Funding an AI Pre-Seed With Small Shipped Apps Instead of Investors
Shipping small paid apps can fund an AI company's pre-seed stage when their combined profit covers founder living costs and compute, and when the apps teach something the main company will use. It works when apps share infrastructure and an audience. It fails when maintenance and support consume the build time the core product needs.
The economics, worked through
The appeal is simple. AI coding tools have cut the time to ship a small paid app from months to days, so a founder can launch several, keep the ones that earn, and use the profit to fund the company they actually want to build. No dilution, no pitch meetings.
The arithmetic decides whether it is a strategy or a distraction. An illustrative portfolio:
- Four small apps with combined gross revenue of 9,000 a month, 6,000 of it through the iOS App Store and 3,000 through web checkout.
- Apple's commission at 15 percent on the App Store portion, assuming enrolment in its Small Business Program, which applies to developers under 1 million in annual proceeds: 900.
- Payment processing on web sales at roughly 3 percent: 90.
- Model inference and API costs at 25 percent of revenue, typical for apps that call a hosted model on every use: 2,250.
- Hosting, domains, and tools: 400. Paid acquisition: 1,000. Refunds and chargebacks: 300.
Net contribution: about 4,060 a month. That covers one founder's living costs in many cities and pays for experiments on the core product. It does not fund two founders in an expensive city, and it does not hire anyone.
Compare it with what investors put in. Carta reported that the average pre-seed instrument on its platform in Q2 2026 was 276,000. At 4,060 a month, matching a single average cheque takes more than five years of profit. App revenue buys time and evidence. It does not replace capital at the same speed.
The costs that grow quietly
Inference on flat pricing. A subscription priced as unlimited loses money on heavy users. Price by usage or cap it.
Maintenance. Model versions are deprecated, platform policies change, and each app needs updates. Four apps at a few hours a week each is a part-time job.
Support and review. App store review cycles and customer email take real hours.
Tax. Digital goods sales can create sales tax or VAT obligations in places you have never visited. App stores usually handle this on their own sales. Your web checkout may not.
When the model works
Three conditions separate a funding strategy from a hobby that eats the company.
Shared infrastructure. The apps run on the same stack, the same auth, the same billing, the same prompt and evaluation tooling. Each new app costs less to ship and maintain than the last. If every app is a fresh codebase, maintenance cost scales linearly and the model breaks at around the third or fourth.
Shared audience. The apps reach the same users through the same channel: one email list, one social following, one community. Distribution is the scarce asset, and building it is the most transferable thing the apps produce.
Every app is an experiment about the main company. Which features people pay for, what they will tolerate in latency and accuracy, which acquisition channel converts. If the apps teach nothing the core product will use, they are a separate business.
Set a kill rule before launch. An app that does not reach a defined revenue level in a defined period gets shut down, and its users get notice.
When it fails
The apps become the company. One app takes off, it demands all the time, and the AI company that was the point never starts. That can be a fine outcome. Decide deliberately rather than drifting into it.
Clones. Small apps built quickly can be copied quickly. Without distribution or proprietary data, revenue erodes as competitors arrive.
Platform and provider dependence. A single app store policy change or a model price increase can wipe out margin across the whole portfolio at once, because every app shares the same dependency.
Attention. Founders underestimate the switching cost of running five products. The core company needs long stretches of uninterrupted work, and support tickets from four apps fragment every day.
How it reads to investors later
Done well, the story is strong. It shows shipping speed, the ability to find and charge users, discipline about unit costs, and a company that stayed alive without outside money. A founder who is default alive negotiates from strength, because the round is optional.
Investors will ask three questions.
What did the apps teach you? Connect each one to a decision in the core product. A portfolio with a clear thread reads as research. A random portfolio reads as a lack of focus.
What happens to the apps now? Wind them down, sell them, keep them as a distribution channel for the main product, or fold them into it. Have the answer, and show the revenue as its own line so it does not blur the core company's metrics.
Who owns them? Apps built before incorporation often sit in a personal developer account or a separate LLC. Investors in the company expect it to own the code, the accounts, the domains, and the user data it relies on.
Move ownership before you raise
Assign the code and IP to the company in writing, transfer the app store accounts and domains where the platform allows it, and make sure privacy policies permit the user data to move with them. App store account transfers have their own procedures and can take time. This is general information about US practice, not legal advice. IP assignment and data transfer depend on your specific documents, so have counsel review them before diligence.
Frequently asked questions
- Can you fund a startup with revenue from small apps?
- Yes, if the combined profit covers founder living costs and the apps share infrastructure and an audience. After platform commissions, inference costs, and acquisition, a few thousand a month in profit is realistic for a small portfolio. That buys time and evidence for the core company. It rarely funds hiring, so most founders who start this way still raise eventually.
- How much app revenue replaces a pre-seed round?
- Usually far more than founders expect. Carta reported an average pre-seed instrument of 276,000 in Q2 2026. Replacing even one such cheque at a profit of 4,000 a month takes more than five years. App revenue replaces the need to raise immediately, not the capital itself. Treat it as a way to raise later and on better terms.
- Do investors mind if a founder runs other apps?
- They mind ambiguity, not the apps. Show what each app taught you about the main product, report its revenue separately, and state the plan: wind down, sell, fold in, or keep as a distribution channel. A founder who has decided reads as focused. One still splitting time across five products reads as a risk to the investment.
- Who should own the apps if you raise later?
- The company, if the core product depends on their code, users, or data. Apps built personally or through a separate LLC should be assigned to the company in writing, with developer accounts and domains transferred where the platform allows. Investors check this in diligence. The details turn on your documents, so get counsel to review them.