When your weekend AI build is ready to become a company: the StudyStash checklist

You have a working AI build that you maintain between work and sleep. The Kortext purchase of StudyStash was seen as a low eight-figure dollar amount. The price was the payoff, not the trigger.

Widespread bottom-up AI adoption can produce high usage metrics without real business value in 2026. AI startup success in 2026 turns on distribution and the other core business dimensions, and those capabilities are harder to build than in older software cycles. The exit question is less about your demo and more about who already needs the product.

Institutional traction is the clearest signal

Only 25% of organizations reported in 2026 that 40% or more of their AI pilots reached production, so institutional production traction remains uneven. That rarity is what buyers pay for. A waitlist is curiosity; a procurement form is demand.

AI adoption in 2026 needs structured oversight, coordinated processes, and intentional change management rather than casual side-project treatment. A weekend experiment can stay a weekend experiment. When it is being used by a department, it starts to need a support path, a data review, and a clear owner.

StudyStash, an AI study platform, began with the University of Birmingham and later reached tens of thousands of students at more than 160 institutions across the UK, Europe, the US, and Africa. Kortext announced the deal on 2 October 2026

The readiness test is a yes/no screen

Run this before you treat the project as a company. Each item is a yes/no question you can answer from your own notes. Without institutional traction, keep the weekend schedule. With it, ask whether the product needs a support queue, a billing system, and a data review. A yes moves it toward company shape.

• Institutional traction: can you name a real organization that uses it in production, not just a waitlist?
• A 10-person team: StudyStash ran with 10 people, so ask whether your product could run at that size, including support, sales, and maintenance
• A strategic buyer with distribution: is there a company that needs your product to reach its existing customers?

Institutions buy outcomes, so the traction item filters out casual praise. A side project that needs a large team is a different workload, and the team item keeps the scope honest. Distribution is the hard part in AI, so the buyer item points you toward a company that already has the channel.

A quick test for the institutional traction item is to open your customer list or email and find a named contact at a named institution who asked for a renewal, a security review, or more seats. A screenshot of a landing page is not the answer. For the team item, list the roles that would keep the product running, then count them. For the buyer item, name a company whose customers match your users and whose sales team would benefit from your product.

The buyer should own the channel

OpenAI held the stronger model for years, but Google's reach through Android, Workspace, and Gemini-powered search shifted the competitive dynamic. A small team rarely wins that competition alone. A strategic buyer with distribution gives you access to customers.

Anthropic earned about 80% of its revenue from enterprise customers, had over 500 accounts with annual spend above $1 million, and Claude Code exceeded a $2.5 billion run rate. Enterprise buyers pay for a product that fits their workflow and can be supported, not just for a model.

StudyStash's founders moved to full-time work after their July 2025 graduation. Ben Ward and Jonathan Graham agreed to the Kortext takeover 13 months after graduating. Kortext reached more than 1,000 universities through its network

Do not quit a day job because a demo got attention. Treat it like a company only when all three points are true. When only one point is true, keep the weekend schedule, and that is a good outcome.