How to sell a half-built SaaS that has no users

The single most common message I get from would be sellers goes something like this: “I have a thing that mostly works, but nobody uses it. Is there even a point in listing?”

Yes. Almost always yes.

Half built projects with zero users sell on Failedups all the time. Not for life changing money, but consistently in the $500 to $5k range, and sometimes higher when the codebase is unusually clean or the domain is unusually good. The buyers exist. The hard part is convincing yourself you have something worth listing.

What buyers actually pay for when there are no users

When I read the buyer side messages on closed deals, the same things keep showing up. Users are not on that list. Here is what is:

Working code that compiles on first clone. Sounds obvious, rarer than you think. A repo that runs end to end with one npm install && npm run dev is worth real money to someone who does not want to spend a Saturday fighting peer dependencies.
A clean stack the buyer can read. Next.js, Postgres, Stripe, Clerk. Boring tools the buyer already knows. Nobody wants to inherit your novel architecture experiment.
A domain. A reasonable .com is worth $500 to $5k on its own. If your project ships with one, you are basically selling the domain plus a starter kit.
Time savings. Even a half built thing represents 60 to 200 hours of someone’s life. The buyer is paying for the parts they do not have to figure out: schema, auth wiring, the hundred small product decisions that look invisible until you have to make them.
Validation already done. You probably interviewed 10 people. Even if it went nowhere, your “what I tried” notes save the next person months of false starts.

None of those require users. They only require that the work actually happened and that you can show it.

Pricing without sunk cost brain damage

The framework I keep coming back to is the one I wrote about in how to value a stalled startup: replacement cost minus stale discount, with a small validation premium if you have receipts.

For a no user project, that math usually shakes out like this:

Estimate honest dev hours. Not “three years of weekends” but “if a competent freelancer rebuilt this from scratch, how many hours?” Usually 80 to 250.
Multiply by a low freelance rate. $60 to $100. That gives you a replacement floor of $5k to $25k.
Apply the stale discount. Six months untouched, take 20 percent off. A year, take 35. Two years, take half.
Subtract another big chunk because the buyer is doing all the distribution work themselves. This is where the “no users” reality hits the price. Realistically, 50 to 70 percent off the post stale number.
Add back $500 to $2k if you own a clean domain.

You usually land somewhere in $500 to $5k. That is not a sad number. That is the honest number. Sellers who anchor at $20k because they “spent a year on it” are the ones whose listings sit forever. Sellers who quote a number they can defend with a straight face are the ones who close in two weeks.

Writing the listing so it does not repel buyers

A no user project lives or dies on the listing copy. The pattern that works:

Be specific about what works. Not “MVP is functional.” Instead: “Signup, email verification, Stripe checkout for one plan, and the core dashboard with three widgets are all wired up and tested.”

Be specific about what is broken. “Password reset is stubbed. The admin panel half exists. The mobile layout breaks below 600px.” Buyers trust sellers who name the holes.

Say why you are letting go. Not vaguely. “I took a new full time job in October and have not opened the repo since.” Or, “I lost interest in the problem space after talking to 20 potential users.” That kind of honesty is correlated with deals closing. Vague hand waving is correlated with listings dying on the vine.

Say what you would do next if you had time. This sells the upside without you having to oversell. “If I had another month, I would replace the homebrew auth with Clerk and ship the team invites flow that is already 60 percent built.” That paragraph is what turns a tire kicker into a buyer.

The seller’s note is the goldmine

Every closed deal I have seen included a private seller’s note the buyer got after purchase. The boring “here is what failed” stuff is what people actually pay for. What to put in it:

Every customer interview takeaway, even the ones that contradict each other.
The pricing experiments you ran and what happened.
The marketing channels you tried and how they performed (yes, even “I posted on Twitter once and got 12 signups”).
The features you cut and why.
The architectural decisions you regret, so the buyer does not repeat them.

A strong seller’s note can comfortably push your asking price up 30 to 50 percent. Buyers know how rare it is.

Red flags that make a project genuinely unsaleable

Some projects really are tough sells. Be honest with yourself about whether yours has any of these:

Wrappers around a dead or about to die API. If your whole product is “ChatGPT plugin for X” and the plugins ecosystem just got deprecated, the value is mostly gone. Do not fake it.
Custom homebrew auth. Buyers do not want to inherit your bespoke session logic. If that is your spine, consider rewriting it onto a standard provider before listing.
Undocumented mess. No README, no env example, no deploy instructions. This is fixable in a weekend and triples your odds of closing.

What to put in the images

Three to five images: the dashboard or main interface, one feature deep dive screenshot, the schema or architecture sketch, the domain rendered in a browser bar, and one screenshot of the codebase showing the file tree. The last two punch above their weight. Buyers want to feel they are getting something real, not a Figma file.

The honest pitch

Yes, your no user project will probably sell. No, you will not get rich. But somebody out there genuinely wants what you built, often more than you do at this point. Listing it is the kindest thing you can do for the work.

If you are ready, add your listing.