I am unsettled this week. I sat down again and did a piece of arithmetic I have been dodging for about a decade.
Ive been saying for years that the world needs a lot more architects than it has. I say it in talks. I say it in posts like this one. And then someone asked me the obvious question I had never actually answered with a number. How many?
So I went and got the small business data… and a spreadsheet. The number is worse than I thought. And I already thought it was bad.
Start With Who Actually Needs One
There are 36,186,089 small businesses in the United States. The SBA counts them in three buckets and I am going to use their buckets, not mine.
29,811,495 have no employees at all. The owner is the staff. 5,720,093 have between 1 and 19 people. 654,501 have between 20 and 499, and that last bucket is the mid sized world we keep talking about and keep ignoring. Inside it, roughly 555,000 sit between 20 and 99 and about 100,000 sit between 100 and 499.
Every one of them buys technology. Every one of them makes build or buy decisions they are not equipped to make. The dental practice with three chairs and a records system that ate a year of patient history. The construction company that put sensors on every truck because the ROI slide looked good. The regional clinic rolling out connected monitoring because a vendor rep made it sound simple.
And who advises them today? A sales engineer.
I was once told by a leader in a well known company, “We do not care about our customers or how our technology fits their needs. We care about sales.” Direct quote. He was not embarrassed by it.
So Let’s Actually Count
A family doctor carries a panel. So does an architect, whether we admit it or not.
For a truly small business, one architect can carry about 20 of them. A few days each per year, a decision or two a month, someone at the other end of the phone when the point of sale vendor starts talking about a migration. Twenty is not a light load… try it for a year and then tell me Im wrong.
And the panel shrinks fast as the business grows. At 20 to 99 people you are down to about five clients each. At 100 to 499 you dont get a panel at all. You need two architects inside the business, and at the top of that range you need more.
So run it.
- 29,811,495 firms with no employees, served the way a clinic serves walk ins, one architect per 200. Call it 150,000 architects.
- 5,720,093 firms with 1 to 19 people, panel of 20. That is 286,000.
- About 555,000 firms between 20 and 99 people, panel of five. Another 111,000.
- About 100,000 firms between 100 and 499 people, two architects each. That is 200,000.
Roughly 750,000 architects. For the United States alone.
Now the supply side. The Bureau of Labor Statistics counts 179,200 computer network architects in the entire country, and that is a networking title doing networking work. There is no federal occupational code for what we actually do. Business technology architect is not a job the government believes exists.
Now look at what this country built for the other professions a small business depends on.
Accountants and auditors, 1,579,800 of them. Actively licensed CPAs, 653,408. Civil engineers, 368,900, and the structural engineers who stamp the drawings are a licensed group inside that number. Licensed building architects, just over 123,000, with about 3,500 new licenses a year.
America decided that every business needs someone who can keep its books honestly. So it made a million and a half of them. It decided buildings should not fall on people, so it made an architect and a structural engineer who both put their names on the drawing and both carry it when the thing comes down. And then it decided that the software running the payroll, the patient records, the door locks and the delivery fleet needs… nobody in particular.
So 750,000 is not a wild ask. It is half the accountants. It is six times the building architects… It is roughly the number this country would have built decades ago if anyone with the authority to act believed a technology decision was as consequential as a tax return.
And the Demand Cycle Is Accounting, Not Buildings
This is the part I got wrong for years, so let me correct myself in public.
I used to reach for the building analogy. Codes, drawings, permits, the cathedral. It is a great analogy for responsibility and a terrible one for demand. A business hires a building architect when it builds or renovates. Twice in the life of the company, maybe. Never, for most of them. That is exactly why 123,000 people can cover a nation of 36 million businesses.
Accounting works nothing like that. Every business, every year, forever. Payroll runs. Quarterly filings. The call before a big purchase. The person who says that lease is going to look ugly on your books. It never stops, which is why it takes a million and a half people.
Now think about what a small business owner faces in technology. The point of sale contract renews. The vendor sunsets a product she has run for nine years. Payment rules change. Someone offers an AI add on for two hundred a month that quietly touches every customer record she has. Her competitor launches something and she has a week to answer it.
That is the books. A decision cycle measured in weeks.
So the profession we should be sizing against is accounting. Which also means my 750,000 is possibly LOW.
And from building we take the other half, the part accounting never had to solve at that scale. A drawing gets a stamp. A named human being with a license is accountable for the load path. Right now nobody signs the architecture of that clinic’s monitoring rollout. There is no drawing, no stamp, no name. There is a purchase order… maybe a sales rep who is uninterested. Lets not pretend big tech cares about small accounts.
Take the cadence from accounting. Take the accountability from structural engineering.
And Then There is the Rest of the Planet
About 400 million small and medium businesses worldwide. Ninety percent of all businesses. Sixty to seventy percent of all employment. Half of global GDP, per the UN.
The US is around nine percent of that.
Do the multiplication and you land somewhere between six and ten million architects. I dont love the width of that range. It is the honest one. Anyone giving you a tighter number is selling you a certification. And yes, we sell certifications, so feel free to point that out in the comments 🙂 but that is meant to scale to universities, training companies and partners… if we can get them speaking the same language.
But This Was Never a Question About Architects
Let me put the whole thing another way, because the architect part is a distraction and I am tired of that argument anyway.
The real question is about technology. Is it important, or is it just cool?
If it is important, if it belongs in the same category as a tax return that has to be right and a building that has to stand up, then everything above follows. You need trained people. You need a path, a standard, an ethic and a name on the decision. That is what a society does with things that matter. Every time. Medicine, law, accounting, structures, aviation. No exceptions.
If it is not important, if it really is just a fun toy that a clever person can pick up over a weekend, then fine. I will drop it. But then somebody has to explain the money.
Explain the salaries. Explain the equity, the recruiters, the retention bonuses, the shortage panic, the conference circuit, the consultancy day rates. Explain why a board that has never asked who designed anything goes white when the system stops. You dont get to have it both ways. Either this work is consequential and it deserves what every other consequential profession has… or it isnt, and we should ALL be paid a great deal less than we are.
In 24 years I have never once heard an executive pick the second answer out loud.
They all say critical. Mission critical. Strategic. Core to the business. And then the same person hands the design of their patient records to a rep with a quota and never asks who is accountable when it goes wrong.
That gap is the whole argument. Everything I calculated above is just the size of it.
But Where Do Those People Come From
Here is where it stops being a fun thought experiment.
It takes 5 to 8 years to make a board certified professional level architect. Not a person with a badge. A person who can sit across from a business owner, understand her business model, price the tradeoffs, make the call and carry the consequences. And 8 to 10 or more years to make a distinguished one.
You cannot compress that. Ive watched people try for 24 years. What you get instead is a confident person with a framework and no scar tissue.
So say we decide to build the American 750,000 over 20 years. That is 37,500 new board certified architects a year.
Now add the fallout. Twenty to thirty percent of the people who start this path leave it. They go back to engineering, which they miss. They go into management, which pays. They find out that being the person who says no to the CEO is a lonely job. So to net 37,500 you have to start about FIFTY THOUSAND people a year.
Every year. For twenty years.
For scale, in 2025 the entire United States enrolled 23,440 first year medical students. Total enrollment across every medical school in the country is 100,723 people.
We would need to start more than twice all of American medicine. Every single year.
And Who Exactly Teaches Them?
This is the part that actually keeps me up.
Fifty thousand entrants a year, each in supervised practice for six years or so, means 300,000 people in training at any given moment. Medicine solved this with the teaching hospital. Residents work on real patients under someone who has done it a thousand times, and the ratio stays small because it has to. Five to eight apprentices per experienced professional. Any more than that and you are not mentoring, you are lecturing.
That means 40,000 to 60,000 mentors. At distinguished level. People who took a decade each to make.
How many exist right now? I could name a few hundred. Maybe a couple thousand worldwide if Im generous and count the ones who never bothered with certification. Ive interviewed and assessed thousands of architects in over 50 countries and Im telling you the deep bench is not there.
That is the wall. You cannot buy past it, you cannot prompt past it, and you cannot hire it from a consultancy because they have the same problem and worse incentives. There is no version of this that runs faster. NONE.
The rate limit on this profession is mentors, not students.
So It Takes a Generation. Fine.
Medicine did not fix itself in a quarter. Abraham Flexner published his report in 1910 and it took the better part of forty years before the public could safely assume the person holding the stethoscope had been properly trained. Schools closed. Standards got written. People fought about it bitterly and some of them lost their businesses over it.
We are somewhere around 1908.
So the work is not a campaign for next year. It is mentoring, at volume, in the open, on real cases with real accountability. It is every senior person in this profession taking two or three people and actually teaching them instead of answering their Slack messages. And it is treating the small business as the teaching case rather than the afterthought, because that is where the volume is and honestly that is where the learning is.
It is also refusing to pretend any of it is quick.
So I Am Asking You
Please dont take my arithmetic as gospel. I probably got something wrong. Run it yourself and argue with me, because I would rather have the profession’s numbers than mine.
How many truly small businesses can one architect really carry… 20, or am I being soft? How many mid sized ones before you are lying to the client? Is 5 to 8 years right, or does serving small business need a shorter path and a different certification than the enterprise one we built? What is the real fallout rate, 20 percent, 30, higher? How many distinguished mentors do we actually have, by name, today? And how do we make the next ten thousand of them when the ones we have are already booked solid?
Halve every number I used. Tell me a good architect can carry 40 instead of 20. Fine. It is still 375,000 people in one country and we have almost none of them.
And before any of that, answer the one that decides it. Is technology important, or is it just cool? Because if it is important then this is a profession and we are twenty years late. And if it is just cool, then send back the money.
So who trains the first ten thousand? And who mentors the mentors?
Come do it with us at Iasa. Bring your arithmetic.
Further Reading
- US SBA Office of Advocacy, Frequently Asked Questions About Small Business 2026. 36,207,130 small businesses, 29,811,495 with no employees, 6,395,635 employers, data year 2022.
- US SBA Office of Advocacy, United States 2025 State Profile. Small business count by size, data year 2022. 29,811,495 without employees, 5,720,093 with 1 to 19, 654,501 with 20 to 499.
- US Census Statistics of US Businesses and SBE Council firm size shares, used only to split the 20 to 499 bucket into roughly 555,000 firms at 20 to 99 and about 100,000 at 100 to 499.
- US Bureau of Labor Statistics, Occupational Outlook Handbook. 179,200 computer network architects in 2024.
- NCARB, the number of US architects rose 6 percent in 2025. Just over 123,000 licensed, about 3,500 newly licensed last year.
- US Bureau of Labor Statistics, Accountants and Auditors. 1,579,800 employed in 2024, about 124,200 openings a year.
- US Bureau of Labor Statistics, Civil Engineers. 368,900 employed in 2024. Structural engineering is listed as a specialty and there is no clean national count of structural engineers, which tells you something on its own.
- NASBA Accountancy Licensee Database. 653,408 actively licensed CPAs as of August 2025.
- AAMC, US medical schools enroll record number of students in 2025. 23,440 first year students, 100,723 total enrolled.
- United Nations, MSME Day. MSMEs are 90 percent of businesses, 60 to 70 percent of employment, about half of global GDP.
- Iasa BTABoK, Decisions. The maturity levels for architectural decision making, and why a small business owner without one is guessing. https://education.iasaglobal.org/browse/btabok/3.2/core-site/core/concept/decisions