You have capital. You want AI income without building from scratch. Most AI projects for sale are demos pretending to be businesses. This course teaches you to tell the difference — and close the right deal.
"I've looked at AI projects for sale on Indie Hackers, Product Hunt, and Twitter. Most of them look impressive in the screenshots. Then you ask for revenue numbers and get silence, or '$200 this month but growing fast.'"
We reviewed 47 AI projects in our last batch. Three qualified as real assets. The other 44 were, in various ways, demos with a price tag. Here's the fastest way to tell them apart.
A real asset answers three questions instantly:
If the answer to any of these is "not yet" or "kind of" — stop. You're looking at a demo.
Ask: "Can you show me 3 months of Stripe or payment processor data?"
A real seller sends it within an hour. A demo seller either doesn't have it, finds reasons to delay, or sends a screenshot of a dashboard you can't verify. End of conversation.
Real AI assets have revenue you can verify, run without the creator, and come with documentation. If any of these three are missing, move on — there are enough real assets out there that you don't need to gamble on potential.
"The seller says '$2,400/month MRR.' I have no idea if that's good for a $28k asking price. I don't know what normal looks like. I'm scared of overpaying for something that collapses in 3 months."
The AI asset market is new. There are no Zillow-style comps. No Kelley Blue Book. But there are financial ratios that apply across every deal — and once you know them, you can evaluate any listing in under 10 minutes.
Healthy range for AI assets: 12–24 months.
Below 12 months: either the price is too low (something is wrong), or revenue is inflated. Ask why before celebrating.
Above 24 months: price is too high, growth is assumed, or seller has optimistic projections. Negotiate down or walk.
Many AI projects run on OpenAI, Anthropic, or other API providers. These costs can spike unexpectedly as usage grows. Always ask: what happens to costs if revenue doubles? If the API cost structure is per-request and there's no cap, your margin can disappear as you scale. Look for projects with fixed-cost infrastructure or predictable per-seat pricing.
Healthy AI assets trade at 12–24× monthly net revenue. Always work with net revenue (after API and hosting costs). Ask for 6-month month-by-month breakdown — not just the last month. A good deal is one where the math works on current numbers, with no assumed growth.
"Even if the numbers look fine, I'm scared of buying something where one small thing breaks the whole business — and I only find out after the transfer."
Most AI asset disasters are predictable. After reviewing 47 projects, we've catalogued the patterns. Here are the 7 things to check before any money moves.
If one customer accounts for more than 30% of MRR, that customer is a business risk you're buying. Ask: what happens if they cancel? Would you still buy at this price? Demand a discount proportional to concentration risk, or a 90-day earnout tied to that customer's retention.
Ask to see the codebase documentation and README before closing. If there isn't one, ask: "Can I pay a developer $200 to review the code and ask questions?" Any legitimate seller will say yes. If they resist, something is hidden — complexity, third-party dependencies, or security issues.
A business built entirely on one LLM provider's API, one social platform's API, or one cloud provider's free tier is fragile. What happens if OpenAI changes pricing? What if the Instagram API breaks? Look for projects with at least one layer of abstraction, or priced to reflect the dependency risk.
Some AI businesses are profitable only because they're using free tiers of APIs (free Notion API, free plan on a SaaS, etc.) or have grandfathered pricing. Check: what does the cost structure look like at current pricing, not at the rates they're currently paying?
Customers who pay because of the seller's Twitter audience, their Telegram channel, or their personal reputation don't transfer. Ask: "Have you ever let someone else handle customer communications? How did customers respond?" If all customer relationships run through the seller personally, expect significant churn at handover.
If the seller can't give you read-only access to a payment processor (Stripe, Paddle, PayPal) for verification — walk. Screenshots are fabricatable. Always verify revenue through a system you can query directly, not through screenshots provided by the seller.
A real asset has a transfer path: a list of accounts to transfer, a list of API keys to hand over, a list of customers to notify (or not), and a timeline. If the seller says "we'll figure it out after the deal closes" — stop. Figure it out before. The absence of a transfer plan is itself a red flag about the state of the business.
Use this list as a checklist. You don't need all 7 to be perfect — but any single red flag should trigger a renegotiation or deeper investigation. The best deals are the ones where the seller is eager to show you everything, not the ones where you have to fight for information.
"What if I buy it, the original developer leaves, and all the customers follow them out? I'm paying for a customer base that might not stick around."
Customer retention during handover is the most underestimated risk in AI asset acquisition — and the most preventable. The first 30 days after transfer determine whether you keep 80% of customers or 40%.
The most critical window is the first 48 hours after transfer. Your goal: no customer should notice a difference in service quality. This means:
For most B2B AI tools ($100-500/mo customers), silent transition is better: the product keeps working, support quality stays the same, and you don't trigger the "should I keep paying?" question in customers' minds by announcing a change of ownership.
For consumer AI tools or products where community is important, a positive announcement works better: "The team is growing, here's the new operator, they're committed to the product." Frame it as a win, not a change.
Days 1-30: Change nothing. Understand the product as the customers use it. Fix bugs only. Do not "improve."
Days 31-60: First outreach to top 20% of customers (by value). Ask one question: "What would make this more valuable to you?" Listen more than talk.
Days 61-90: Implement one small improvement. Announce it to customers. This establishes you as an active steward, not an absentee owner.
The handover is not the end of the deal — it's the start of the relationship with customers. Plan it like a product launch. Test everything before you touch anything. Change nothing for 30 days. The deals that fail at handover almost always fail because the buyer moved too fast, too soon.
Submit it for a free AI Asset Score — we'll give you an independent score across 5 dimensions before you commit to due diligence.
Get a AI Asset Score →