A useful competitor analysis answers three questions: what are they saying that we are not, where are they getting attention, and what are they deliberately not doing. It takes a day for three or four competitors, and it produces decisions rather than a document.
The failure mode is scope. Ten competitors across twenty dimensions produces a matrix that is impressive, unfinishable and never used.
Pick the right competitors, which are not the obvious ones
Your competitors are whoever your customer considered instead of you, and that list usually contains something that is not a company.
Include the two or three direct rivals, one adjacent alternative that solves the problem differently, and the do-nothing option, which is the one that wins most often and is almost never analysed.
The fastest way to get this right is to ask five recent customers what else they looked at. The answers are frequently not what the team assumed.
What to record, and what to skip
Limit yourself to things that change a decision.
- The first sentence on their homepage: their claim, in their words
- Who they say it is for, and who they visibly exclude
- Pricing, if public, and what the tiers are gated on
- Their top ten organic pages, which tells you what actually earns them traffic
- What they publish weekly, and what they have stopped publishing
- Their reviews, filtered to the negative ones, which is where the honest product gaps live
Where the useful information actually is
Two sources are worth more than the rest combined and both are free.
Their negative reviews tell you what their customers wanted and did not get. That is a list of positioning openings written by the people you are trying to win.
Their changelog, careers page and recent posts tell you where they are going. A company hiring three enterprise salespeople is leaving the small end of the market, whatever their homepage still says.
Reading their content without copying it
Look at which of their pages rank, then ask why. Usually it is a term they covered early and thoroughly, and the lesson is the pattern rather than the topic.
Copying their best page is the standard move and it rarely works, because they have the head start and the links. The better move is to find the question their page raises and does not answer, and to own that instead.
Turn it into three decisions or throw it away
The analysis is only finished when it has produced decisions. Force three.
One thing to say differently, because everyone in the category says the same sentence and it has stopped meaning anything. One gap to occupy, where demand exists and nobody is answering well. One thing to stop, because a competitor is clearly better at it and the fight is not worth having.
If a day of work produces none of those three, the honest conclusion is that you are less differentiated than you thought, which is itself the most useful finding available.
How often to redo it
Twice a year is enough for most markets, plus whenever a competitor does something loud. Monthly monitoring of rivals is a way of being reactive on a schedule.
Set a recurring reminder to check three specific things rather than repeating the whole exercise: their homepage sentence, their pricing, and their most recent five posts. Changes in those three predict the rest.
Common questions
- Who counts as a competitor?
- Whoever your customer considered instead of you, which usually includes something that is not a company. Include two or three direct rivals, one adjacent alternative, and the do-nothing option, which wins most often and is almost never analysed.
- What is the most useful thing to look at?
- Their negative reviews. That is a list of positioning openings written by the people you are trying to win, in their own words, for free.
- How often should I redo a competitor analysis?
- Twice a year for most markets. In between, check three things only: their homepage sentence, their pricing, and their last five posts. Changes in those predict the rest.
Keep reading