agentleFS
iangogenticiangogentic/venture-lab/.claude/skills/analyze-competition/SKILL.md

analyze-competition

Skill0 starsChanged 4 months agoMITView on GitHub

Analyse the competitive picture for each opportunity.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Start here, not with vendors.

    line 29
  2. Distinctive:

    If you are unsure whether a product exists, describe the category instead of inventing a brand.

    line 49
  3. Distinctive:

    Do not omit the strongest incumbent because acknowledging it weakens the opportunity.

    line 55

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Analyse the competition

For each opportunity below, work out what these people already use to get this job done — including everything that is not a product — and what it would cost them to stop.

The question

${question}

The opportunities

${opportunities}

One analysis per opportunity

Link it to the opportunity by its real id.

Substitutes are usually the real competition

Start here, not with vendors. What you actually compete with is what the ICP does today, and it is rarely a product:

  • a spreadsheet, kept by one person, with the logic in their head;
  • a script somebody wrote once that nobody else can run;
  • a Slack channel, an inbox rule, a recurring meeting;
  • a junior hire, a contractor, an agency;
  • an existing tool used sideways for a purpose it was not built for;
  • doing nothing — living with the problem.

Doing nothing wins more evaluations than any vendor, so assess it seriously: what does it actually cost this segment to keep enduring the pain, and is that cost enough to make them move? A substitute that is ugly but free, already paid for, already trusted and already in the workflow beats a better product that requires change. Say plainly when a substitute is good enough — that finding is worth more than a competitor list.

Named competitors

  • Name only products you are certain exist, and characterise each concretely: who they serve, what they are genuinely good at, where they leave this segment badly served.
  • If you are unsure whether a product exists, describe the category instead of inventing a brand. A fabricated competitor is as damaging as a fabricated source, and it distorts the decision in both directions.
  • Judge strengths and weaknesses from the user's point of view, in the terms the upstream artifacts use. Vendor marketing is not a strength, and "clunky UI" is not a weakness unless someone said the work is slower because of it.
  • Do not omit the strongest incumbent because acknowledging it weakens the opportunity. That omission is the most expensive error available at this stage.

Switching cost is usually the whole story

A better product loses to switching cost more often than to a better product. Be specific about what moving actually costs this segment, in their terms:

  • migrating history, and whether it can even be exported;
  • retraining people who are fluent in the current way;
  • reopening a security review, a procurement process or a signed contract;
  • integrations, reports and downstream habits built on the incumbent's shape;
  • the political cost to whoever chose the incumbent last year;
  • the risk of being the person who broke a working process.

A switching cost with no named mechanism is not a switching cost. If you conclude switching is cheap here, say why — that is a strong claim and the most valuable one on the page.

Moats and differentiation

  • Be sceptical about moats. Most early opportunities have none, and "no durable moat identified" is a real, useful answer. Data network effects, workflow lock-in and regulatory position are claims needing an argument, not labels to reach for.
  • Differentiation is something an incumbent could not trivially ship this quarter. "Better UX", "AI-powered", "modern stack" and "purpose-built" are not differentiation; they are what everyone claims. Genuinely different means different in what the product does for the workflow, not in how it feels.

Failure modes

  • A tidy table of vendors with interchangeable generic weaknesses.
  • "No direct competitors" — which nearly always means either you have not looked, or the segment does not buy in this category at all. Say which.
  • Grading rivals on the axes this opportunity happens to win on.
  • Listing substitutes without asking whether they are already good enough.
  • Treating a large incumbent's current inattention as permanent.
  • Any competitor, funding figure, customer count or price you did not read in the input.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.