How to Validate a SaaS Idea Before You Write Code
To validate a SaaS idea, prove three things before you write code: a specific group describes the problem and the workarounds they hate, they already pay a monthly price for something close, and you can reach them through communities or search. Read Reddit threads and comments for the pain, list the competitors and their plans for the price, then run a landing page with a waitlist and a small ad budget for a week. Build only when people sign up at a cost you can live with.
A SaaS needs no inventory and no factory, so it takes little to start. That is also why it goes wrong so often: code feels like progress, so founders build first and look for buyers later. The cost is months of evenings spent on software that nobody pays a monthly fee to use.
What makes SaaS validation different?
A SaaS asks for money every month, so the bar is higher than a one-time sale. The buyer has to feel the problem often, not once. And most software problems already have a free option: a spreadsheet, a notes app, a general tool that does most of the job.
That changes what you look for:
- A problem frequent enough to justify a subscription.
- The tool people use today, and what they dislike about it.
- Competitors that charge monthly, and at what level.
- Who pays: a person with their own card, or a team that needs approval.
Where do you find proof of the problem?
Start with conversations, because that’s where people describe their workarounds. Reddit is a rich source of software pain: people post their stack, their frustration and what they tried. Search the job the software would do (“track client hours”, “send invoice reminders”), not the category name.
Look for three patterns in comments:
- A workaround. Someone describes doing the job with a spreadsheet and a reminder app. That’s a job done by hand.
- A complaint about a named tool. Someone says a known tool costs too much for a freelancer, or does everything except one job. That’s room.
- A request. Someone asks if anything does the job. That’s demand you can link to.
How to find pain points on Reddit shows the search terms and how to tell real pain from polite interest. X posts help too, especially for developer and founder tools, where people complain in public.
How do you check that buyers pay monthly?
List every competitor you can find, then write down their entry plan and their middle plan. If several tools charge monthly for the same job, the market accepts a subscription. If everything close is free or a one-time purchase, a monthly price will be a hard sell.
| What you find | What it means for a SaaS |
|---|---|
| Several tools with monthly plans | Buyers accept a subscription; look for a gap |
| One giant tool everyone uses | Hard to be seen; sell to a group the giant serves badly |
| Only free tools and templates | Payment is unproven; test a price early |
| “Alternatives” pages with complaints | Room: people are actively looking to switch |
| Nothing at all | Either a new problem or no buyers; find the conversation first |
Search Google for “[the job] software” and “[the main competitor] alternatives”. Alternatives pages and review sites (G2, Capterra, Product Hunt) show what people switch from and why. How to find competitors for a product idea covers the search patterns in detail, and how to price a new product turns competitor plans into your first price.
Ads are another signal. A software company running the same Meta ad for months is paying to acquire customers and, most likely, getting them back. The Facebook Ad Library guide shows how to read run length.
What does weak evidence look like?
Here is one example check Vettoz ran on October 18, 2026: a SaaS that emails remote teams a daily reminder to drink water, with a team leaderboard. It scored 46 out of 100, a “rethink”.
The reason is useful. There was plenty of conversation, but most of it was about something else: threads about electrolyte drinks and supplements, not about teams or managers. No competitor showed a price. Reach was weak because the communities that came up didn’t point at the buyer (remote team managers). The check flagged the evidence as off topic, which keeps the verdict at “test” at most even when the numbers look big.
That’s a common SaaS trap: a real-sounding problem with lots of chatter that isn’t about your buyer. Signs your product idea is bad lists this and other patterns.
Which test settles a SaaS idea in a week?
Once the evidence says “worth testing”, run one small experiment with a number decided in advance.
- Write a one-page landing page with the promise in the buyer’s own words (lift phrases from the comments you found), one screenshot or mock, and a waitlist form.
- Send traffic you control. A small search or Meta ad budget, posts in the subreddits where the pain showed up (follow each community’s rules), or 20 direct messages to people who described the problem.
- Decide the go number first. For example, a cost per signup you can live with, or a share of conversations that end with “tell me when it’s ready”.
- Ask for money early. A pre-order or a founder price on the waitlist tells you more than any number of signups.
If nobody signs up, change one thing (the buyer, the promise or the channel) and run it again before you write code.
Should you build an MVP to validate?
Usually not first. An MVP answers “does this solution work?”, but the earlier question is “does anyone want this enough to pay?”. Evidence and a landing page answer that in days. Build the MVP after people have signed up, and build it for them.
Our general guide on how to validate a business idea before building covers the same five checks for every kind of product.
FAQ
How many signups prove a SaaS idea?
There’s no universal number. Decide yours before the test from what a customer is worth to you and what each signup costs.
Is a market with big competitors a bad sign?
It proves people pay. You need a group or a job the big tools serve badly, and the complaints in their alternatives pages usually show it.
Should I validate with friends?
Friends are polite. Validate with strangers who already have the problem and already spend time or money on it.
Can I validate a B2B SaaS the same way?
Yes, with more weight on direct conversations, since a small number of buyers can make the business work.
Do I need a tool for this?
No. You can do every step by hand in a few hours. A tool saves the collection time and keeps you honest with a score.
Check your SaaS idea with Vettoz
Vettoz reads the market for a SaaS idea inside Claude, ChatGPT or Cursor: Reddit threads and comments, X posts, videos, active Meta ads, and Google results for software and alternatives pages. He scores pain, demand, pay, room and reach with the evidence linked, and your AI writes the report with a price and a test. The setup guide takes a few minutes, and the plans start with free checks.