Competitor analysis that changes something

A competitor analysis is easy to produce and easy to waste. You can fill thirty slides with screenshots of other people's Instagram grids and change nothing about how your own business operates.

The version I ran for an insurance-approved repair centre was useful because it was scoped to produce three specific artefacts, not a document.

Start from the decision, not the competitors

Before looking at anyone else, write down what you would change if the research said X. If you cannot name a decision the analysis could plausibly alter, you are producing a report, not doing research.

For the repair centre the open questions were narrow: which search terms should the site target, what proof does a customer need before they trust a bodyshop, and what should we photograph on every job.

Turn findings into things that persist

Research that lives in a slide deck decays. Research that lives in the site, the workflow and the shoot checklist does not.

• The keyword work went into the site copy, where it keeps working after the project ends
• The trust findings became a decision to publish the Volvo Approved Bodyshop accreditation as third-party proof at the consideration stage, rather than as a badge in the footer
• The photography findings became a shoot-and-edit checklist the wider team adopted, so every future job produces usable assets without anyone re-deriving what to capture

The third one is the one that compounds. A checklist adopted by other people outlives the project that produced it.

Be honest about what you did not find

Some questions the research could not answer, and saying so is more useful than filling the gap with a confident guess. If you do not have conversion data, do not present engagement numbers as though they were conversion numbers. A recommendation with a stated confidence level is actionable. One with borrowed authority is a liability.

The test

Six months later, can you point at something that is different because of the analysis? If the only answer is "we understand the market better", it did not work.