Awesome post - thanks for putting it together. As a one-man business, I agree with this data :) I am not hiring, and don't plan to, only because of have AI at my disposal (for editing, designing, and more). AND, Stripe makes it incredibly easy to process payments. So thank you for helping my little 1-person company run.
Really good data here — the cross-validation across Census, Stripe, and international filings makes the "this isn't fraud" case well.
One thing the AI section understates, I think: you're measuring AI's effect on the capability side (one person doing work that used to need a team). But there's a second-order effect on the discovery side that doesn't show up in this data. A team produces legibility as a byproduct — separate employee profiles, a comms function, a site someone was paid to structure.
A solopreneur running real revenue off a Stripe link and not much else often has none of that residue, not because anything's wrong, but because there was never a team to leave the trail.
That was a minor disadvantage when humans did most of the finding — people fill gaps with context. It gets sharper as more discovery work (vendor search, supplier evaluation, candidate sourcing) shifts to agents reading structured data instead. Solo revenue going up and solo legibility going up are not the same curve.
Would be curious whether you're seeing anything on the discovery/search side that tracks this independently of the revenue numbers in this report.
I would encourage you to be more precise in your accounting.
$100k in revenue is not $100k in earnings. It could be $0 or $15k.
I presume by revenue you means cash receipts, as that is what Stripe can provide. That ignores any COGS (including token costs), operational costs, and stripe/processor fees.
This matters because 4m people with $100k in income would be interesting (still likely not a living wage, but interesting). The reality is that the vast majority is likely making less than $50k, which is basically poverty.
The line about AI being "revenge of the idea guys" is going to stick with me. I spent years assuming I needed a co-founder to cover the skills I didn't have, and it turns out the gap was smaller than I thought once the right tools existed.
The cross-validation is what makes this convincing. Census, Stripe proprietary data, cross-country filings, Delaware incorporations. Any one source could be noise. Four independent sources moving in the same direction with the fraud explanation systematically eliminated is a structural signal.
The question the data can't answer yet is survivorship. AI lowers the barrier to start and accelerates the path to revenue. Both are proven here. But does it lower the barrier to sustain? Customer retention, competitive moats, recurring revenue, all require durability that formation speed doesn't measure. The traditional employer business failure rate is roughly 50% within five years. If the 2025 solopreneur cohort starts 3x faster but fails at the same rate, the net effect on the economy is a faster churn cycle, not a permanent expansion of the business base. The formation and revenue data are compelling. The survival data doesn't exist yet because the cohort is too young. That's the chart to watch in 2028.
The most interesting data point in this piece isn't the headline solopreneur count. It's that the share crossing $5M and $10M thresholds is also rising.
AI isn't just creating more businesses. It's enabling solo operators to reach scale that previously required teams. The implications for what "fundable" means at formation stage are real.
Great to hear about solopreneurs, and their increase. Lots of good looking graphs, and makes me believe the qualitative conclusion.
But I don’t like the “share” graphs, based on some starting index number. A 5 or 10 or 50% share out of 100 makes sense. But what are the 5 lines of the y axis in the share over $1 million rev after 12 months? Can’t be quintiles, maybe deciles or merely 1 %. So in 2025, over 50% made it, or maybe over 5%; while in 2019 less than 20%, or less than 2%. Or some other level.
But ok, lots more in 2025. Similarity the solopreneur index share by income, lots more of the 115 over $10 million. That’s a chart asking for yearly bars for 6 exclusive, not inclusive, ranges:
<100k, <500k, <1m, <5m, <10m, >10m (the rest), which total 100% of all 115 firms.
Thanks for at least showing in data how you’re getting your conclusions.
Awesome post - thanks for putting it together. As a one-man business, I agree with this data :) I am not hiring, and don't plan to, only because of have AI at my disposal (for editing, designing, and more). AND, Stripe makes it incredibly easy to process payments. So thank you for helping my little 1-person company run.
Really good data here — the cross-validation across Census, Stripe, and international filings makes the "this isn't fraud" case well.
One thing the AI section understates, I think: you're measuring AI's effect on the capability side (one person doing work that used to need a team). But there's a second-order effect on the discovery side that doesn't show up in this data. A team produces legibility as a byproduct — separate employee profiles, a comms function, a site someone was paid to structure.
A solopreneur running real revenue off a Stripe link and not much else often has none of that residue, not because anything's wrong, but because there was never a team to leave the trail.
That was a minor disadvantage when humans did most of the finding — people fill gaps with context. It gets sharper as more discovery work (vendor search, supplier evaluation, candidate sourcing) shifts to agents reading structured data instead. Solo revenue going up and solo legibility going up are not the same curve.
Would be curious whether you're seeing anything on the discovery/search side that tracks this independently of the revenue numbers in this report.
I would encourage you to be more precise in your accounting.
$100k in revenue is not $100k in earnings. It could be $0 or $15k.
I presume by revenue you means cash receipts, as that is what Stripe can provide. That ignores any COGS (including token costs), operational costs, and stripe/processor fees.
This matters because 4m people with $100k in income would be interesting (still likely not a living wage, but interesting). The reality is that the vast majority is likely making less than $50k, which is basically poverty.
Is there a solopsistic pun dying to surface in this piece? Enjoy...
The line about AI being "revenge of the idea guys" is going to stick with me. I spent years assuming I needed a co-founder to cover the skills I didn't have, and it turns out the gap was smaller than I thought once the right tools existed.
The cross-validation is what makes this convincing. Census, Stripe proprietary data, cross-country filings, Delaware incorporations. Any one source could be noise. Four independent sources moving in the same direction with the fraud explanation systematically eliminated is a structural signal.
The question the data can't answer yet is survivorship. AI lowers the barrier to start and accelerates the path to revenue. Both are proven here. But does it lower the barrier to sustain? Customer retention, competitive moats, recurring revenue, all require durability that formation speed doesn't measure. The traditional employer business failure rate is roughly 50% within five years. If the 2025 solopreneur cohort starts 3x faster but fails at the same rate, the net effect on the economy is a faster churn cycle, not a permanent expansion of the business base. The formation and revenue data are compelling. The survival data doesn't exist yet because the cohort is too young. That's the chart to watch in 2028.
The most interesting data point in this piece isn't the headline solopreneur count. It's that the share crossing $5M and $10M thresholds is also rising.
AI isn't just creating more businesses. It's enabling solo operators to reach scale that previously required teams. The implications for what "fundable" means at formation stage are real.
Great to hear about solopreneurs, and their increase. Lots of good looking graphs, and makes me believe the qualitative conclusion.
But I don’t like the “share” graphs, based on some starting index number. A 5 or 10 or 50% share out of 100 makes sense. But what are the 5 lines of the y axis in the share over $1 million rev after 12 months? Can’t be quintiles, maybe deciles or merely 1 %. So in 2025, over 50% made it, or maybe over 5%; while in 2019 less than 20%, or less than 2%. Or some other level.
But ok, lots more in 2025. Similarity the solopreneur index share by income, lots more of the 115 over $10 million. That’s a chart asking for yearly bars for 6 exclusive, not inclusive, ranges:
<100k, <500k, <1m, <5m, <10m, >10m (the rest), which total 100% of all 115 firms.
Thanks for at least showing in data how you’re getting your conclusions.
Hi, I'm happy that my concept gain interest, but I believe you should at least mark sources ;)
https://mosor.pl/technology/the-liquid-future-ai-economy/
https://mosor.pl/economy/economical-deconstruction-through-decentralisation-to-liquidity-ai-economy/
https://mosor.pl/economy/soloeconomy/
Hi great article! Is this across certain sectors or are solo owners only showing up in certain areas?
I think if a company employees you for those skills, certainly your own clients would too
I think this can add to the perspective:
https://manshidixit.substack.com/p/what-it-actually-takes-to-become-a-solopreneur?utm_source=share&utm_medium=android&r=2z82yq