Most competitor analysis is theater: a feature matrix where your column is all green checkmarks, produced once for a pitch deck and never opened again. It convinces nobody and informs nothing.
Useful competitor analysis answers three questions: what do buyers in this market already pay for, what do they hate about it, and which of those complaints will you build your position on? Here is the framework we use.
Start with the right list
Your competitor set is whatever your buyer considers instead of you, which is rarely just software lookalikes. It usually includes an incumbent or two, a cheaper point tool, an agency or consultant, and the status quo: a spreadsheet, an intern, or doing nothing. Five to eight entries. The status quo goes on the list every time, because for early-stage products it wins more deals than any named company.
The six reads per competitor
- Pricing pages. What buyers pay, how the packaging works, where the price cliffs are. This anchors your own pricing and your TAM math (see TAM, SAM, SOM).
- Negative reviews. The one-to-three-star G2 and Capterra reviews are paying customers describing, in their own words, what they wish existed. Read for verbs: what were they trying to do when the tool failed?
- Positioning language. Their homepage headline tells you which buyer they optimize for, and therefore which buyer they are ignoring.
- Changelogs and roadmaps. What they ship reveals what their biggest customers demand, which predicts who gets neglected next.
- Hiring pages. Roles reveal strategy quarters early: a first enterprise AE, a compliance hire, a new region.
- Community complaints. Reddit, Slack groups, and forums carry the unfiltered version of the reviews, plus the workarounds people build, and a workaround is a feature request with proof of effort attached.
One output per competitor: a two-sentence card. Who they serve well; where their customers hurt. Resist the matrix.
Turn the reads into positioning
Lay the cards side by side and look for the overlap: the complaint that recurs across competitors, or the segment every incumbent underserves. That intersection is your entry point. Incumbents optimize for their largest accounts, which structurally forces them to neglect someone: smaller teams, a vertical, a geography, a workflow. Your position is not "better"; it is "built for the people they forgot", said in the words those people already use in reviews.
This also disciplines your ideal customer profile: the segment you found in the gap is the ICP, and the complaint you found is the first line of your outreach.
The mistakes that waste the exercise
- The all-green feature matrix. Built to flatter, not to decide. If your analysis contains no bad news for you, it is not analysis.
- Ignoring non-software competitors. The agency and the spreadsheet take more of your deals than the incumbent does.
- Copying incumbent positioning. If you describe yourself in their words, buyers default to the safer, bigger option. Own the gap language instead.
- Doing it once. Pricing moves, funding lands, roadmaps shift. A quarterly refresh, or a re-run triggered by your own campaign data, keeps the read alive.
Where this fits, and the shortcut
Competitor analysis is step three of the broader validation loop in how to validate a startup idea: demand first, buyer second, competition third, willingness to pay fourth. If you want the desk-research half done for you, the competitive intelligence module in Cafiyn Lens produces the teardown (including hiring-signal reads and linked sources) as part of every market assessment, and the resulting positioning feeds FlyWheel outreach directly, so the analysis becomes campaigns instead of a slide.