the dashboard’s client-needs and service categories, with readable summaries of the category names.
A useful service-design signal. Requests connected to funding and business assistance deserve a closer look. The next distinction is between finding a program, completing an application, receiving money and having enough cash to operate. This dashboard does not measure those separate results.
A smaller category is not a smaller citywide problem. The 40 public-safety/livability/vandalism entries describe this office’s recorded requests. Owners may take those concerns elsewhere or never contact OSB.
How contact happens, and where referrals lead
The contact-method chart reports 61.5% phone and/or email, 22.3% in-person meetings and 16.2% virtual/phone meetings. These are the source’s categories; the distinction between the two phone labels needs clarification.
Liaison outreach accounts for 30.7% of the inquiry-source chart and OSB appointments for 26.1%. Relationships remain an important route into this service. These percentages have no printed field-level denominator.
Permitting & Development represents 41.4% of the relevant-City-program chart. That is a share within a referral-related chart, not 41.4% of all businesses. It supports examining the permit handoff; it does not quantify permit delays. The service chart’s largest label is truncated, so we have not assumed it has the same definition as the year-one report’s Prosper-resources category.
Needs are on page 1 and services on page 2. The categories sum to 1,716 needs and 3,104 services, exceeding 1,170 engagements and 2,183 logs. They overlap or use different counting rules; these are not exclusive shares of clients. Chart labels are editorial summaries because the PDF cuts off some official field names. Original text is preserved in the download; unspecified funding types and referral destinations remain unknown. The first service count is printed as 1.181k.
SourcePortland Office of Small Business Dashboard, October 20, 2026 ↗Download the data ↓
The geographic detail improves. The totals still need reconciliation.
Businesses with action logs and group events by council district or other category. Counts, not rates of access or measures of need.
Seven more category entries than reported unique businesses.
Two more category entries than reported events.
The new dashboard explicitly includes N/A, empty and outside-city business categories, plus virtual and city-wide events. That is useful context for the older report’s incomplete district totals. It does not retroactively explain the older gaps. District 1 has fewer listed businesses than District 2, but these counts alone cannot establish unequal access or insufficient outreach.
Business categories on page 1 sum to 947 against 940 reported unique businesses. Event categories on page 3 sum to 180 against 178 events. Overlap, classification or filters could explain the differences; the PDF does not establish which. No eligible-business denominator is supplied.
SourcePortland Office of Small Business Dashboard, October 20, 2026 ↗Download the data ↓
Most recorded survey answers are positive. We cannot generalize to everyone served.
Follow-up survey, page 2. Fifty responses received overall; question-level answer counts and the number invited are not supplied.
“I feel supported by the business liaison.”
- Strongly agree
- 60%
- Somewhat agree
- 18%
- Neutral or disagree (combined remainder)
- 22%
“How likely would you recommend OSB?”
- Very likely
- 64%
- Somewhat likely
- 18%
- Neither likely nor unlikely
- 8%
- Very unlikely
- 10%
This is encouraging feedback from people who answered, including a visible dissatisfied group. The next step is to learn whether nonrespondents had different experiences, and whether feeling supported led to a completed task.
What the language and appointment demographics can tell us
The business-language chart shows 92.6% English. The dashboard also reports 315 appointments, with demographic charts beneath that heading. Printed shares include 17.5% Black or African American, 9.0% Asian, 42.7% White and 11.9% preferring not to disclose; the gender chart shows 51.5% women and 40.4% men.
These are useful starting points for an access review, but the PDF does not state each chart’s denominator, whether repeat appointments count again or how missing answers are handled. They cannot describe all 940 businesses or establish equal access. We preserve the printed figures without filling in unlabeled slices or inferring identities.
Positive shares add the two positive categories: 60% + 18% = 78% supported; 64% + 18% = 82% likely to recommend. The other 22% on the support question combines neutral and negative answers whose individual percentages are not printed. Fifty divided by 940 is not a valid response rate without an invitation list and a matched cohort. These are experience measures, not evidence of economic impact.
SourcePortland Office of Small Business Dashboard, October 20, 2026 ↗Download the data ↓
Does the help leave a business better off?
A referral can be useful, but an owner’s problem may remain unresolved. For someone trying to open a shop, the result that matters is being able to open. For an established business, it may be steadier cash flow, fewer hours lost to paperwork or better jobs. Different needs call for different measures of success.
The dashboard shows that assistance is happening and that most recorded survey answers are positive. It does not establish how often problems are resolved, whether the benefits last or how much would have happened without public help. Those gaps leave Portlanders without enough evidence to judge which services work best or whether the city is spending the right amount.
Read what the dashboard can and cannot tell us ↓ · Download the documented extract ↓
Earlier evidence · The office’s first year
A person who stays with the problem.
OSB launched at Prosper Portland in May 2025. Its first-year report describes a practical role: help an owner find the right bureau, sort out a utility bill, understand permits, reach a business advisor or identify funding. The September 10 announcement puts trust and follow-through at the center of that work. It describes the same businesses and period as the PDF, not a second set of results. Report, p. 2 ↗ Read the announcement ↗
Download the page-by-page source review ↓
Keep the earlier report alongside the October dashboard: its period and counting terms differ. The report is strongest at showing what assistance looks like. An owner can need language support, a document and a useful introduction before a grant application is even possible. The cases below make that work visible. They also help separate a task completed from a business outcome that still needs to be measured.
Information owners can use
The website offers startup instructions, permitting guidance, community connections and resources for managing a business. The report says these were available in five languages.
Bureaus compare notes
Staff describe regular coordination meetings to connect processes that had operated separately and work through business issues together.
Business districts get a hearing
District leaders meet elected officials or their staff. Office hours, webinars, event promotion and visits to businesses provide other ways to reach the office.
Delivery model reported in the manager’s letter, p. 2 ↗. The report also documents an on-site visit with Professional Auto Body & Paint on p. 5 ↗. These describe access and activity; they do not establish time saved or equal access across languages.
Who asks for help—and what the office offers.
Office of Small Business, first-year report. Every category on page 4, with percentages as printed.
Outreach is a route into the system. Liaisons generated 30% of inquiries; appointments accounted for 25%. A website alone would miss how much of this service begins with a person.
The 7% capital category is easy to misread. It means referrals to funding outside Prosper Portland. Prosper’s own grants and loans sit inside the much larger 50% resources category. Neither percentage tells us how many clients actually received money.
Keep the rounding visible. The printed industry values add to 102% plus agriculture’s “under 1%.” We preserve those values; we do not force them to 100% or turn them into exact business counts. Agriculture has no bar because its precise share is not given.
These are the office’s clients, inquiry channels and recorded service categories—not the mix of all Portland businesses. Exact category counts and counting rules are not provided. Open the original charts, p. 4 ↗
SourceOffice of Small Business Year One Impact Report ↗Download the data ↓
OSB showed up in every district. Its totals still need explaining.
Businesses by district from page 2; group events attended or hosted from page 5. First-year activity, not a measure of need or success.
District 1
- Businesses
- 103
- Group events
- 34
District 2
- Businesses
- 182
- Group events
- 27
District 3
- Businesses
- 154
- Group events
- 33
District 4
- Businesses
- 142
- Group events
- 32
A difference of 178. The report does not explain it.
A difference of 10. The report does not explain this gap either.
The letter’s “more than 135” events is consistent with the 136 headline. The district map is not. Separately, page 2 uses 1,600 touchpoints, while page 3 and the announcement say 1,700+ interactions. The counting terms or periods might differ; we cannot assign a reason from these documents. Announcement ↗
The business counts and event counts measure different things. Neither is divided by a district’s eligible businesses, so these are not rankings of access. Business counts, p. 2 ↗ Event map, p. 5 ↗
SourceOffice of Small Business Year One Impact Report ↗Download the data ↓
Four businesses, four different problems
The work a referral count cannot describe.
These are the four stories OSB selected for its report. They show how support was delivered, not how often it works. Business details and outcomes below are attributed to that report, not independently verified follow-up interviews.
District 1 · Starting childcare
Uluf Hassan
After a decade in early childhood development, Hassan wanted to open a preschool in Division Midway. The report identifies the business as new and home-based childcare; it also describes a search for suitable commercial space.
From an idea to the requirements for opening
At District 1 office hours she met liaison Jon Bebe, Permitting & Development staff and Division Midway Alliance. They connected her with help on childcare requirements, premises that could meet the rules, grant and loan opportunities, business coaching and technical training. She described gaining confidence to navigate zoning, regulations and financing.
What the report establishes: connections and guidance. It does not confirm an opening date, a license or a funding award. Those are the next milestones to follow.
District 2 · Recovering from a fire
One Taekwondo Academy
Vicente founded the studio in 2016 to teach children and adults. An August 2025 fire forced it and three other businesses to relocate or close. Liaison Julieanna Elegant reached out to all four.
Spanish-language help, documents and a new location
OSB helped Vicente communicate with the investigator and obtain the fire report, connected him to the Hispanic Metropolitan Chamber through the Inclusive Business Resource Network, introduced business associations near his old and new locations, and promoted the studio on social media. It also helped with his St. Johns Small Business Grant application.
Vicente reported help getting a sign for the new location and valued being able to work in Spanish: “I could get help directly in Spanish.”
What the report establishes: specific assistance during a move. A grant application is described, not a grant award. It does not isolate how much this help changed the studio’s survival prospects.
District 3 · Buying and adapting a building
Hey Doc
The clinic opened in November 2021, offering chiropractic, pelvic health, acupuncture, massage and mental health care. The report describes growth from one provider and an office manager to 12 providers and five administrative staff, and ownership of its building.
Working through permits with the clinic’s architect
Founder Montserrat first contacted Jon Bebe when OSB launched, then returned for help as the clinic acquired a building in early 2026. Bebe and manager Mitch Daugherty worked with architect Kaeli Nolte of Zone Design Group and Permitting & Development, helped navigate a fee-reduction application, and answered detailed application questions intended to prevent delays.
What the report establishes: coordination and application support. It gives no measured permit-time reduction or fee savings. The clinic’s reported growth is not a count of jobs caused by OSB.
District 4 · Keeping an established restaurant going
Tangier
Founded in 2010, Najia’s downtown family restaurant serves Moroccan and Mediterranean food. What began as a request for help applying for a grant became an ongoing relationship with liaison Julieanna Elegant.
Online ordering, equipment and new customers
OSB connected Najia with the City Revenue Division and Oregon Secretary of State. Xcelerate Women, an Inclusive Business Resource Network partner, helped set up online ordering. The report says she received a Prosper Building Energy Efficiency grant the previous fall and was working out its scope for restaurant equipment. Travel Portland connections offered ways to promote the restaurant to visitors.
What the report establishes: online-ordering help, connections and a reported grant award. It does not give the grant amount, confirm completed upgrades, or measure energy savings or extra sales. Najia emphasized prompt, continuing follow-up in her testimonial.
What the city partners say changed
The report includes statements from three partners. They describe useful ways of working together; they are not independent evaluations.
Transportation
PBOT director Millicent Williams describes clearer construction updates, promotion of events such as Sunday Parkways, and a channel for addressing business concerns early.
PBOT statement, p. 7 ↗Public spaces
The Public Environment Management Office describes mutual referrals, help with commercial-corridor problems and connections through its Problem Solver network.
PEMO statement, p. 9 ↗Permitting
PP&D’s Alice Nielsen describes better access to permitting and property information, and staff learning what owners will encounter. She also says more work remains.
PP&D statement, p. 11 ↗Build on the relationships. Measure what gets finished.
The announcement’s year-two priorities are better coordination, visibility and responsiveness. A useful next report would follow the milestones these stories reveal: a childcare license, a completed move, a permit decision, an installed upgrade and an owner’s time saved. It would include people who stopped seeking help. Year-two direction ↗
This also sharpens the AI question: software might help prepare documents or track a case. It still takes people and institutions to resolve a fire-report request, judge a site’s requirements or coordinate a permit.
Activity is visible. The path to better outcomes is not.
What the October 20 dashboard establishes, next to what would need to be measured to judge the service.
Businesses that need help
Nobody has counted them, including the ones that never ask.
Know about the help and can apply
The dashboard describes client language; awareness, unmet access needs and abandoned applications remain unmeasured here.
Businesses with action logs
October 20, 2026 dashboard snapshot. Provider-reported unique businesses, including outside-city and unclassified records.
Problems solved
Being referred somewhere does not mean the problem was fixed.
Better off afterward
Would need income, hours, survival and job quality, measured the same way for everyone.
These are five things to measure, and the boxes are not drawn to scale. We can’t divide 940 by a county or metro count of businesses to get a share served: geography, eligibility and counting units must match. Survey satisfaction also does not fill the missing problem-resolution box.
The earlier year-one report also has unresolved totals
The report lists 581 businesses across the four council districts and 759 businesses overall, a difference of 178 that it does not explain. It also reports 1,600 “touchpoints” on page 2 and “1,700+ interactions” on page 3. Those may be different terms or different time periods. These are open questions about the reporting. They are not evidence that anyone did anything wrong.
The dashboard adds activity and client feedback. This chain asks whether help reached those who needed it, resolved their problems and improved business outcomes. It is not an observed conversion funnel.
SourcePortland Office of Small Business Dashboard, October 20, 2026 ↗
Who gets the chance to own a business?
Whether someone can start a business depends on savings, collateral, connections, language, time, health, family care and access to customers. A program can serve many owners from underrepresented groups and still miss people who never heard of it, couldn’t meet its conditions or gave up on the application.
The network’s client figures show who its services reach. They don’t show who owns businesses across the city, or whether everyone eligible had the same access. The fair comparison is with the businesses and would-be owners a program is meant for. Comparing with the whole population hides differences in age, work, industry and the barriers that kept people from starting.
Better evidence would combine Census data on owners, city data with identities protected, program records, and interviews with people who didn’t use the programs or were turned down. Small groups would need their privacy protected.
| Who | What they do | What would show it works |
|---|---|---|
| Office of Small Business ↗ | Points a business to the right service or city bureau. | Problems solved, fewer repeat calls and less time from first contact to an answer. |
| Inclusive Business Resource Network and community groups ↗ | Advice, training, trusted relationships and referrals to specialists. | Tasks finished and stronger owner finances, compared with similar businesses that got no help. |
| MESO ↗ / Livelihood NW ↗ / SBDCs ↗ | Different mixes of advice, training and loans. | For each service: who qualifies, who follows through, what it costs and what changes. A client who uses several should be counted once. |
| Loans and grants ↗ | Fill a financing gap, or pay for something the public benefits from. | Investment that would not have happened otherwise, repayment the owner can afford, and a clear answer on who ends up with the benefit. |
| Public contracting ↗ | Connects capable firms with government contracts. | Contracts actually won, how fast they are paid, whether the margins hold up and whether firms win again. |
| Permits and basic city services ↗ | Make it predictable and practical to operate legally. | Time from start to finish, less redone work, safety and services that show up. |
This is a map of who does what. It can’t tell you what you qualify for, so check with each provider. Prosper Portland supplied the October dashboard credited above; the analysis and recommendations here are Portland Civic Lab’s.
Is Portland spending too much or too little?
We can’t tell from the public record. Prosper Portland’s budget also covers real estate, workforce programs and other work. A loan is not the same as a grant. An approval is not the same as money paid out. One-time money is not the same as yearly staffing.
Prosper’s 2025 report cites $37 million in grants and loans approved or awarded to businesses and nonprofits. That is not what small-business help costs in a year. A clear accounting would list staff and contracts, grants paid, loans made and repaid, guarantees, tax breaks and money passed through to other groups, and then match each line to the agency’s actual accounts. Report summary ↗ Fiscal year 2025 financial statements ↗
Then ask what one more dollar would do. Is the shortage advisors, clear eligibility rules, a permit that holds everything up, an affordable buildout or customers? Each option should be compared with the next best use of the money, including simply running everyday city services well.
Happy clients don’t prove a program worked.
The City Auditor’s review of pandemic emergency grants found weaknesses in how awards were decided and in the evidence on results. Those findings are about that program and don’t automatically apply to the Office of Small Business today. The broader lesson is to decide what success means, and who to compare against, before the money goes out. 2021 audit ↗
Training got more people to start businesses without raising what they earned.
Project GATE assigned people to entrepreneurship services at random so the results could be compared fairly. Its long-term report found an early rise in business ownership that faded, and no effect on self-employment earnings. That doesn’t settle the question for every Portland program. It shows why counting new businesses is not enough. U.S. Department of Labor evaluation ↗
A better system
Better help would finish the task, with a person in charge of it.
AI can make some of the work owners now do alone cheaper and faster. The hard part for the public is turning that into finished, reliable work, and fixing the processes that shouldn’t need a helper in the first place.
Start with what the owner is trying to get done.
Picture an owner with a specific goal: open this shop, see next month’s cash, submit this bid or answer this tax notice. The service asks only for the information it needs, checks the rules that apply, prepares the paperwork and puts a named advisor in charge of the next step. The owner can see what is waiting, who has it and when to expect an answer.
AI can draft, sort, translate, organize and check for inconsistencies. A qualified person still has to review anything with real consequences in accounting, law, licensing or lending. Government still makes the official decisions. The owner approves what gets submitted and controls who sees the business’s records.
How many businesses already use AI depends on who is asking. In the Federal Reserve’s 2025 survey, 46% of responding businesses with employees said they use it, and accuracy was a leading concern. The Census Bureau found about 17% to 20% between December 2025 and May 2026, using a different sample and a different question. Neither figure is specific to Portland. Federal Reserve ↗ Census Bureau ↗
How the service would work, step by step.
Our proposal. A person stays responsible at every step.
Understand the need
The owner explains the problem in their own language, online, by phone or in person.
A person checks what the real problem is, how urgent it is and what the owner qualifies for.Do the preparation
Organize the records, lay out the cash flow, find bids the firm could win or fill in an application.
Anything AI prepares is checked against current rules and the owner’s own records.Clear the roadblock
Get the specialist’s answer, the agency’s decision or a complete submission. If none of those is possible, find another route.
Named professionals and bureaus make the decisions. Software does not issue a permit.Confirm it worked
Did the owner finish the task? Was it done correctly? Did it save time or money, or steady the business?
Follow up after the referral and publish results without identifying anyone.These are ideas to test. They draw on existing services in Seattle, Singapore and New York and on audits of them. Those examples show how such services work. They do not prove the same thing would work in Portland.
SourcesSeattle small business consulting programs ↗SMEs Go Digital Advisor announcement ↗CTO-as-a-Service factsheet ↗MyCity system audit ↗
| Task | What AI can help with | What still takes people or policy |
|---|---|---|
| Bookkeeping | Sort documents, suggest categories and flag missing records. | The owner confirming the entries, an accountant’s review, tax decisions and accurate records to begin with. |
| Cash-flow planning | Build what-if scenarios from invoices, regular bills and timing. | Judging future demand, negotiating with creditors, and the cash itself. |
| Bidding on contracts | Find relevant bids, summarize the requirements and build a checklist. | The capacity to do the work, certification, bonding, fair purchasing rules and payment on time. |
| Applications and grants | Reuse verified facts, draft answers, check for gaps and translate. | Clear eligibility rules, simpler forms, open selection and enough funding. |
| Permits | Explain the documented steps and assemble a complete package. | Official interpretations, inspections, professional design and fixing the process itself. |
| Marketing and sales | Draft materials, improve accessibility and help follow up with customers. | An offer that stands out, real trust and customers with money to spend. |
Better applications don’t create more grant money.
If the pot is fixed, better applications mostly make the competition tougher. An application assistant should be compared with a shorter form, a business profile that is verified once and reused, clearer eligibility rules or a simpler way of choosing winners. Time saved for the people who don’t win counts too.
The same goes for navigation. A tool that explains a confusing process is useful. Making the process less confusing may be worth more and last longer.
A wrong answer can be an expensive service.
The New York City Comptroller’s audit of the city’s MyCity chatbot documented inconsistent and inaccurate answers, and it disputed the city’s performance measures. Portland should test any tool on real tasks with independent checks, in several languages, with ambiguous questions and with rules that change. A task that was not resolved should not count as a success because the system replied. NYC Comptroller ↗
Why doesn’t this exist already?
No single institution controls the whole path. The rules sit with different agencies and levels of government. The data is scattered, each funding source has its own eligibility rules, and the group that refers a business often has no say over the next decision. Trusted community groups need steady capacity, and owners have little time to learn unfamiliar tools.
AI makes some information work cheaper. It doesn’t change who is in charge of what, and it doesn’t produce affordable space, working capital or customers with money to spend. A system that works needs agreements between agencies, a budget, usable data and someone answerable for the cases that don’t get solved.
There are tradeoffs too. A subsidized AI service could take work from local bookkeepers, designers and office workers. The city could instead hire those people to supervise and improve the service. That choice should be made in the open. The effects on service quality, workers’ income, competition and customer demand should be tracked along with the time owners save.
Three ways to spend the next dollar.
Three example programs with rough first-year costs. These are our scenarios and not a city budget. Move the sliders to change the assumptions.
Example first-year budget
Give small firms a back office
Human advisors who use AI to help prepare documents and cash-flow plans, with specialists reviewing the work.
- 8 staff × $145,000
- $1,160,000
- Systems, tools and operations
- $300,000
- Language help, quality checks and targeted support
- $280,000
- Evaluation and accessibility
- $150,000
$1,890,000first-year cost
150 more problems solved than would have been without the program. That effect is our assumption. Nobody has measured it.
Who would run itProsper Portland and community groups working under contract.
What has to be in placeOwners’ consent to share records, service in several languages, licensed experts where the law requires them, secure tools and enough staff at the provider groups.
How to test itCompare today’s help, human help with AI and simpler forms. Measure tasks verified as finished, errors, time saved and how businesses are doing after six and twelve months.
These are planning numbers. They are not appropriations or bids, and nobody has proven the effects. The three options overlap, so their results can’t simply be added. Loans, major construction and agency overhead are left out.
Serving more businesses improves the result only if the program solves problems that would not have been solved anyway.
SourcesFramework for Evaluation of SME Policies 2023 ↗Project GATE final evaluation ↗Download the data ↓
What Portland should decide
Aim to be the easiest place to make a living from a business.
Being the best city for small business only means something if owners can feel it and the city can measure it. The goal should reward opportunity people can reach, work that is useful and prosperity that lasts. A rising count of small businesses does none of that.
A predictable start
How long it takes to get from a complete application to opening day or a final decision, for the typical case and for the slowest tenth.
Count projects that were abandoned. Note how complex each one was, how much of the time was the applicant’s, and the safety results.Help that finishes the task
How many more problems get solved because of the help, and what each one costs.
Needs independent quality checks and a fair comparison group.An income owners can live on
What owners earn per hour worked, what they have saved, how much their income swings and how much risk the household carries.
Measure costs and hours. Include businesses that closed and owners who did not answer.Jobs worth having
Hourly pay, benefits, steady hours, how long people stay and whether they move up.
Compare similar industries, jobs and workers.Opportunity people can reach
For each group a program is meant to serve: who knows about it, who applies, who is turned down, who gives up and how they fare.
Compare with the right group, offer help in people’s languages and protect privacy when groups are small.A productive local economy
Value added where it can be measured, customer demand, growth among suppliers and services people use.
Subtract business taken from competitors, gains captured by landlords and costs that fall on residents.First 90 days
Find out where things stand.
Publish one reconciled list of what is actually spent and what services exist. Agree on what counts as a case and what counts as solved. Clear up the Office of Small Business totals and the city’s definition of small. Talk to owners who didn’t use the programs, were turned down or closed. Name the two or three delays owners keep hitting and can’t fix on their own.
Months 4 to 9
Run a small, fair test.
Choose specific tasks that come up often enough to evaluate. Compare today’s help, human help with AI and a simpler process. Decide before launch what error rate is acceptable and when a case goes to a person. Track cost and the owner’s effort from first contact to finish.
Months 10 to 18
Publish the results and change the rules.
Measure results at six and twelve months, and say how many people didn’t respond. Expand what clearly helps. Redesign or stop what doesn’t. Where cases keep failing at the same point, change the permit, purchasing, payment or eligibility rule behind it. Keep the results public.
What would change our minds
If owners lose most of their time to filling out the same paperwork again and again, a shared back office looks better. If workable projects keep failing at the same city-controlled step, fixing that step comes first. If businesses open easily and then can’t find customers, the focus should move to demand, household incomes and the condition of business districts.
If grants mostly move customers from one nearby business to another, the growth claimed for them should be marked down. If landlords capture most of a space subsidy, the program needs a new design. And if a trusted advisor talks an owner out of debt they can’t afford, or out of a business that can’t work, that is a good result even though no new business shows up in the statistics.
Celebrate the business that opens. Then ask whether the people in it can make a living.
Sources and methods
Check our work.
Every source, data file and definition we used is here. Search the sources, download the numbers and tell us what we got wrong.
How we calculated things
We use four federal datasets, and each one counts something different. The Census Bureau’s Statistics of U.S. Businesses gives jobs, pay and sales by company size. Its Nonemployer Statistics covers businesses with no employees. The Bureau of Labor Statistics’ Quarterly Census of Employment and Wages gives yearly average jobs and pay by industry. The Census Bureau’s Business Dynamics Statistics tracks openings, closings and job changes.
Each share is a group’s total divided by the total from the same source. To adjust for industry mix, we take each metro’s small-company share within each industry and weight it by Portland’s jobs in that industry. Where the Census withheld a number, we did not treat it as zero.
What this covers and what it doesn’t
This is not a count of every business in the city, and we have not read everything written on the subject. The list below includes sources we reviewed and leads we have not finished checking. Each one is labeled.
Still missing: what owners earn, benefits by company size, how long city businesses survive, who wins public contracts, a sound city GDP figure and proof of what the programs cause. The October 20 supplement adds aggregate administrative data from Prosper Portland’s Office of Small Business. It is not a representative owner survey or an independent audit.
Five things to keep in mind when reading the numbers
- Place. The city, the county and the metro area have different boundaries. A Portland mailing address does not mean a business is inside the city.
- What is being counted. A registered business may not be operating. A location is not a company. A job is not a full-time worker. Sales are not profit and are not GDP.
- Dates. Each chart uses the year its data was collected, and the newest year differs by source. Pay figures are not adjusted for inflation, so they can’t show real gains.
- Cause and effect. Businesses choose to join programs, some results would have happened anyway, and surveys miss the people who don’t answer. A claimed effect needs something to compare against.
- Loose ends. Official reports sometimes contradict themselves. The Office of Small Business totals, the Chamber’s 20% and 28%, and the city’s definition of small stay flagged here until someone resolves them.
73 of 73 sources. Being listed here does not mean we verified every finding in a source.
US Census2022 SUSB MSA 3-digit NAICS datasetReviewed and archived
Enterprise size; revised MSA dataset; SHA in checks.json
- Place
- US metropolitan areas
- Period covered
- 2022
- Published
- 2025-07 revision
- Where in the source
- c01: MSA 38900 NAICS -- ENTRSIZE 01 | c02: MSA 38900 NAICS -- ENTRSIZE 05 | c03: MSA 38900 NAICS -- ENTRSIZE 08 | c04: MSA 38900 NAICS -- ENTRSIZE 08/01 PAYR | c05: MSA 38900 NAICS -- ENTRSIZE 08/01 RCPT | c09: MSA 38900 NAICS 54 and 72 ENTRSIZE 01 | c10: MSA 38900 NAICS 62 ENTRSIZE 01
- Method and limits
- Enterprise size; revised MSA dataset; SHA in checks.json
US Census2022 SUSB county enterprise size workbookReviewed and archived
Multnomah County extract; SHA in checks.json
- Place
- US counties
- Period covered
- 2022
- Published
- 2025-05-29
- Where in the source
- c06: State 41 County 051 NAICS -- Enterprise Size 1 | c07: State 41 County 051 NAICS -- Enterprise Size 2+3+4 | c08: State 41 County 051 NAICS -- Enterprise Size 2
- Method and limits
- Multnomah County extract; SHA in checks.json
US CensusSUSB methodologyReviewed
Employer universe and enterprise versus establishment
- Place
- US
- Period covered
- 2022
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Employer universe and enterprise versus establishment
US Census2022 SUSB dataset noticeReviewed
Revised MSA 3-digit file notice
- Place
- US
- Period covered
- 2022
- Published
- 2025-08-04
- Where in the source
- c27: Notice Revision of Dataset
- Method and limits
- Revised MSA 3-digit file notice
US Census2023 Nonemployer Statistics county fileReviewed and archived
Cached pinned raw in maker-economy corpus; receipts are not profit
- Place
- US counties
- Period covered
- 2023
- Published
- 2025
- Where in the source
- c11: ST 41 CTY 051 NAICS 00
- Method and limits
- Cached pinned raw in maker-economy corpus; receipts are not profit
US CensusNonemployer Statistics methodologyReviewed
Mailing address and threshold limitations
- Place
- US
- Period covered
- 2023
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Mailing address and threshold limitations
US CensusNonemployer Statistics by Demographics APISkimmed, not fully reviewed
Potential owner demographics
- Place
- US; subnational
- Period covered
- 2023
- Published
- 2025
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
US CensusAnnual Business SurveyNot yet reviewed
Employer ownership and innovation; check metro coverage
- Place
- US; subnational
- Period covered
- multiple
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
US CensusCounty Business PatternsSkimmed, not fully reviewed
Establishments not enterprises
- Place
- county; metro; ZIP
- Period covered
- multiple
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
US CensusBusiness Dynamics StatisticsSkimmed, not fully reviewed
Age; size; births; deaths
- Place
- US; metro
- Period covered
- multiple
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
US Census2023 BDS data filesSkimmed, not fully reviewed
MSA age and size tables
- Place
- US; metro
- Period covered
- 2023
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
US CensusBusiness Formation StatisticsNot yet reviewed
Applications are not firm births
- Place
- US; state; county
- Period covered
- multiple
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
US BLS2019 QCEW county annual bulkReviewed and archived
Original cached in maker-economy corpus
- Place
- Multnomah County
- Period covered
- 2019
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Original cached in maker-economy corpus
US BLS2025 QCEW county annual bulkReviewed and archived
Covered annual-average payroll jobs
- Place
- Multnomah County
- Period covered
- 2025
- Published
- n.d.
- Where in the source
- c12: area_fips 41051 own_code 5 industry_code 10 size_code 0 qtr A | c13: area_fips 41051 own_code 5 industry_code 62 size_code 0 qtr A | c14: area_fips 41051 own_code 5 industry_code 31-33 size_code 0 qtr A
- Method and limits
- Covered annual-average payroll jobs
US BLSQCEW overviewReviewed
Excludes most proprietor work
- Place
- US; county
- Period covered
- multiple
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Excludes most proprietor work
Oregon Employment DepartmentOregon Employment Department data hubSkimmed, not fully reviewed
Regional employment and wages
- Place
- Oregon; county; metro
- Period covered
- current
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
Oregon Employment DepartmentBusiness Formation and Business Employment Dynamics in OregonSkimmed, not fully reviewed
Context for formation and UI coverage
- Place
- Oregon; metro
- Period covered
- multiple
- Published
- 2026-02-05
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
US BEAGDP by County 2024 releaseReviewed
No firm-size split
- Place
- US counties
- Period covered
- 2024
- Published
- 2026-02-05
- Where in the source
- No claim extracted yet
- Method and limits
- No firm-size split
US BEASmall Business satellite-account program statusReviewed
Experimental national series discontinued as regular output
- Place
- US
- Period covered
- multiple
- Published
- 2026-02
- Where in the source
- c25: Opening notice
- Method and limits
- Experimental national series discontinued as regular output
US BEAGeographic aggregator technical documentationSkimmed, not fully reviewed
Need current-dollar sum and real GDP aggregation method
- Place
- county-derived metros
- Period covered
- 2024
- Published
- 2026-02
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
US SBA Office of AdvocacySmall Businesses Generate 44 Percent of U.S. Economic ActivitySkimmed, not fully reviewed
Historical national finding; not Portland allocation
- Place
- US
- Period covered
- 1998-2014
- Published
- 2019-01-30
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
Prosper PortlandOffice of Small Business Year One Impact ReportReviewed and archived
13-page operator report; full numerical tables, outreach model, four selected business stories and partner statements reviewed October 5, 2026
- Place
- Portland city
- Period covered
- 2025-05 to 2026-05
- Published
- 2026-09-10
- Where in the source
- Pages 2-5: launch, service model, business/intake/industry/service/event counts; pages 6/8/10/12: owner stories; pages 7/9/11: partner statements; page 13: help entry point
- Method and limits
- Self-reported outputs and selected testimonials, not causal outcomes; district businesses total 581 vs 759, events 126 vs 136; 1600 touchpoints vs 1700+ interactions; rounded industry shares exceed 100 percent
Prosper PortlandOSB impact report articleReviewed
Article says 1700+ interactions
- Place
- Portland city
- Period covered
- 2025-2026
- Published
- 2026-09-10
- Where in the source
- September 10 announcement: same first-year report, more than 750 businesses and 1700+ follow-up interactions; trust, multilingual access and year-two priorities
- Method and limits
- Same operator and evidence base as PDF, not independent corroboration or a separate reporting period
Prosper PortlandWorkforce and business programs board updateSkimmed, not fully reviewed
621 businesses served; survey response about 30 percent
- Place
- Portland city
- Period covered
- FY2023-24
- Published
- 2025-03-12
- Where in the source
- c22: Inclusive Business Resource Network section | c23: Inclusive Business Resource Network section
- Method and limits
- Not fully reviewed; inspect publisher methods before use
Prosper PortlandProsper Portland FY2025 annual comprehensive financial reportSkimmed, not fully reviewed
Fund and budget context; not all spending targets firms
- Place
- Portland city
- Period covered
- FY2025
- Published
- 2025-11
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
Prosper PortlandInsights and Indicators 2025Reviewed and archived
City small-business employment Figure 1.06 and central-city Figure 1.07; size-band inconsistency
- Place
- Portland city
- Period covered
- 2025
- Published
- 2026-02
- Where in the source
- c35: Figure 1.06 PDF page 17 (printed page 11) | c36: PDF pages 17 and 88 (printed pages 11 and 82)
- Method and limits
- City small-business employment Figure 1.06 and central-city Figure 1.07; size-band inconsistency
City of PortlandFive Year Loan and Grant ActivitySkimmed, not fully reviewed
Geographic spending; do not sum principal with grants
- Place
- Portland city
- Period covered
- FY2019-20 to FY2023-24
- Published
- 2025-01
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
City of PortlandFY2026-27 proposed Prosper budget reportSkimmed, not fully reviewed
Budget proposal not actual expenditures
- Place
- Portland city
- Period covered
- FY2026-27
- Published
- 2026
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
Portland Civic LabPortland City Performance DashboardSkimmed, not fully reviewed
City performance indicator republished; trace back to original
- Place
- Portland city
- Period covered
- multiple
- Published
- 2026
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
City AuditorEmergency grants auditReviewed
Selection and outcome assessment findings
- Place
- Portland city
- Period covered
- 2020
- Published
- 2021-05-12
- Where in the source
- c24: Outcome assessment section
- Method and limits
- Selection and outcome assessment findings
City AuditorProsper Portland 2022 audit follow-upSkimmed, not fully reviewed
Follow-up status; historic program
- Place
- Portland city
- Period covered
- 2022
- Published
- 2022
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
City AuditorOngoing City Auditor workSkimmed, not fully reviewed
Grants-management audit unfinished at cutoff
- Place
- Portland city
- Period covered
- 2026
- Published
- 2026
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
City of PortlandPermit Review Dashboard descriptionReviewed
Commercial review times; application cohorts needed
- Place
- Portland city
- Period covered
- current
- Published
- 2026
- Where in the source
- c26: Understanding commercial and residential permits
- Method and limits
- Commercial review times; application cohorts needed
City of PortlandPermitting MetricsSkimmed, not fully reviewed
Dashboard and performance claims
- Place
- Portland city
- Period covered
- current
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
City of PortlandBusiness Tax Filing and Payment InformationSkimmed, not fully reviewed
Net-income tax; distinct from permit
- Place
- Portland city
- Period covered
- current
- Published
- 2026
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
MetroEconomic Value AtlasSkimmed, not fully reviewed
Uses Esri/DataAxle; vendor inventory limits
- Place
- Portland region
- Period covered
- 2025 data
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
City of PortlandPlanning maps and city boundariesSkimmed, not fully reviewed
City limits and business district boundaries
- Place
- Portland city
- Period covered
- current
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
Oregon Secretary of StateOregon business statistics and filingsSkimmed, not fully reviewed
Registration is not operating business
- Place
- Oregon
- Period covered
- current
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
Oregon Secretary of StateActive Businesses Socrata datasetNot yet reviewed
Entity types and mailing address require filtering
- Place
- Oregon
- Period covered
- current
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
Portland Metro Chamber2023 State of Small BusinessReviewed
One page states both 20 and 28 percent for metro share
- Place
- Multnomah; metro
- Period covered
- 2017-2022
- Published
- 2023-09
- Where in the source
- c20: Report at a Glance and What the data shows | c21: Introduction
- Method and limits
- One page states both 20 and 28 percent for metro share
City of PortlandCouncil acceptance of Chamber reportSkimmed, not fully reviewed
Uses 28 percent metro and 35.1 percent county; under 50 definition
- Place
- Portland city
- Period covered
- 2022
- Published
- 2024
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
Bricks Need MortarWe Are Here 2025Skimmed, not fully reviewed
Respondent survey; check recruitment and sample
- Place
- Portland brick-and-mortar
- Period covered
- 2024
- Published
- 2025-03-24
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
Downtown Portland Clean & SafeDowntown Portland research reportsSkimmed, not fully reviewed
Foot traffic and business survey; not citywide
- Place
- Central city
- Period covered
- multiple
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
City of PortlandLegacy businesses study announcementSkimmed, not fully reviewed
Succession and displacement
- Place
- Portland city
- Period covered
- 2026
- Published
- 2026-03-20
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
US SBA7a and 504 FOIA dataSkimmed, not fully reviewed
Loan approval is not additional economic impact
- Place
- US; loan level
- Period covered
- through 2026-06
- Published
- 2026-04 update
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
FFIECCRA public data productsSkimmed, not fully reviewed
Reporting lenders only; loan amount is not small-firm count
- Place
- US; county; metro
- Period covered
- 1996-2024
- Published
- 2026-02-23
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
Federal Reserve SystemSmall Business Credit Survey reportsSkimmed, not fully reviewed
Sample and weighting limitations
- Place
- US; selected metros
- Period covered
- 2025
- Published
- 2026
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
Business OregonOregon State Small Business Credit InitiativeSkimmed, not fully reviewed
Capital and technical assistance
- Place
- Oregon
- Period covered
- current
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
Port of PortlandPort of Portland small business enterprise programSkimmed, not fully reviewed
Procurement goal and eligibility
- Place
- Portland region
- Period covered
- current
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
City of PortlandPortland 2026 disparity studySkimmed, not fully reviewed
Study underway at cutoff; no findings yet
- Place
- Portland region
- Period covered
- 2026
- Published
- 2026
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
Oregon Secretary of StateState Office of Small Business AssistanceSkimmed, not fully reviewed
State uses 0–100 employee definition
- Place
- Oregon
- Period covered
- current
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
OECDFramework for Evaluation of SME Policies 2023Skimmed, not fully reviewed
Evaluation framework not Portland impact evidence
- Place
- international
- Period covered
- 2023
- Published
- 2023-02-13
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
US Department of LaborProject GATE final evaluationSkimmed, not fully reviewed
Random-assignment entrepreneurship support study
- Place
- US study sites
- Period covered
- 2002-2009
- Published
- 2010-08-24
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
NBER; original authorsWho Creates Jobs? Small vs Large vs YoungReviewed
Size-growth association changes after age adjustment
- Place
- US
- Period covered
- multiple
- Published
- 2010-2012
- Where in the source
- No claim extracted yet
- Method and limits
- Size-growth association changes after age adjustment
NBER; original authorsWhat Do Small Businesses Do?Skimmed, not fully reviewed
Many owner goals do not involve high growth
- Place
- US
- Period covered
- multiple
- Published
- 2011
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
City of SeattleSeattle small business consulting programsSkimmed, not fully reviewed
Accounting legal permitting and space support
- Place
- Seattle
- Period covered
- 2026
- Published
- 2026-01-15
- Where in the source
- c29: Helping Seattle small businesses grow
- Method and limits
- Not fully reviewed; inspect publisher methods before use
IMDASMEs Go Digital Advisor announcementSkimmed, not fully reviewed
Existing AI-enabled and human-guided shared services
- Place
- Singapore
- Period covered
- 2025
- Published
- 2025
- Where in the source
- c28: SMEs Go Digital Day
- Method and limits
- Not fully reviewed; inspect publisher methods before use
IMDACTO-as-a-Service factsheetSkimmed, not fully reviewed
Digital strategy advice and solution discovery
- Place
- Singapore
- Period covered
- 2021
- Published
- 2021-03-02
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
NYC ComptrollerMyCity system auditSkimmed, not fully reviewed
AI business guidance accuracy and oversight
- Place
- New York City
- Period covered
- 2025
- Published
- 2025-12-30
- Where in the source
- c30: Audit Summary of Findings
- Method and limits
- Not fully reviewed; inspect publisher methods before use
US CensusBusiness Trends and Outlook Survey downloadsSkimmed, not fully reviewed
AI adoption and business conditions
- Place
- US; selected geographies
- Period covered
- 2023-2026
- Published
- 2026
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
Federal Reserve Bank of San FranciscoAI adoption in small businessesSkimmed, not fully reviewed
Qualitative adoption findings
- Place
- US
- Period covered
- 2024 survey
- Published
- 2026-07
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
Oregon SBDC networkOregon SBDC center finderSkimmed, not fully reviewed
PCC local center; no reach or outcomes extracted
- Place
- Oregon
- Period covered
- current
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
MESOMESO services and lendingSkimmed, not fully reviewed
Provider self-description; verify contracts and results
- Place
- Oregon and Southwest Washington
- Period covered
- current
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
Livelihood NWLivelihood NW servicesSkimmed, not fully reviewed
Provider-reported reach; not city-unique or causal
- Place
- Oregon and Southwest Washington
- Period covered
- 2025
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Not fully reviewed; inspect publisher methods before use
US Census2023 BDS MSA all-years CSVReviewed and archived
2019-23 metro entry exit and job flows; raw hash in checks.json
- Place
- US metropolitan areas
- Period covered
- 1978-2023
- Published
- 2025-11-25
- Where in the source
- c31: year 2023 msa 38900 estabs_entry and estabs_exit
- Method and limits
- 2019-23 metro entry exit and job flows; raw hash in checks.json
US Census2023 BDS MSA by firm age coarse CSVReviewed and archived
2023 Portland MSA firm-age extract; left-censored age category
- Place
- US metropolitan areas
- Period covered
- 1978-2023
- Published
- 2025-11-25
- Where in the source
- c32: year 2023 msa 38900 fagecoarse a) 0 firms
- Method and limits
- 2023 Portland MSA firm-age extract; left-censored age category
US CensusBDS methodologyReviewed
Establishment entry exit and parent firm-age definitions
- Place
- US
- Period covered
- 1978-2023
- Published
- n.d.
- Where in the source
- No claim extracted yet
- Method and limits
- Establishment entry exit and parent firm-age definitions
City of Portland; Prosper PortlandProsper FY2025-26 proposed budget committee reportReviewed
Proposed $189m agency budget includes real estate and other activities; targets are not actuals
- Place
- Portland city
- Period covered
- FY2025-26
- Published
- 2025-05-13
- Where in the source
- c33: Financial and budgetary impacts paragraph | c34: Key Performance Indicators paragraph
- Method and limits
- Proposed $189m agency budget includes real estate and other activities; targets are not actuals
Federal Reserve Banks2026 Report on Employer Firms: 2025 Small Business Credit SurveyReviewed
6,525 employer-firm responses; weighted convenience sample, not Portland estimates. AI module optional (about 81% response). 42/36/22% financing outcomes among applicants; 46% AI use.
- Place
- United States
- Period covered
- Survey Sep 3–Nov 14, 2025
- Published
- 2026-03-03
- Where in the source
- Web sections Debt and credit demand; Use of artificial intelligence; About survey. PDF p. 28 for AI module caveat.
- Method and limits
- Nonrandom sample; self-reported outcomes; no causal effects.
City of Portland Revenue DivisionBusiness tax filing and payment: gross-receipts exemption thresholdsReviewed
City gross-receipts threshold less than $50,000 before 2026, $75,000 in 2026, $100,000 from 2027. City/county exemption still requires filing.
- Place
- Portland / Multnomah / Metro, distinct tax jurisdictions
- Period covered
- Tax years 2025–2027
- Published
- Living official guidance; accessed 2026-10-03
- Where in the source
- Exemptions section
- Method and limits
- Gross-receipts eligibility is distinct from taxable net-income base; not individualized tax advice.
City of Portland Mayor's OfficeSmall Businesses Get Tax Cut: adopted city threshold changeReviewed
Confirms the threshold changes were adopted before the Sep 30 evidence cutoff.
- Place
- Portland city
- Period covered
- 2026/2027 tax years
- Published
- 2026-04-08
- Where in the source
- Opening paragraph
- Method and limits
- Policy announcement; not an impact evaluation.
US Census BureauLarge Firms With at Least 20 Employees Biggest AI UsersReviewed
BTOS AI use about 17–20%; revised question covers any business function. Not comparable to older goods/services-only question or Fed sample.
- Place
- United States
- Period covered
- December 2025–May 2026
- Published
- 2026-05
- Where in the source
- Article and methodology explanation
- Method and limits
- Different universe/question from SBCS; cannot interpret survey gap as a trend.
Prosper Portland / Office of Small BusinessPortland Office of Small Business Dashboard, October 20, 2026Reviewed and archived
Supplied by Dan Ruan and Mitch Daugherty. All three pages visually reviewed; selected exact values transcribed. Administrative activity and survey feedback, not causal outcomes.
- Place
- OSB service population; includes outside-city and unclassified records
- Period covered
- Snapshot supplied October 20, 2026. Engagement chart spans February 2025–October 2026; precise date filters and cohort definitions are not shown. October is incomplete.
- Published
- 2026-10-20
- Where in the source
- p.1 engagements, client needs, businesses and geography; p.2 action logs, follow-up survey and appointments; p.3 group events
- Method and limits
- Image-only PDF; filters, denominators, deduplication and survey invitation count unspecified. Geography sums exceed headline totals. Truncated labels and unlabeled values retained.
Originally published October 3, 2026, with a September 30 research cutoff. Updated October 27 with the OSB dashboard dated October 20, 2026, and November 6 with the business feedback dashboard. Other economic and policy sections retain their original source dates; this is not a full refresh of every dataset.
