Competitive analysis
How to Run a Competitive Analysis in 90 Minutes
Most competitive analyses founders produce are lists of company names. The list looks thorough, but it does not change a decision: it does not say which wedge would actually open the market, which buyer is underserved by every alternative, or which claim to test next. This page is a 90-minute procedure that produces a different output — a one-page map that names the wedge, the alternatives, the alternatives' weaknesses, and the gap the new product would close.
Last updated · September 8, 2026
Quick answer
How does a solo founder run a competitive analysis in 90 minutes?
Eight blocks of 10–15 minutes each, run in order, on a single page. (1) Name the buyer in one sentence. (2) List the four kinds of competitor: direct, adjacent, status quo, substitute. (3) For each competitor, write the one-sentence claim they make to the buyer. (4) For each claim, write the assumption the claim rests on. (5) For each assumption, write the strongest counter-evidence you have seen publicly. (6) Write the wedge — the assumption that, if it broke, would let a new product win. (7) Identify the cheapest experiment that tests the wedge. (8) Write the next decision the experiment would inform. The output is a one-page map that a stranger can read in two minutes and that drives the next experiment.
Key takeaways
The page in 30 seconds
- Four kinds of competitor matter: direct (same product), adjacent (same buyer, different product), status quo (the buyer does nothing), and substitute (the buyer solves it manually with a spreadsheet, an agency, or a contractor).
- The 90-minute procedure is eight blocks of 10–15 minutes each. Each block produces one input; the eight inputs together produce the one-page competitive map.
- The map is a one-page table: the wedge, the alternatives, each alternative's one-sentence claim, the assumption behind the claim, and the counter-evidence. The output fits on one screen.
- The wedge is the assumption that, if it broke, would let a new product win. The wedge is the hypothesis the next experiment tests — not a feature list, not a positioning sentence.
- Five recurring mistakes: confusing 'no competitor' with 'no substitute' (most categories have a status quo), copying the incumbent's positioning, listing competitors without naming the assumption behind their claim, treating the analysis as a one-time exercise, and stopping before the wedge is written.
- The map is not the deliverable; the experiment the map produces is. If the procedure does not end in a named experiment with a named decision, it has not finished.
What it is
What a competitive analysis is — and is not
A competitive analysis is the procedure that turns 'we have competitors' into 'this is the wedge a new product would open, and this is the experiment that tests it.' The output is a decision-ready map, not a competitor spreadsheet. Porfer's 'How Competitive Forces Shape Strategy' (HBR, 1979; updated 2008) is the canonical reference for industry-level competitive analysis; April Dunford's 'Obviously Awesome' (2019) is the canonical reference for product-level positioning against an incumbent. The procedure below uses the product-level frame — it is the version a solo founder can run alone in an afternoon.
What it is not
A list of competitor names. A feature comparison matrix. A 'market share' chart. A pricing table. Each of these is a tool that can support a competitive analysis, but none of them is the analysis itself. The analysis ends in a wedge and an experiment.
What it is
A one-page map that names the wedge, the alternatives, the assumptions behind each alternative's claim, and the next experiment. The map fits on a single screen and a stranger can read it in two minutes.
The procedure
The 90-minute procedure — eight blocks of 10–15 minutes
Run the blocks in order. Each block produces one input. Do not edit inputs from later blocks back into earlier ones; the procedure is one-directional.
- 1
Block 1 — Name the buyer in one sentence
10 minWrite the buyer sentence exactly the way it appears in the report: persona, industry, size band, geography. 'US accounting firms with 5–25 employees' is a buyer. 'Small businesses' is not. If you cannot write the sentence in one line, the analysis is not specific enough to run.
- 2
Block 2 — List the four kinds of competitor
15 minDirect: a product that does the same thing for the same buyer. Adjacent: a product that does something different for the same buyer. Status quo: the buyer does nothing today. Substitute: the buyer solves it with a spreadsheet, an agency, a contractor, or a manual process. Most categories have all four; the analysis is incomplete if any of the four is missing.
- 3
Block 3 — Write each competitor's one-sentence claim
10 minFor each competitor on the list, write the one sentence they use to describe themselves to the buyer. The sentence must be one the buyer would recognize, not one the competitor would use internally. Source: the competitor's homepage, the headline of their last press release, and the first sentence of their 'About' page.
- 4
Block 4 — Write the assumption behind each claim
15 minFor each one-sentence claim, write the single assumption the claim rests on. 'For [persona] who [need], unlike [alternative]' makes the assumption explicit: the alternative is supposedly worse on the dimension the claim highlights. The assumption is what a new product would have to falsify to win.
- 5
Block 5 — Write the counter-evidence for each assumption
15 minFor each assumption, write the strongest counter-evidence you have seen publicly. A G2 or Capterra review that contradicts the claim. A Reddit thread where users complain about the dimension the claim highlights. A competitor's own marketing that admits the gap. The counter-evidence is what the wedge would exploit.
- 6
Block 6 — Write the wedge
15 minThe wedge is the single assumption that, if it broke, would let a new product win. The wedge is not a feature list and not a positioning sentence; it is the falsifiable claim the new product will make to the buyer. Write it in one sentence, in the buyer's words, with the counter-evidence from block 5 built in.
- 7
Block 7 — Identify the cheapest experiment that tests the wedge
10 minWrite down the experiment that tests the wedge. The experiment must be cheaper than building the product and must produce a result that a stranger, given the evidence, would consider informative. A landing page with a paid-traffic variant, a concierge MVP, a smoke test, or a five-call customer-research round are all acceptable.
- 8
Block 8 — Name the next decision the experiment would inform
10 minWrite the decision the experiment would inform: continue / pivot / kill on the wedge, narrow the persona, change the channel, change the price point. If the experiment does not produce a named decision, it is not yet an experiment — it is more research.
The output
The one-page competitive map
The eight blocks produce this table. The table fits on a single screen; a stranger can read it in two minutes; it ends in a named experiment and a named decision.
| Competitor | Type | Weakness | Source |
|---|---|---|---|
| Incumbent A (direct) | Direct — same product, same buyer | Their claim rests on X; counter-evidence is Y | Their 'About' page, last press release, G2 reviews |
| Incumbent B (adjacent) | Adjacent — different product, same buyer | Their claim rests on X; counter-evidence is Y | Their 'About' page, last press release, G2 reviews |
| Status quo (no tool) | Status quo — the buyer does nothing today | Their claim is 'we don't need this'; counter-evidence is the workaround they already run | Customer interviews, Reddit threads, public complaints |
| Substitute (manual process) | Substitute — the buyer solves it manually | Their claim is 'spreadsheet is enough'; counter-evidence is the time or money the manual process costs | Customer interviews, time-and-motion studies, contractor rates |
| Wedge (your product) | Wedge — the assumption a new product would falsify | The wedge is the single assumption that, if it broke, would let a new product win | The output of blocks 5 and 6 above |
Each row ends in a source. The sources are public (homepages, review sites, public interviews, public complaints), and the counter-evidence is something a stranger could check. If a row has no public source, the analysis has skipped a step.
Common mistakes
Five mistakes that turn a competitive analysis into a list of names
Each of these is a way to spend 90 minutes and produce a deliverable that does not change the next decision. None of them are reasons to ship the analysis; they are reasons to redo the procedure.
Confusing 'no competitor' with 'no substitute'
'There are no direct competitors in this niche' is almost always true for a new product, and almost always irrelevant — because the buyer is still solving the problem with a status quo or a substitute. The status quo is 'do nothing'; the substitute is 'do it manually with a spreadsheet, an agency, or a contractor'. Most categories have a substitute even when they do not have a direct competitor.
Copying the incumbent's positioning
'For [persona] who [need], unlike [alternative]' is a useful template. Copying the template is a mistake when the founder fills the bracket with the same words the incumbent uses. The wedge has to falsify an assumption the incumbent's claim rests on; a positioning sentence that repeats the incumbent's positioning does not falsify anything.
Listing competitors without naming the assumption
A competitive map that lists 'Incumbent A, Incumbent B, Incumbent C' with no per-row claim, no per-row assumption, no per-row counter-evidence is a competitor list, not an analysis. The list does not produce a wedge. The analysis must produce a wedge, and the wedge requires per-row claims and assumptions.
Treating the analysis as a one-time exercise
A competitive analysis is a snapshot. The map is correct on the day it is written and starts to drift the moment a new competitor launches, an incumbent raises a round, or a substitute emerges. The procedure should be re-run whenever any of those three events happens — usually every six to twelve months for an early-stage startup.
Stopping before the wedge is written
The procedure ends with a wedge and an experiment. An analysis that stops at block 5 ('we have listed the assumptions and the counter-evidence') has not finished. The wedge is the output; the experiment is the deliverable; without them the procedure has produced research, not a decision.
Connecting to Yibud
How the Yibud tools help
The 90-minute procedure on this page produces a wedge and an experiment. The Yibud tools carry the wedge into the same report so the competitive map and the score move together.
Run the idea evaluator
The Startup MRI evaluator includes a competition dimension in its eight-dimension scoring. The dimension is informed by the one-page competitive map: which competitors exist, what each one claims, and what each claim rests on. The map makes the score auditable.
Open the evaluator →Use the positioning page
The positioning page documents the four-wedge framework and the '[for–unlike]' template for writing the positioning sentence. The wedge from block 6 above is the input; the positioning sentence is the output. The two pages together close the loop from analysis to copy.
Open the positioning page →Connect to the risk assessment
The risk assessment categorizes the per-dimension risks the report surfaces. The competitive map is one input: which risks come from the direct competitors, which from the adjacent, which from the status quo, and which from the substitute. The categorization makes the risk list actionable.
Open the risk assessment →FAQ
Frequently asked questions
What is competitive analysis in simple terms?
Competitive analysis is the procedure that turns 'we have competitors' into 'this is the wedge a new product would open, and this is the experiment that tests it.' The output is a one-page map with a wedge and a next experiment, not a list of competitor names. Porter's 'How Competitive Forces Shape Strategy' (HBR, 1979; updated 2008) is the canonical reference for industry-level analysis; April Dunford's 'Obviously Awesome' (2019) is the canonical reference for product-level positioning.
What are the four kinds of competitor?
Direct (same product, same buyer), adjacent (different product, same buyer), status quo (the buyer does nothing today), and substitute (the buyer solves it manually with a spreadsheet, an agency, or a contractor). Most categories have all four. An analysis that lists only direct competitors is incomplete.
How long does a competitive analysis take?
The 90-minute procedure on this page runs eight blocks of 10–15 minutes each. It is intentionally timeboxed; the timebox is what prevents the analysis from growing into a research project. A useful competitive map fits on a single screen; an analysis that takes longer than a few hours to produce is usually producing a research deliverable instead.
What is a competitive matrix?
A competitive matrix is a table that compares competitors on a small number of features or attributes. The matrix is a useful input but not the analysis itself. The procedure on this page uses a different output: a one-page map that names the wedge, the alternatives, each alternative's one-sentence claim, the assumption behind the claim, and the counter-evidence. The map ends in a named experiment.
How do I find competitors if I am early?
Start with the buyer, not with the category. Search G2, Capterra, Product Hunt, and Reddit for what the buyer would search for when they are trying to solve the problem today. Search for 'spreadsheet for X', 'agency for X', 'how to do X without a tool'. The first three pages of results usually surface the direct, adjacent, status quo, and substitute in one pass.
What if there are no direct competitors?
'No direct competitors' is almost always true for a new product and almost always irrelevant — because the buyer is still solving the problem with a status quo or a substitute. The status quo is 'do nothing'; the substitute is 'do it manually'. Most categories have a substitute even when they do not have a direct competitor, and the substitute is often the harder competitor to displace.
What is a wedge?
A wedge is the single assumption that, if it broke, would let a new product win. The wedge is not a feature list and not a positioning sentence; it is the falsifiable claim the new product will make to the buyer. The wedge is the hypothesis the next experiment tests, and the experiment is the deliverable that ends the procedure.
How does this connect to positioning?
The 90-minute procedure produces a wedge (block 6) and an experiment (block 7). The positioning page documents the four-wedge framework and the '[for–unlike]' template for writing the positioning sentence. The wedge from the analysis is the input; the positioning sentence is the output. The two pages together close the loop from analysis to copy.
Run the idea evaluator
Generate a Startup MRI report. The competition dimension is informed by the one-page competitive map produced by the 90-minute procedure; the other seven dimensions — market, distribution, monetization, build, founder fit, opportunity, and validation — are independent and feed into the same report.
Analyze my idea →