PropGPT: AI Props Analysis

Does PropGPT: AI Props Analysis have a moat?

No.

19/100 · Moatless

“An LLM over a sports odds feed takes a week or two to build. What a clone can't copy fast: 2,700 App Store ratings at 4.8 and a tested paywall.”

Code to copy

2 weeks

Moat to cross

3 months

Market

Crowded market

Share on X

Moat depth

MoatlessShallowReal

Verified revenueMRR $93,223Last 30 days $90,049All time $998kCustomers 2,363Founded 2024TrustMRR

01/The verdict

An iOS and Android app that gives AI analysis of player props, spreads, totals and moneylines. US sports bettors pay $9.99 a week.

PropGPT makes real money, about $93k MRR from 2,350 weekly subscribers, with no VC funding. Its edge is execution, not protection: an App Store listing with 2,738 ratings at 4.8 and about 100k+ downloads, plus a paywall tuned through Superwall price tests. The product is an AI layer over sports data you can license, no licence or platform approval is needed, and weekly subscribers can cancel any time, in a market already full of AI picks apps.

02/2 moats

Distribution

Shallow

~4 mo to cross

The iOS app has 2,738 ratings at 4.8 and ranks for 'AI sports picks' searches. Third-party trackers estimate 100k+ downloads, and Android adds 5,000+ installs. Its social channels are small: 2,215 Instagram followers.

To cross it: Buy installs on paid social and through sports-betting influencers for a full NFL and NBA season, while prompting users for ratings, until search ranking catches up.

Real-world operations

Shallow

~3 mo to cross

Superwall case studies show the team tested weekly prices ($6.99 vs $9.99, a 47.5% lift in proceeds per user) and annual plans, and runs at an 80% margin. That know-how lets them outspend weaker clones on acquisition.

To cross it: Use Superwall or RevenueCat paywall templates and run the same price and trial tests in your first few months.

03/The gap

Easy to copy

The AI analysis itself: an LLM prompt over licensed stats and odds APIs (SportsDataIO, The Odds API and similar) that anyone can buy
The PropGPT name and the 'first app of its kind' claim, which Rithmm, PropsBot, ScoreGPT and others already contradict
The UI: props boards, grades and value and risk explanations
No compliance barrier: it is a picks app, not a sportsbook, so no gaming licence is needed

How you'd go after it

A public, timestamped pick record that includes losses. Reviewers and comparison sites already rank apps on this, and some PropGPT reviews question where its data comes from
Undercut on price with a cheaper monthly plan or a free tier against the $9.99 weekly price, which works out to about $40 a month
Go after a niche it serves loosely: DFS pick'em (PrizePicks, Underdog), WNBA, college sports or esports, or a non-US market
Ad creative with sports-betting TikTok and Instagram creators ahead of NFL and March Madness, the high-intent windows when new bettors install apps

04/Build against it?

Only with an edge

Demand is clearly there: $93k MRR from 2,350 subscribers at an 80% margin. But with no barriers, winning comes down to paid acquisition and conversion. Only go in if you already know how to scale app installs profitably or have a betting audience. A real edge would also help, such as a verified track record or an underserved sport.

Already selling this:RithmmPropsBotOutlierPickFinderProps.CashStat Sniper

05/Receipts

Checked 48m ago by Claude Opus 5.5 on high effort, with web search. Confidence: medium.