How to do a competitive analysis, step by step

A competitive analysis goes wrong the moment it becomes a feature checklist — a spreadsheet of who has what, ending in a resolve to match everything. That produces a feature-parity treadmill, not a strategy. A good competitive analysis answers a sharper question: given the field, where can we win, and where should we not even try?

1. Map the competitors — including the non-obvious ones

Start by identifying who you're actually competing with, and go beyond the obvious direct rivals. Direct competitors solve the same problem the same way. Indirect competitors solve it differently — a spreadsheet competes with a lot of SaaS tools. And don't forget the "do nothing" option: often your real competition is the customer's status quo and inertia, not another product. Missing the indirect and status-quo competitors is how analyses produce a falsely comfortable picture.

2. Choose dimensions that matter to customers

Compare on what actually drives the customer's decision, not on a generic feature list. Useful dimensions usually include positioning (what category they claim), target customer, pricing and packaging, go-to-market motion, integrations, and trust signals (security, social proof, reviews). A feature you both have isn't interesting; a dimension where customers feel a real difference is.

3. Ground it in current facts

The fastest way to a useless analysis is filling it from memory and assumption. Get current facts: real pricing from their pricing page, real positioning from their homepage, real strengths and weaknesses from their actual customer reviews. Things change — a competitor you "know" raised prices, pivoted, or added the feature you thought you uniquely had. Verify before you conclude.

4. Read their strengths and weaknesses honestly

This is where founders fool themselves. It's tempting to minimize competitors' strengths and exaggerate their weaknesses. Resist it — an analysis that flatters you is worse than none, because it builds confidence on a false picture. Name where they're genuinely better than you. That's where you either need to improve or deliberately cede ground.

5. Produce a strategic read, not a matrix

The deliverable isn't the comparison table — that's the input. The deliverable is a conclusion: where you're genuinely differentiated and should compete hard, where you're at parity and shouldn't bother, where you're exposed and need to defend, and the shared risks that hit the whole category. A competitive analysis that ends in a matrix is unfinished; one that ends in "here's where we play to win" is doing its job.