Reading a competitor's public artifacts is not the same as beating them

A competitor has spent years and a lot of money learning things about your market. Most of what they learned is sitting in public, in artifacts they publish for other reasons. Reading those artifacts properly is the cheapest research available to you, and it is not the same activity as making a feature comparison table.

You are not looking for a gap. You are reading somebody else's lab notebook.

What to take away

What their public artifacts actually record

Every one of these was produced under commercial pressure, which is what makes it informative.

The pricing page is a record of segmentation decisions. Which axis they charge on, where the tier breaks fall, and what they moved behind the top tier all tell you what they discovered buyers will pay more for. When a feature sits in the highest tier, someone found out it correlates with budget.

The onboarding flow is a map of where they believe people get stuck. Every extra step in a signup exists because removing it cost them something. Every hand-holding prompt marks a place customers got lost.

The changelog shows which complaints were loud enough to fund. Read a year of it in one sitting and the pattern of what customers actually pushed for becomes obvious in a way no single entry conveys.

Job listings tell you where money is going now, before it shows up in the product. A run of hires in one function is a strategy announcement with a salary attached.

Support documentation is the failure catalog. The articles that exist describe the things that go wrong often enough to be worth writing about.

Cancellation and refund policy is a small window on churn anxiety. Long lock-ins, aggressive win-back offers, and hard-to-find cancel paths all suggest a retention problem somebody is managing.

Case studies and logos tell you who they wish they sold to, which is not always who they sell to. Look for the mismatch between the customers on the homepage and the customers in the support forum.

Formal filings and registers are the artifacts nobody reads. Where a competitor is a listed company, it describes its own market in its own words, with dates, and full text search across electronic filings will find the passage rather than the company. Smaller firms leave a thinner trace, but the goods and services claimed in a federal trademark filing still describe the market somebody thought they were entering.

When you find no competitors

"There are no competitors" is almost never the finding it feels like. It usually means one of four things, and they call for opposite responses.

The problem is real but not worth paying for. Common, and the most likely explanation by default. People live with it. It annoys them for ten minutes a week and then the day moves on. Whether a cost is real enough to buy against is the test set out in problem discovery.

The buyers exist but cannot be reached economically. Somebody probably tried and quietly gave up. The absence of competitors is then evidence about distribution cost, not about demand, which is the third of the four questions market validation separates out.

You have drawn the category so narrowly that it has no name. Widen the description until you can name what people currently use. There is always something.

It is genuinely new. This happens. It usually requires something to have recently changed in the world that made it possible or legal or cheap, and you should be able to name that change in one sentence. If you cannot name it, be suspicious of your own answer.

The test that separates these: find the substitute. Everyone with the problem solves it somehow, even if the solution is bad. Find out what they do now. If the honest answer is that they do nothing at all and are content, that is your finding, and it is the most valuable negative result you can get this cheaply.

Substitutes count, and they usually win

The thing you are actually competing with is rarely another product in your category. It is the spreadsheet somebody built and now maintains. It is the junior member of staff who does it manually. It is a phone call to a person who knows. It is a general-purpose tool bent into shape. It is doing nothing and absorbing the cost.

These substitutes have advantages that are easy to miss because they are not features. They are already paid for. Nobody has to be persuaded to adopt them. They are infinitely flexible, because a human being is doing the adapting. They carry no procurement process, no security review, and no risk of looking foolish for having chosen them.

Beating a competitor's product is a design problem. Beating a habit that nobody has to justify is a harder one, and it is the one most first products face.

Switching costs are the real barrier

Whatever exists now has accumulated a set of costs to leave. List them explicitly for the substitute you are actually replacing:

Add them up honestly. Your advantage has to exceed the total, not exceed the competitor's product. This is why obviously better products lose, and it is worth deciding early whether you are prepared to do the unglamorous work of reducing switching costs rather than the pleasant work of adding capability.

Talk to their customers, not about them

The single most useful conversation in competitor work is with someone who bought a competing product. Three questions carry most of the value:

The customer interview rules apply here without exception. Ask about what happened, not what they think of the competitor.

Reading without copying

A competitor's feature set is the output of decisions you cannot see: their funding, their existing customers, their technical debt, a large client who demanded something, a founder's preference. Copying the output imports constraints that were never yours.

The useful move is to work backwards. For any feature that surprises you, ask what would have to be true about their customers for that to be the right choice. Then ask whether it is true about yours. Sometimes the answer explains the whole market. Often it explains only their situation, and you can safely leave the feature alone.

Be equally careful with what they do not do. An obvious missing thing is sometimes an opportunity and sometimes a hole they already fell into and climbed out of.

What this can and cannot settle

The line to hold: competitor research narrows what you need to test, but it never replaces the test. Everything you learn here is about other people's customers.

The only way to learn about your customers is to put something in front of them that they can turn down. Whether you can act on any of it is separate, covered in founder readiness.

Common questions

Should I sign up for a competitor's product?

Yes, under your own name, on a real trial, doing a real task, and log where you got stuck. Do not misrepresent who you are to obtain a demo or a quote. In a small market the reputational cost is real, and a conversation built on a false premise teaches you less anyway.

How many competitors should I study properly?

Three, thoroughly. Take the one buyers name most often, the one they compare on price, and the substitute that is not a product at all. Twelve at a glance is worse than three in depth.

What if a competitor already does exactly what I planned?

Then somebody has established that people pay for this class of thing, which is more than most ideas can claim. Go and find out what their customers still do by hand. The surviving workaround is where the next offer lives.

Does a well funded competitor mean the market is closed?

No, and nothing public tells you whether a market is saturated. Funding tells you somebody believes. It says nothing about how well they serve the segment you can reach or how much of their revenue is at risk.

How often should I redo this?

Ten minutes a month on three artifacts: the pricing page, the job listings, and the changelog. Direction only appears over time, and a single reading cannot show it.