How to Research Your Competitors Before You Build (Without a Feature Matrix)

The standard advice is to build a feature matrix, and the standard outcome is a spreadsheet of ticks that tells you nothing you can act on. Every tool has the features. What differs is who each one fits, what it is quietly bad at, and why people leave. That information exists, it is public, and almost none of it is on the marketing sites you have been reading.

Table of contents

• What you are actually looking for
• Finding the real list, not the obvious one
• Where the honest information is
• Reading a pricing page properly
• The "no competitors" trap
• Turning it into a position
• Frequently asked questions

What you are actually looking for

Three things, and none of them is a feature list.

• Who each tool is really for. Not who their homepage says. The size of company, the job title, the budget. Two products with identical features and different buyers are not competitors.
• Why people leave. The specific, repeated complaint. Price rises after a threshold, an integration that breaks, support that stopped answering, a rewrite that lost features. This is the whole prize.
• What they will not do. Every mature product has things it structurally cannot build, because of its architecture, its bigger customers, or its business model. Those are the gaps a small product can actually hold.

A feature matrix tells you none of these and gives you the comforting feeling of having done research. If the exercise ends with "we should add SSO too", it has gone wrong.

Finding the real list, not the obvious one

You already know the two market leaders. They are rarely who you lose to. The real list comes from three places:

• Your interview notes. The question "what do you use for this today?" from talking to users before you build produces the honest answer, and it is usually a spreadsheet, a notebook, or a colleague. Those are your real competitors and they are undefeated.
• The alternatives pages. Search for "alternatives to" plus the market leader. Someone has already assembled the list and ranked it, and the comment sections underneath are more honest than the articles.
• What people recommend in public. Search the communities where your buyers are for the problem, not the product. The tools people suggest to each other unprompted are the ones with real traction, which is not always the ones with the best marketing.

Five to eight names is plenty. Beyond that you are collecting rather than deciding.

Where the honest information is

Their marketing site is the least useful page they own. These are the ones that tell you something, roughly in order of value:

Keep one document with a line per complaint and a link to the source. When the same complaint appears from three unrelated people about the same product, you have found something real rather than one annoyed customer.

Signing up for a trial is worth doing once, for the onboarding rather than the features. The first ten minutes of a competitor's product show you what they think matters, and where they lose people.

Reading a pricing page properly

A pricing page is a description of the customer, not a number. Read it for three things.

What the tiers are measured in. Seats means they sell to teams. Contacts or records means they sell to people whose usage grows. Whatever the meter is, that is what they think you value, and it is the assumption you can disagree with.

Where the cliff is. Most complaints about price are not about the starting number, they are about the jump when you cross a threshold. Find that threshold. Founders who get stranded just above it are the most motivated buyers you will ever meet.

What is missing from the cheap tier. If the thing your buyer needs most is gated two tiers up, that is a position, not a feature request. There is more on setting your own number in how to price your first indie SaaS.

The "no competitors" trap

If you cannot find competitors, one of two things is true, and only one of them is good news.

Either you are searching for your solution instead of the problem, which is the usual cause. Nobody has built your exact thing, so nothing matches your search terms, while five products solve the same job differently. Search for what the person is trying to achieve and the list appears.

Or the market genuinely is empty, which is much more often a warning than an opportunity. Someone has usually tried. Look for the abandoned products, the archived repositories, the forum threads from four years ago. Finding out why the last three attempts died is worth more than any competitive matrix.

The healthy version of "no competitors" is a crowded market where everyone serves a different buyer than yours.

Turning it into a position

The output is one sentence: for whom you are better, at what, and what you give up in exchange. The last part is not optional. A positioning statement with no tradeoff in it is marketing copy, and buyers discount it automatically.

Two disciplines make this honest. Say plainly what the competitor does better, because everyone comparing you already knows and your admitting it is the only thing that makes the rest credible. And pick one axis, not four. Cheaper, simpler, faster, and more powerful is not a position, it is a wish.

This also decides scope. Once you know the one axis, the cutting exercise in how to scope your MVP gets much easier, because anything that does not serve that axis is an obvious cut rather than a painful one.

Frequently asked questions

How long should this take?

A day, spread over a week so you notice patterns. It is not a project, and treating it as one is a common way to delay building while feeling productive.

Should I write public comparison pages?

Yes, once you have something to compare. They rank well because people search for exactly that, and they work only when honest. A comparison page where you win every row convinces nobody and reads as an advert.

What if a competitor copies my idea?

Larger products copy features slowly and rarely copy positioning at all, because serving your narrow audience properly would cost them their existing one. The idea is not the moat, and worrying about this before you have users is time spent on the wrong problem.

Do I need paid research tools?

No. Everything described here is free: review sites, forums, changelogs, job pages, and search. Traffic estimation tools are interesting once you are choosing between markets, and irrelevant when you are choosing what to build.

Should I talk to their customers?

Talk to people with the problem, which usually means talking to their customers. Ask about their work, not about the competitor, and do not use the conversation to pitch. The rules are the same as any other user interview.

Competitor research goes wrong when it becomes a list of features to match, because matching a bigger product later and with fewer people is the one strategy guaranteed to fail. Done properly it produces the opposite: a short list of specific, repeated complaints, and a clear sense of which one you intend to fix. That is a position, and it decides both what you build and who you talk to.

Lighthouse asks your survey question on the signup form, so when people tell you what they use today the answers sit next to the emails and the competitor list writes itself. From an indie dev, for indie devs and makers.