← All posts

How to run a competitive teardown in 30 minutes.

Most competitive analysis takes four hours and produces a deck nobody reads. Here's the prompt that does the work in thirty.

Most teams spend four to six hours on competitive analysis. They open seven browser tabs, paste screenshots into Notion, build a comparison table, write an executive summary, and produce a deck that nobody on the leadership team actually reads. The work product is the procrastination.

This is the prompt-and-process I use to do the same job in thirty minutes. Same depth, less theatre.

What competitive analysis is actually for

One thing: finding where to position. Not "knowing the competitive landscape." Not "having a deck for the board." Finding the spot in the category where you can credibly own ground.

Everything in the teardown should serve that. If a section doesn't help you decide where to position, cut it. The "feature matrix with 47 checkmarks" doesn't help. The "positioning axes the category clusters on" does.

The Competitive Landscape Teardown prompt

The free version is here. The prompt forces five things:

1. You have to paste real competitor content (URLs, hero copy, pricing pages, G2 reviews). No "guess at the competitive landscape."
2. For each competitor, you get five fields, positioning, ICP, pricing model, three strongest signals, two structural weaknesses.
3. A comparison matrix that doesn't compress into uselessness.
4. A 150-word executive summary that names where the white space sits.
5. Three positioning hypotheses, each with a kill criterion.

The kill criterion is the part most competitive teardowns skip. A positioning hypothesis without a kill criterion isn't a hypothesis, it's a wish. The kill criterion turns the deck into a testable thing: "we'll test this positioning for 90 days; if X happens in user research, kill it."

The inputs you need to gather

Per competitor (10-15 minutes total)

• Homepage hero + sub-hero copy
• Pricing page (the whole thing, tiers, fine print, the comparison tables)
• About / story page
• 2-3 G2 or TrustRadius reviews (both sides, what people love, what they hate)
• A paragraph from their last earnings call if they're public
• Any recent press or product launch announcement

Don't try to scrape it all. Open each site, grab the visible text into a single document, move on. The goal is real signal, not completeness.

About your own company (5 minutes)

• Your current positioning (one sentence)
• Your ICP (one sentence)
• What you can credibly defend (real product moats, not aspirational ones)

Without your own context, the teardown will produce positioning hypotheses you can't actually take. Claude doesn't know your engineering velocity, your data moat, or your team's edge. You have to tell it.

Run the prompt

Paste the Teardown prompt. Paste all the competitor content you gathered. Paste your company context. Hit go.

About 30-60 seconds later, you have the output. Read it once. Some things to verify before you trust it:

• Every pricing claim. Did Claude infer pricing it couldn't see? If yes, the prompt should have marked it [INFERRED, verify against {{competitor}}'s pricing page]. If it didn't, push back and ask for the inference flags.
• The "structural weaknesses" claims. Are these real gaps in the competitor's offering, or things they explicitly don't do because their ICP doesn't want them? "Asana doesn't have a free tier" is a fact. "Asana is bloated for small teams" is a take. The take needs to be defensible.
• The white-space framing in the executive summary. Is the "nobody owns X" claim actually true? Sometimes Claude over-claims white space that someone is already entering, flag those for verification.

The positioning hypotheses are where the conversation starts

This is the bit most teams skip in their version of competitive analysis: actually using the output. The three positioning hypotheses are designed to be talked about, not filed.

For each one, the conversation in the next leadership meeting should be:

1. Could we credibly take this position? (Honest answer, not aspirational.)
2. What would happen in user research / sales calls in the first 60 days that would prove it works?
3. What would kill it? (The kill criterion the prompt gave you, but adapt it to your reality.)

That conversation, with the matrix and the hypotheses on the table, takes maybe 45 minutes. The whole thing (gather inputs, run the prompt, hold the conversation) is under two hours. Same job. Less procrastination.

What I'd cut from the traditional process

• The slide deck. Send the brief in a shared Google Doc. The conversation matters more than the format.
• The 47-checkmark feature matrix. Reduce to the 4-5 axes that matter for positioning.
• The "history of the category" section. Nobody reads it. Nobody decides anything from it.
• The "trends shaping the future" section. Trends are noise unless they map to a competitor's strategic move.

What you keep: the positioning hypotheses, the kill criteria, and a one-page summary anyone can read in 60 seconds.