How to Find SaaS Ideas From Competitors: 8 Patterns and 2 Frameworks From 100 Founder Interviews

The short version: A former quant trader spent two and a half weeks studying every screen of a habit-quitting app — then rebuilt it for a different audience: women quitting sugar. Five months later, the app was making $12,000 a month. He never invented anything.

I went through 100 founder interviews from Starter Story, and that pattern repeats across the whole dataset. The founders who found ideas fastest did not invent. They took products that already had paying customers and found the part nobody was serving well. Below are 8 patterns that worked, 5 free places to dig for gaps, and 2 step-by-step frameworks — plus the traps that turn a promising gap into a dead end.

Why is "no competitors" actually a red flag?

Finding a gap means looking inside a product that already has paying customers for the part nobody serves well. A price. A feature. An audience. A platform.

Most first-time founders read an empty market as good news. The interviews point the other way.

"Not having competitors means that there's no market," says Thomas, who built Unid, a launch platform doing $10,000 a month. "You're not Steve Jobs. Creating a new market is nearly impossible. If you don't have competitors you don't have a market, and your idea won't sell."

He learned that lesson the expensive way. Before Unid, he spent two years on Gum Affiliates — a marketplace he had spotted with no competitors at all. It made $500. You read that right: two years, five hundred dollars.

Anton learned the same lesson from the other direction. He spent 15 years trying to build something unique — six or seven startups, all original. "I was always trying to make something unique," he says. By his own account, the result was products people did not really need. Then he did the opposite: took a crowded category, stripped it down to one thing, and built Letterly — $250,000 a month.

A market with no competitors is usually a market with no customers. The gap has to sit inside proven demand, not replace it.

If you already have a pain point of your own, the other route in this series starts there — validating an idea from your own pain point. This article is for the opposite case: no pain point, no audience, no obvious idea. Just a market full of products people already pay for.

What 8 patterns actually worked?

I sorted every case by the move the founder made. Eight patterns came up — some repeated by multiple founders, some from a single case. Every figure is self-reported by the founder.

The pattern behind the patterns: none of these eight started from a blank page. Each started from something already working — and one specific reason it was not working for everyone.

This is one route out of the full list of patterns and angles from those 100 interviews, and it is the route that needs no personal pain point.

Where do you find these gaps for free?

You do not need paid tools to dig any of these up. All five are places where complaints pile up in public — with the first step for each.

1. Review sections — read the worst reviews first.

Will Cannon, who built two companies this way, goes straight to the bottom of the scale. His method: read the 1- and 2-star reviews, and look at what people hate about the top players in the industry — those complaints, he says, point to the easiest customers to win.

Cedric ran the same play at scale: he pulled every review he could find for apps in his category, fed the complaints into Claude and ChatGPT, and shaped his product around the output. Seven weeks after launch, his app had made $50,000.

*First step: open G2 or your app store category, sort by lowest rating, and read 20 reviews. Write down every complaint you see twice.*

2. The platform's own forum.

Before Andy Cloak built Data Fetcher, he read Airtable's community forum to see what problems people kept running into. The same request appeared again and again: a flexible way to pull data in from other tools. That one thread led to a solo business at $23,000 MRR — and a product deliberately parked between scripting and no-code imports, a spot the platform itself was unlikely to build for.

*First step: search your target platform's forum for the most-upvoted feature requests that have been ignored for a year or more.*

3. Marketplaces where projects are for sale.

Adrian found his business on Acquire.com, a marketplace where founders sell their products. He filtered for listings with real revenue, reverse-engineered how the seller got customers, then built his own version. His scraping API now runs at $20,000 a month.

*First step: browse listings in a niche you understand — not to buy, but to read which products already make money, and how.*

4. Comment sections.

Michael found his app in the comments of TikTok videos: under every video about height, teenagers were asking how much taller they would grow. He built the prediction app, answered those comments, and had a sale two days after launch. Reported revenue: $100,000 a month.

Louis did the same in beauty TikTok — reading comment sections to see what women were asking for. His app hit 100,000 users in three days.

*First step: find 10 high-comment videos in your category and read the questions people repeat.*

5. Trends and search data.

David Adius validated his niche with two free tools: Google Trends, where "stop sugar" had been climbing for five years, and TikTok and Instagram, where a wave of creators was already talking about it.

Bhanu took the keyword route — filtering for terms with low competition and steady volume, then shipping a tiny free tool for each. Fifty small tools pull 50,000 visits from Google every month and feed his main product, at $13,000 MRR.

*First step: run your topic through Google Trends and check the five-year direction. Then search TikTok for the same phrase. Creators mean it is being thought about; repeated questions mean people are shopping for it.*

How do you evaluate a gap before you build?

Two founders in this set turned "find a gap" into a routine you can borrow.

Andy Cloak's six steps — for building on someone else's platform:

1. Find a platform that is growing fast.
2. Find a pain point on it. Read the forums, Reddit, and Twitter.
3. Borrow a proven pattern from a more established platform — for Andy, the Google Sheets add-on.
4. Check you can integrate: is there a public API, a marketplace, an extension SDK?
5. Do the napkin math. How many users does the platform have? How common is the problem? What will people pay?
6. Ask: will the platform crush you? Read the roadmap and the support forums, and work out whether they are about to build this themselves.

Step 6 is the one people skip, and it decides whether your business has a future. Andy's test for his own product: it sits between two capabilities the platform already has — close enough to be useful, awkward enough that they are unlikely to build it. His read: "It's safe for now." His framework is also the backbone of the separate guide on building on top of a platform's ecosystem, if platforms are where you want to play.

Adrian's eight steps — for rebuilding something that is already working:

1. Go to a marketplace where projects are for sale.
2. Filter for products with real revenue.
3. Pick a market you actually know.
4. Work out which product is behind the listing.
5. Study the site and the product.
6. Reverse-engineer how they get customers. Adrian calls this "arguably the most important part."
7. Build your version.
8. "Don't copy word for word. Don't copy everything literally... you just want to copy the concept."

That last step is the one that keeps you out of trouble — and Adrian's own version of it is blunter: "If something is working, you have a moral obligation to copy it." Together, the two lines point to one rule: take the concept, leave the execution. And if you want to go further than these two frameworks — getting paid before you build — that is covered in the article on pre-selling a SaaS product.

When does a promising gap turn into a trap?

A gap being real does not mean your version of it will work. Four traps from the same interviews:

You can copy the screens. You can't copy the reason people stay. Genuinely solving a real problem for real users is what works.

Where is the line?

One honest note: in 100 interviews, almost nobody talked about the legal side of this. No trademarks, no warning letters, no platform disputes. The founders talked about their methods, not their lawyers. So what follows is not legal advice — it is the line that shows up in what they actually did:

1. Take the concept, not the execution. Adrian's step 8. The code, the copy, the design files — that work stays yours.
2. Do not touch the brand. Nevo watched people remix Postiz's code and self-host it. Within weeks, they were gone. The brand was the moat, and it was the one thing they could not take.
3. Bring something the original did not have. EUform brought a lower price and a one-click migration tool. Stoppr brought a new audience. Data Fetcher brought a new platform. Yataphone brought a business model — pay-as-you-go in a market of subscriptions.

Every founder in this dataset took the structure. None of them took the brand.

If your plan touches trademarks, code licensing, or platform terms, talk to a professional before you build. That question is outside what this data can answer.

FAQ

Is finding an idea from competitors the same as cloning an app?

No. Every case here changed something material: the audience, the price, the platform, or the form. The people who took code directly — the Postiz remixers — all quit within weeks.

What if the platform builds my feature into its official product?

That is platform risk, and step 6 of Andy's framework exists for it: read the roadmap and the support forums, and pick a spot the platform is unlikely to build for itself — ideally something that sits between two of its existing capabilities.

Can I do this without paid tools?

Yes. Every source in this article is free: reviews, forums, marketplaces, comment sections, Google Trends. The only paid tool mentioned by any founder was a keyword tool for the search-data route.

How long before I know a gap is working?

Faster than you would expect. A fake demo video takes a day. Dennis's first sale came minutes after a Reddit post. Andy's step 6 — checking whether competitors already win through search — takes one afternoon.

Does this work outside the US or Europe?

The interviews suggest some of the strongest versions of this play work outside the English-speaking market: take a product validated in the US, rebuild it for a different language or country, where ad costs are lower. Gravl translated its app into Spanish and ran ads across South America. Stoppr targeted France. The giants rarely bother with those markets.

---

One more thing. The 8 patterns and the 5 digging spots are on one page in the checklist I put together from these interviews. It is free, and it costs you an email.

Get the checklist →

---

Related reading

• How to Find a SaaS Idea That Actually Makes Money
• How to Validate a SaaS Idea From Your Own Pain Point
• How to Find a SaaS Idea Without an Audience

*A note on sources: every case in this article comes from founder interviews published by Starter Story. I went through 100 of them and cross-checked the load-bearing numbers against the original transcripts. Results are self-reported by the founders.*