Emergent Index 2026: The Age of Custom Software

A third of small business owners have built their own software to solve operational challenges that no existing off-the-shelf solution has been designed to address.

That is the central finding of Emergent’s Age of Custom Software Index 2026, published this week by the AI software platform Emergent and based on 300 structured interviews with operators alongside a sample of more than 50,000 live, deployed applications built on the vibecoding platform.

The reason the software never arrived was partly price. Operators who asked a development agency what a basic custom build would cost reported a median quote of $20,000, with some quoted $100,000 and timelines running beyond a year. Cost was the blocker they cited most. Most of them walked away and carried on without.

These custom-built applications are not experiments. One in four takes payment directly. One in three sends transactional emails and one in three runs on a custom domain. Emergent counts roughly eight million people using the apps in the sample, across 28 industries from healthcare and retail to construction and poultry farming. Almost none of it was built by a development team.

Key findings

• Operators reported a median agency quote of $20,000 for a basic custom build, with quotes ranging up to $100,000 and timelines running from several months to more than a year. Cost was the most cited blocker.

• For one in three of the 300 operators interviewed, the software they built replaced nothing at all. Not a spreadsheet, not a subscription, not an agency. The tool did not exist at a price they could pay. A further 33 per cent dropped a SaaS subscription, 21 per cent retired a manual process, and 13 per cent cited independence from developers.

• Only one in seven operators interviewed was well served by existing software. Among the rest, 33 per cent had no approach that solved the problem, 27 per cent ran manual processes, 27 per cent used SaaS that did not fit, and 12 per cent paid for custom development.

Mukund Jha, co-founder and chief executive of Emergent, said: “The number that stayed with me is the one in three who had nothing to replace. We assumed we would mostly be looking at people swapping out a subscription. Instead, a third of them were solving a problem that no product had ever been built for, because their market was too small or their problem too specific for a vendor to chase. That software was never going to arrive. It only exists because the person closest to the problem finally had a way to make it.”

The index lands in a market where software buying is already under pressure. Research from Retool in 2026 found 35 per cent of enterprise s had replaced at least one SaaS tool with something built in house, and 60 per cent of respondents had shipped software without IT sign-off in the past year. That is the enterprise account of build versus buy. Emergent’s data is an attempt to document the same shift among businesses with no IT function at all.

The backdrop is an SMB sector where AI use is now common but shallow. US Bank’s June 2026 Small Business Perspective put generative AI adoption among US small businesses at 75 per cent.Thryv’s 2026 survey put SMB adoption at 66 per cent, up from 55 per cent a year earlier, with 70 per cent of owners saying they needed more training to use it effectively. Adoption is mainstream. What owners do with it, beyond drafting copy and summarising email, is largely undocumented.

Small firms are 99.9 per cent of US businesses and employ 62.3 million people, close to 46 per cent of the private sector workforce. Globally, the World Bank puts SMEs at around 90 per cent of businesses and more than half of employment. The index is Emergent’s case that this group is not simply adopting AI tools but starting to produce software of its own.

Emergent closed a $130 million Series C led by Creaegis on 15 July 2026 at a $1.5 billion valuation, bringing total funding to $230 million.

The full Emergent Index 2026, including industry portraits, verbatim operator prompts, nine named operator stories and the complete methodology appendix, is available at smb-report.emergent.sh. A summary is published at emergent.sh/blog/emergent-smb-report.

About Emergent

Emergent is an AI software creation platform that lets people build full-stack, production-ready web and mobile applications using autonomous AI agents. Founded in 2024 by Mukund Jha and Madhav Jha and publicly launched in 2025, the company is backed by Creaegis, Claypond, Khosla Ventures, SoftBank Vision Fund 2, Lightspeed, Sentinel Global, Y Combinator, Prosus, Together and Google’s AI Futures Fund. More than 12 million applications have been built on the platform. Emergent is headquartered in San Francisco with its engineering team in Bangalore.

Methodology

The index draws on two sources - the first is Emergent platform data covering a sample of more than 50,000 applications that were live and deployed as of the report snapshot. Deployment is the inclusion bar for every platform figure. Custom domain, payment and transactional email shares are computed on the deployed-app universe. Integration counts are computed per app and exclude the default database. Return-to-build, at 39 per cent, is computed per operator and measures the share still actively building on an app one month after launch. The eight million figure counts customers, clients and staff using the apps in the sample. It is measured on that sample only and is described by Emergent as a floor rather than a platform total.

The second source is 300 structured interviews with operators drawn from the deployed cohort. Interview findings describe that group and are not extrapolated to the wider population. The coping distribution is computed among operators who were not already well served by existing software, roughly six in seven of those interviewed. Development cost figures are quotes operators reported receiving. Operator prompts quoted in the report are verbatim, lightly cleaned for filler and anonymised by business. All shares are rounded to whole percentage points.

Reporters should note the sample is drawn entirely from Emergent’s own platform. Every operator surveyed had already chosen to build custom software, so the findings describe the population of people who built rather than small businesses generally. Emergent states the interview findings are not extrapolated beyond the interviewed group. The full methodology appendix is published with the report.

Contact Information

Michelle Mead

PR for Emergent

VSC for Emergent

emergent@vsc.co