The State
of Startups 2026

Solo founders were already the largest group in 2025 at 53%. In 2026 they are 61% of respondents, up 8%. Cofounder pairs slipped.

Technical-founder share dropped from 82% to 78%. The same cohort leans harder on AI code generation than anyone else.

Every age band above 40 grew by a statistically significant margin. The 22 to 29 cohort shrank by 4%. Seasoned operators are filling the gap, often with AI doing the typing.

The top metros are still the top metros. But AI has flattened the development gap everywhere else. Europe and Africa grew 3%. Startups are setting up across Toronto, Chicago, Denver, and Austin, and throughout Europe, Asia, and Africa.

Cursor dropped 19%. v0, Bolt, and Windsurf each lost 6–9%. VS Code held flat. Antigravity appeared for the first time and took the 4th spot in its debut year.

Paid Claude subscriptions jumped from 28% to 59% of respondents. Paid OpenAI dropped from 57% to 39%. Gemini entered the list at 27%.

Anthropic/Claude climbed from 38% to 64%; OpenAI fell from 69% to 52%. Gemini entered at 44%. Hugging Face and custom models lost material share, a sign fewer teams run their own inference.

Among non-technical founders, the 76-to-100% share rises to 54%. Among 50-to-59-year-olds it rises to 60%. Flip the age cohort toggle to see it climb with age.

The more AI-generated a startup’s codebase, the less likely it is monetizing yet, and the more likely it names customer acquisition as its biggest challenge. One way to view this, early-stage startups are the ones more likely to be heavy users of AI, and they’re also the ones most likely to be early to monetization.

Postgres went from 76% to 80%. Every legacy NoSQL option lost share: MongoDB dropped 5%, MySQL 3%, Firebase 2%. Neon, DynamoDB, and Convex appeared as options for the first time and took small but measurable shares.

This question was new in 2026, so there is no prior baseline. Three in four respondents who answered the auth question picked Supabase Auth. The firm migration floor is the ~25% who picked no Supabase option at all.

Supabase held its lead. Vercel was already ahead of AWS in 2025, and in 2026 it extended that lead by 9%. Cloudflare grew fastest of all, crossing 27% and passing AWS on the way up. Every hyperscaler lost share.

React and Next.js both grew. But four tools went from effectively zero to real share in 12 months: Expo 10%, TanStack 8%, HTMX 4%, Astro 3%. Native mobile also picked up 3%.

Agent-building share is statistically flat year over year. The "not sure" cohort shrank, which means undecided builders are making up their minds and shipping. Agents are also getting more sophisticated and handling workflow and data analysis, not just customer support.

24% of people building agents are orchestrating multiple agents. The advanced builders are racing ahead of the pack.

Agent usage and construction is growing. But the operational layer (monitoring, versioning, evaluation, etc.) is still missing. If the agentic trend continues, this represents a real opportunity for the next wave of builders.

A year after the Model Context Protocol launched, more than half of respondents are already active users. Even as the debate over MCP and APIs continues, people still find MCP valuable in at least some contexts.

53% of startups with a GTM motion have no formal CRM, up from 43%. Every named CRM lost share: HubSpot, Salesforce, Notion/Airtable, Google Sheets. Teams build their own or do without.

"Custom solution" is the fastest-growing answer, up 2.5%. Datadog and Prometheus both lost share. Sentry kept its lead and grew slightly. The observability market is bifurcating between Sentry for errors and custom dashboards for everything else.

"I don’t track this yet" grew 5%. HubSpot, Salesforce, Mixpanel, Segment all lost share. The pattern repeats: skip the vendor, ship something internal, or don’t track at all.

Personal networks remain the top source of initial paying customers. 67% of respondents still haven’t tried paid acquisition at all, up from 62% last year.

Dedicated full-time sales hires usually do not arrive until after the tenth employee. Product-led growth as a motion climbed 4% to half of respondents. "Not sure yet" is shrinking.

For the first time in the survey, startups are picking a pricing shape earlier in their lifecycle, and they are picking the same one. Tiered feature plans went from 23% to 36% of respondents who picked any pricing model. The "still experimenting" cohort shrank.

Most startups skip community entirely. 48% have not built one, up 4%, and the "planning to" middle is shrinking. Only 11% have built one. That small group is the one to watch: the chart below shows they grow from a completely different mix of channels.

When anyone can ship anything, distribution is the bottleneck. Personal networks run dry. Cold outreach gets ignored. The startups that built a developer community ship into a different funnel: open-source users and developer communities convert into paying customers at multiples of the rate non-community teams see. The chart below splits the customer-source mix by community status.

X lost 6%. LinkedIn lost 3%. Reddit and Discord lost 3-4%. TikTok was the only platform that grew. The "I have no online persona" share grew 5% to 33%. One in three respondents is fully offline.

The "none of the above" cohort jumped 10%. Google Cloud Next, AWS re:Invent, Microsoft Build, and Y Combinator Demo Day all lost share. Conference-led developer marketing is working for a smaller slice of the market every year.

The largest year-over-year shift in any single category. Three new challenge options came online: burn out, AI competition, runway anxiety. Together they absorb roughly the same share that used to pick technical complexity. Among 1-10 person teams, burn out has already overtaken technical complexity as the second-biggest challenge.

57% say they are optimistic, down 1% from last year, not statistically significant. Founders are 58% optimistic; non-founders are 49%.

Optimism rises with AI codebase share: 49% of zero-AI founders feel optimistic about the world, 61% of heavy AI users do. Revenue runs the other way. 56% of zero-AI users are currently monetizing. 31% of heavy AI users are. One reading is that hands-on AI experience demystifies the technology and turns it into something founders feel they can wield. The heaviest AI users are also the most likely to be bootstrapped, the earliest in their lifecycle, and the least exposed to whether the market actually wants what they have built.