How to tell if people will actually pay for your startup idea

The polite 'I'd use that' answer is the most expensive response a founder can collect. It costs you a weekend of building, then another, then the slow discovery that 'use' and 'pay' are not the same word. People will pay when a current workaround hurts enough that switching feels cheaper than staying. Until you see that switch in their behavior, you have a hobby.

A pay signal is a behavior, not a sentence. Quoting a price they would accept. Bookmarking a competitor's pricing page. Naming the exact amount they currently spend on the workaround. A comment that says 'I'd pay for this' without a number is enthusiasm. A comment that attaches a recurring amount to the product is a market signal you can act on. This piece walks through the pay signals already hiding in the research most founders have already collected, and three tests that turn vague interest into something you can compare.

Pay signals are actions, not opinions

Enthusiasm confirms a problem exists. Pay signals confirm a business can exist. Someone saying they like your idea is enthusiasm. Someone quoting a price they would pay, or switching away from a competitor's pricing page, is a pay signal. Observable behavior beats stated preference every time.

The split is the same one every sales team has known for decades: what people say and what they do. A founder who runs twenty customer interviews and collects twenty 'sounds great, I'd definitely use it' answers has run a popularity contest. A founder who runs twenty interviews and watches how many people name a dollar amount, click a pricing link, or describe a tool they already pay for has run a different experiment.

The framing matters because the work differs. Enthusiasm tells you to write better copy. Pay signals tell you what to charge and who to charge it to. The first exercise is a mood. The second is a market.

A research board that only returns enthusiasm has told you nothing. Confidence is computed from the signals you can verify, not from the kind words in your inbox.

What existing research looks like under the pay-signal lens

Review each source you already have for behavioral signals, not just thematic relevance. A thread where people name their current workaround is stronger than one describing the pain in general terms. A pricing-page bookmark is stronger than an open tab. Research tools that synthesize across multiple sources can surface patterns manual review misses. Look for evidence of commitment, not just expression.

Your existing research may contain pay signals you did not know to look for.

Most pre-build research settles into the same five sources: a Reddit thread, two or three product reviews, a competitor's landing page, a Discord or Slack screenshot, and the resulting notes in a doc. Under a thematic read, all of that answers one question — does the problem exist. Under a pay-signal read, it becomes a different set of questions. Does the thread mention what people currently pay. Does the review mention cancellation. Does the screenshot show a usage cap. Each of those is a different kind of evidence.

Tools that synthesize across multiple sources can surface patterns a manual review misses. mvanhorn/last30days-skill pulls material from Reddit, X, YouTube, HN, and the web, then writes a grounded summary. assafelovic/gpt-researcher runs an autonomous research agent against a query. dzhng/deep-research is the lighter open-source take on the same idea. khoj-ai/khoj, marketed as an AI second brain you can self-host, keeps the same questions running against a growing corpus. The point is not which tool you use. The point is the question you ask it. Asking 'what is the pain' returns a thematic read. Asking 'what do people currently pay, and what do they complain about' returns a pay-signal read.

Every quote you collect should land on your board copied from the page you can open, never written by the model. The model never writes a quote.

Pricing-page behavior on competitor sites

Visit the pricing pages of two or three tools your target users rely on. Note whether there is a free tier, what the lowest paid tier costs, and whether users complain about price in reviews or support forums. If people negotiate price in public threads, that is a pay signal. You want a market that pays and still complains, not a market that has never paid for anything in this space.

Complaints about current pricing outweigh complaints about product quality.

A market that has never paid for a category is not your market. It is a market full of users who will demand a free tier and resent you the day you introduce one. A market that already pays, even badly, has accepted that the category is worth money. Your job is to be a better version of what they already pay for, or a different one they will switch to.

The signals to look for are blunt. People arguing about whether the cheapest paid tier is worth what it costs. Threads titled 'is [tool] worth the price.' Commenters comparing the cost of two tools they already pay for. A common pattern is the user who has outgrown the cheapest tier and is shopping for a way to avoid the next one. That user has a budget, a current bill, and a reason to switch. That is a pay signal with a price tag attached.

Two tests that matter on any pricing page: does the cheapest paid tier cost enough to make a user think before clicking, and do people in public threads defend or attack that price. A market that ignores price has not decided yet. A market that argues about price has decided.

For B2B tools, your buyer may already have a line item you never see. Funded competitors, public programs, and visible revenue in the category all point in the same direction: someone has decided this category is worth money. The European Innovation Council, as one example, blends a non-dilutive grant of up to €2.5 million with equity of €1 to €10 million through its STEP Scale-Up route. A founder pitching a tool to a funded buyer is talking to someone with a real budget, not a hobbyist. The same pattern shows up in any category where the buyer has revenue, has been funded, or is replacing a tool they already pay for.

Community threads where people name a specific cost

A comment attaching a dollar amount to a capability is a stated price point. When multiple people in different threads name similar amounts, you have a signal. If no one names an amount, the interest is real but the business model is unvalidated. One named price point is useful. Several matching price points is a market signal.

Without a named price, you have interest. With matching price points across sources, you have a direction.

The pattern shows up in a few predictable places. Indie communities talking about tools they would pay for versus tools they currently pay for. B2B communities comparing per-seat costs across the category. Consumer threads where users name what they would pay for a missing feature. Read these with a single question: are the named amounts clustering, or scattering. Clustered numbers across independent sources are the closest you get to a market-clearing price before you have a product.

The absence of a number is itself a signal. If you read fifty threads about a problem and not one person names what they would pay, you have evidence of pain and no evidence of a market. The next move is not to invent a price. It is to run a test that surfaces one.

A spreadsheet with one column for source, one for the exact quoted amount, and one for what the user gets in return will tell you in a single glance whether your evidence clusters or scatters. If it scatters, your price is a guess. If it clusters, you have an anchor. That is what 'from raw to well done' looks like in practice — five minutes of structure on top of an hour of reading.

Three moves that convert interest into comparable intent

After finding pay signals in your research, run one of three tests before writing code. A price-anchor test: add a pricing page with a tier at the high end of what you found, and watch whether visitors stay. A pre-order page: describe the product as it will exist, attach a price, and count email signups. A refundable deposit: offer early access at full price with a full refund guaranteed at launch. The refundable deposit is strongest because it requires real money for something that does not exist yet.

A deposit requires a decision. A pre-order requires commitment. A pricing visit requires curiosity.

The three tests are not equal. They live at different points on a scale of how much you are asking the visitor to give up, and how much you can learn from the answer.

Pick the test that matches the evidence you already have. If your research is full of named price points, jump straight to the deposit. If you have scattered signals, the pre-order is the right intermediate step. The pricing visit is something you do almost as a habit, on every idea you consider, because it costs nothing and tells you something.

A research board that tracks the count, the price, and the date of each commitment is the closest a pre-build founder gets to a forecast. Confidence is computed from that record, not from how the post felt to write.

The bottleneck for most startup ideas today is no longer writing the code. Fully AI-generated code went from 1% to 27.6% of all pull requests in the past year, and the bottleneck moved from writing code to validating it. The test you run before you build is now the test that determines whether the build is worth it. Validate your startup idea before you build it.