What Buyers Actually Care About When You List Your App

What Buyers Actually Care About When You List Your App

You're sitting on a working app doing $400 a month, and you've been "almost ready to list" for six weeks. Be honest: it's not the listing form that's stopping you. It's the fear that some buyer will open the repo, see Cursor-generated React, and ridicule you in public.

That fear is fiction. Here's what's actually true.

TL;DR

Five worries you can drop today, because no buyer pays them a second of attention:

1. ·"My code is messy AI output"
2. ·"I used Cursor / Bolt / Replit"
3. ·"It's too niche to sell"
4. ·"There's no patent, no moat, no IP"
5. ·"The stack isn't impressive enough"

Three worries you actually need to fix this weekend before you list:

1. ·There's no documentation
2. ·The app dies the second you stop touching it
3. ·Hosting costs are eating the MRR

That's the whole post. Want the proof and the fix recipes? Keep reading.

The Five Worries You Can Drop

"My code is messy AI output"

A buyer who's about to spend $25,000 on a SaaS doing $1,200 MRR is looking at three things on the listing page: revenue, churn, and the trailing 6-month chart. They are not pulling up your App.tsx to grade the import order.

Here's the part new sellers miss: the buyer almost always plans to rewrite the parts they care about anyway. They want the customers, the domain, and the Stripe account. The code is scaffolding. Listings on this marketplace and others routinely close where the buyer's first message after the wire clears is "ok sending it to my dev to refactor next week."

The only code thing buyers actually flinch at is a 4,000-line single file with no folder structure. If your project has more than three folders and zero files over 1,500 lines, you're fine.

"I used Cursor / Bolt / Replit"

This was a real concern in 2023. In 2026, "I built it with Cursor" is closer to "I built it with VS Code" than to a confession. Look at any indie-hackers thread about recent micro-acquisitions: the builder tool barely makes the writeup.

The one nuance: if the app is hosted on Bolt or Replit infrastructure (not just built with their editor), call that out plainly in the listing. Buyers don't care about the editor, but they do care if migrating means a weekend of yak shaving. Disclose it, price it in, move on.

"It's too niche to sell"

The opposite is true. The dental-clinic appointment reminder doing $900 MRR with 12 customers and 2% monthly churn sells faster than the generic "AI productivity app" with 8,000 free signups and no revenue. Narrow + paying always beats wide + free.

A useful rule: if you can describe your buyer in one sentence ("solo dentists who run their own front desk"), the app is not too niche. It's positioned.

"There's no patent, no moat, no IP"

Nobody buying a $5K to $50K app expects a patent. The moat is the customer list, the Stripe history, and the SEO position if you have one. A 24-month-old domain with 40 customers paying $29/mo is more defensible than any patent you could write up in a weekend, because the next person trying to clone it still has to convince those 40 customers to switch.

Stop apologizing for not having a moat. Show the retention curve instead.

"The stack isn't impressive enough"

Next.js + Supabase + Stripe is the modal stack for everything that sells under $50K. It's not impressive. It's not supposed to be. It's the marketplace equivalent of a Honda Civic engine: buyers know it, can hire for it, and won't be surprised by it. That's the point.

If you're tempted to rewrite in something more "serious" before listing, don't. You'll burn three weeks and the buyer will be mildly annoyed they have to learn a new framework.

The Three Worries You Actually Need to Fix

These are the ones that kill deals. Each one is a weekend of work, max.

1. No documentation

This is the single most common reason buyers ghost after the first call. They ask "how do I run it locally?" and the answer is a 20-minute Loom video of you remembering which env vars matter.

The fix (3 to 4 hours):

Write a README.md with exactly these sections:

• ·What this app does (3 sentences, plain English)
• ·Stack (one line: e.g. "Next.js 14, Supabase, Stripe, Resend, deployed on Vercel")
• ·Run locally (the actual commands, copy-pasteable, including cp .env.example .env)
• ·Environment variables (a table: name, what it's for, where to get one)
• ·Deploy (which Vercel/Render/Fly project, where the DNS lives)
• ·Known weirdness (the 2 or 3 quirks the buyer will hit in week one)

That last section, "Known weirdness," is the one buyers love most. It's the difference between "this founder is hiding things" and "this founder is honest."

Commit a real .env.example while you're there. If you have one secret hardcoded somewhere, fix it now.

2. The app dies when you stop touching it

If the answer to "what happens if you take two weeks off?" is "support tickets pile up, three customers churn, and the embedding job stops running because the API key is in my notes app", that's a deal-killer. Buyers call this founder dependency and they price it in heavily, often as a 30 to 50 percent discount on the multiple.

The fix (one weekend):

• ·Move every secret out of your notes/keychain into the hosting provider's env vars. Document where.
• ·Set up uptime monitoring (UptimeRobot free tier, 5 minutes to configure).
• ·Write the 5 customer-support replies you send most often as a one-page doc. Most "support" is the same three FAQs.
• ·If any cron job or background worker exists, write down how to restart it.
• ·Stop touching the app for 7 days as a test. What broke is what you fix before listing.

3. Hosting costs eat the MRR

If you're paying $200/mo to host an app doing $400 MRR, fix that before you list, or the buyer will. A 50 percent gross margin on a small app is a yellow flag. A 25 percent margin is a red one. Buyers underwrite on net, not gross.

The fix (varies, usually one afternoon):

• ·Open every service in your stack and write down the monthly cost.
• ·The usual offenders: an oversized Supabase plan, a dedicated Render instance for a 300-request-per-day app, a Vercel Pro account that could be Hobby, Pinecone when pgvector would do, Postmark when Resend's free tier fits.
• ·If you're on a paid plan that's 5x what you need, downgrade. The buyer doesn't need your headroom.
• ·Aim to walk into the listing with at least 70 percent gross margin. Anything lower needs an honest line in the listing about why.

A founder I talked to last month had a $1,400 MRR app burning $680/mo on infra. Two hours of work got it to $190/mo. The listing multiple went up roughly $8,000 on the same revenue, because the buyer was now underwriting $1,210/mo net instead of $720.

Real Talk: The One Thing That Actually Matters

Strip all of this away and the one number buyers stare at is revenue stability over the trailing 6 months. Flat $600 MRR for 6 straight months sells faster than $400 → $1,800 → $300 → $1,100 over the same window, even though the second one looks more "exciting" on paper.

If your numbers are bouncy, ship one boring month before you list. Don't over-discount, don't run a flash promo, don't change pricing. Let it sit. The chart you want is a slow uphill or a quiet flat line. That's the chart that gets full asking.

Now close this tab and go fix your README.