GPT Wrapper SaaS: Score 43 — Why the $42k Ask Collapses When You Read the API Bill

The Asset at a Glance

This is an AI writing assistant — a B2C SaaS product built on top of OpenAI's API. It generates marketing copy, social captions, and blog outlines via a subscription model. Typical profile: $15–$29/month plans, self-serve signup, no sales team.

The listing reports $2,800 MRR, approximately 180 paying subscribers, and "passive income" positioning. Asking price: $42,000.

We scored it using AI Asset Score v1.0. The analysis is based on the listing data and a standard diligence framework for AI-native assets. We did not conduct a full due diligence review.

AI Asset Score: 43 / 100 — Demo

The Revenue Problem: Gross vs Net

This is the most common mistake buyers make on AI SaaS listings. MRR is a gross number. For any GPT wrapper, the OpenAI API bill is a direct cost of revenue — not an operating expense you can cut. It scales with usage.

In this case:

• Reported MRR: $2,800
• Monthly OpenAI API costs: ~$1,100 (estimated from typical usage per active user at gpt-4o-mini rates)
• Hosting, Stripe fees, support tools: ~$200
• Net monthly profit: ~$1,500–$1,700
• Annualized net: ~$18,000–$20,400

At $42,000 asking, the buyer is paying ~2.1–2.3x annual net profit — which is within our 1–2x operator norm at the high end only if revenue holds. But it doesn't, because of churn.

Traction: 10 / 25 — Flat and Fragile

180 paying subscribers sounds reasonable. The problem is the trajectory. MRR has been flat for three months. Churn is running at roughly 30–35% per month on an annualized basis — meaning the business is on a treadmill, acquiring just enough new subscribers to replace the ones leaving.

Flat MRR with high churn is not a stable business. It's a business where the founder's sales effort (paid ads, Product Hunt bumps, AppSumo deal) is masking structural weakness. When the buyer takes over and the ad spend pauses, MRR typically drops 20–40% in the first 60 days.

• No organic search traffic: the product ranks for nothing on its own
• No referral loop: users don't invite other users
• Acquisition is 100% paid — and the founder runs the ads

Transferability: 9 / 20 — Technically Movable, Operationally Fragile

The code can be transferred. API keys are in environment variables. The Stripe account can be handed over. On paper, this is transferable.

In practice:

• No written SOPs for customer support, refunds, or billing issues
• The founder handles all support personally via email — ~15 tickets/week
• Onboarding relies on a welcome email sequence that references "the founder's personal blog"
• No technical documentation — a new owner would need a developer to add features or debug issues

This is the difference between "can be moved" and "can be operated by someone who didn't build it." The score reflects operational fragility, not technical impossibility.

Risk: 3 / 10 — The OpenAI Problem

This is where GPT wrappers consistently score worst, and for good reason. The moat is thin — near zero.

Three specific risks apply here:

1. OpenAI pricing risk. API costs went up 3 times in 18 months across different models. Each price increase compresses net margins directly. The buyer has no leverage.
2. Feature cannibalization. OpenAI added marketing copy generation to ChatGPT's standard interface in 2025. Every GPT wrapper competes with a product that 200M people already have for free or at $20/month.
3. No niche moat. This product serves "anyone who writes marketing content." That's not a niche. Niched alternatives (e.g., AI copy specifically for real estate agents, or SaaS landing pages) have higher retention because users can't find exact substitutes. A generic writing assistant can be replaced by dozens of competitors.

Automation: 8 / 20 — Not as Passive as Advertised

The core product runs without human involvement — the AI generates output, Stripe handles billing, the app handles authentication. That's true automation of the revenue engine.

But "passive income" claims break down in practice:

• Customer support: ~10–15 tickets/week, all manual
• Failed payments and dunning: handled manually when Stripe flags an issue
• Model updates: when OpenAI changes API behavior (prompt formats, token limits), someone needs to update the prompts
• Competitive feature pressure: users compare weekly with what ChatGPT just added for free

For an operator buying this as a side asset, budget 5–8 hours per week minimum. That's not passive.

Valuation: What the Asset Is Worth

The asking price is defensible only in the optimistic scenario — and only if you believe MRR will hold for 24 months post-transfer. Given the churn rate and platform risk, that's a bet we would not take at $42k.

Verdict: Pass at $42k

The asset is real. The revenue is real. But the price assumes a stability that the fundamentals don't support.

For an operator to break even in 12–18 months — our standard payback window — the entry price needs to be $18,000–$25,000. At that range, even with post-transfer churn, the math works.

What would move us toward a conditional pass at a higher price:

• Verified MRR with 6 months of Stripe + bank statements
• Churn rate under 10% monthly (currently estimated at 30–35%)
• API cost audit confirming sub-$900/month for current user base
• A written SOP for support and billing handover
• Evidence of organic traffic (any SEO footprint at all)

The GPT Wrapper Pattern

This case is not an outlier. It is the median GPT wrapper listing. The pattern repeats across Flippa, MicroAcquire, and Acquire.com:

• Gross MRR headline, net margin buried or absent
• Flat or declining trajectory masked by recent promo
• Founder-operated support called "passive"
• No niche, competing directly with OpenAI's own product
• 3–5x MRR asking price that assumes perpetual growth

Our valuation norm for AI assets is 1–2x annual net profit. For GPT wrappers specifically, we lean toward 1–1.5x unless there is a defensible niche, strong organic acquisition, or low churn below 10% monthly. Most do not qualify for the upper end.

This doesn't mean GPT wrappers are worthless. At the right price — typically $15,000–$30,000 for a $2,000–$3,000 net MRR product — they can be reasonable operator assets. The question is always: is the seller pricing it like a real business, or like the growth they hoped to build?