Course resource
Competitor Analysis Template
A structure for understanding competitors without producing forty pages nobody reads.
The rule that governs this
AI will confidently invent competitor facts. Funding rounds, headcounts, pricing, customer numbers — all plausible, all fabricated. Use AI to structure the analysis and to interpret what you find. Get the facts from primary sources: their site, their pricing page, their job ads, their filings, their customers.
Every cell in the table below should be traceable to something you can point at.
Who to analyse
Three, not ten. Pick:
- The direct competitor — closest to what you do, for the same buyer
- The alternative — what customers use instead, which is often a spreadsheet, an agency, or doing nothing
- The one that worries you — different model, moving toward you
The second one is where most of the insight is and the one most analyses omit.
The table
| Us | Competitor A | Competitor B | The alternative | |
|---|---|---|---|---|
| Who they sell to | ||||
| Stated positioning | ||||
| Price and model | ||||
| What they are visibly good at | ||||
| What customers complain about | ||||
| How they acquire customers | ||||
| What they have that we cannot copy quickly | ||||
| What we have that they cannot | ||||
| Direction they are moving |
Where to actually look
| Source | Tells you |
|---|---|
| Pricing page | Model, tiers, who they optimise for, what is gated |
| Job ads | Where they are investing — the most reliable signal of direction |
| Changelog / release notes | Real pace and priorities |
| Review sites, sorted by lowest rating | Where they genuinely fail |
| Their customers' public posts | What people actually use it for |
| Support docs | Complexity, and which problems recur |
| Their homepage over time, via an archive | How their positioning has shifted |
Job ads and one-star reviews are the two highest-signal sources and the two most people skip.
Prompts
Structuring what you found
Here is what I gathered about [COMPETITOR], all from their own site
and public sources:
[PASTE RAW NOTES]
Organise this into: who they target, positioning, pricing model, apparent
strengths, apparent weaknesses, direction of travel.
Use only what I gave you. Where something is missing, write UNKNOWN.
Do not fill gaps from your training data.
That last instruction is the important one.
Reading the reviews
Here are [N] negative reviews of [COMPETITOR]:
[PASTE]
1. Group the complaints by underlying cause, not surface wording
2. Which are fixable and which are structural to their model?
3. Which of these do WE also have?
4. Which represents a genuine opening for us?
Point 3 is the one people avoid and the one that is useful.
Reading the job ads
Here are [COMPETITOR]'s current job openings:
[PASTE TITLES AND KEY REQUIREMENTS]
What does this tell me about where they are investing?
What are they building that they have not announced?
What does the absence of certain roles suggest?
The honest positioning check
Our claimed differentiator: [WHAT]
Competitor claims: [THEIRS]
Is ours actually a differentiator, or is it table stakes that we simply
state more loudly?
What could they copy in a month? In a year? Not at all?
If a customer compared us side by side today, what would they conclude?
The output
One page. Four sections:
Where we genuinely win. Specific, and defensible for more than a quarter.
Where we genuinely lose. Be honest. This section is the point of the exercise.
What is changing. Direction of travel matters more than the current snapshot.
What we will do about it. Maximum three actions, each with an owner.
If your analysis has no "where we lose" section, you have not done the analysis.
Keeping it current
A competitor analysis is a snapshot that starts decaying immediately.
- Quarterly refresh of pricing and positioning only — twenty minutes
- Alerts on their announcements
- Full re-analysis annually, or when something material changes
- Date every version
| Date | What changed | Action taken |
|---|---|---|
What not to do
- Do not obsess. Time spent watching competitors is time not spent with customers, who are a better source of what to build.
- Do not copy their roadmap. You will always be behind, and you do not know why they made those choices.
- Do not cite AI-generated numbers. Ever. A fabricated funding figure in a board deck is a bad afternoon.
- Do not analyse ten. Three, properly, beats ten superficially.