Competitor teardown

Use this when someone asks about a competitor. A teardown that ends in observations has failed; it ends in a decision.

Work from what they say in public

Four sources, in this order, because each one contradicts the last in a useful way:

1. Homepage and pricing page — what they claim, and who they will take money from. Pricing tells you the customer more honestly than the copy does.
2. Changelog, blog and job ads — where they are actually spending. A posting for three enterprise sales roles says more than a positioning statement.
3. Reviews and support forums — the complaints that repeat. One angry reviewer is noise; the same complaint fifteen times is a roadmap.
4. Their own customers talking unprompted — social, communities, podcasts.

The five questions

Answer these and nothing else:

• Who is this obviously built for, and who is it obviously not built for?
• What is the one promise on the homepage, in their words?
• What do their customers complain about that they cannot fix without changing what they are?
• What do they do better than us, stated plainly?
• What would we have to become to beat them, and do we want to be that?

The fourth question is the one people skip. A teardown with no honest praise in it is not research, it is reassurance.

What to hand back

A page. Their promise, their customer, three strengths, three structural weaknesses, and one recommendation with a reason. Mark every claim as observed or inferred — a reader needs to know which parts would survive a phone call.

What not to do

Never state a competitor's revenue, headcount, churn or customer count as fact unless it is published — say "inferred from" and name what you inferred it from. Never repeat a rumour. Never write a comparison that a competitor could sue over, and never recommend copying a feature without saying what it would cost us.