agentleFS
Natan-MohartNatan-Mohart/24-strategy-skills-for-claude/skills/customer-segmentation/SKILL.md

customer-segmentation

Skill41 starsChanged 3 months agoMITView on GitHub

Builds needs-based, mutually exclusive customer segments using Jobs-to-be-Done rather than demographic personas, then scores each segment on attractiveness and your actual right-to-win. Use whenever the user wants to segment customers, define an ICP, asks who to target, or has "personas" that describe demographics/psychographics but don't actually predict different buying behavior.

4 other projects use a version of this.

Adopted by

Most recently taken up by

The file

Files
SKILL.md

Customer Segmentation

When to use

Use when existing segmentation is demographic or psychographic ("millennial urban professionals") and doesn't actually predict different purchase behavior, price sensitivity, or channel preference — a sign the segments aren't doing real analytical work. Also use when deciding where to focus limited go-to-market resources across multiple candidate customer groups.

What it does

Segments customers by the job they're hiring the product/service to do — the underlying need and the context that makes it urgent — rather than by who they demographically are. Produces segments that are mutually exclusive and collectively exhaustive (MECE), then scores each on market attractiveness and your specific right-to-win, so the output is a prioritized target list, not just a taxonomy.

Method

  1. Interview or infer the underlying job: for each customer group, identify the functional job (what task are they accomplishing), the emotional job (how do they want to feel), and the social job (how do they want to be perceived) — Jobs-to-be-Done, not attribute lists.
  2. Cluster by job similarity, not by demographic proxy. Two customers of different ages/industries with the same job and trigger event belong in the same segment; two customers of the same demographic with different jobs do not.
  3. Check MECE: can any real customer plausibly belong to two segments as defined? If so, the segment boundaries are drawn on the wrong variable — usually a demographic proxy leaking back in.
  4. Size each segment using the same triangulation discipline as market-mapping (independent top-down and bottom-up estimates).
  5. Score attractiveness per segment: growth rate, willingness to pay, switching cost/lock-in potential, and lifetime value — not just headcount.
  6. Score right-to-win per segment separately from attractiveness: does your current product, channel, brand, and cost structure actually fit this segment's job, or would winning it require capabilities you don't have yet? A segment can be highly attractive and still be a poor fit.
  7. Plot attractiveness × right-to-win on a 2x2 to force the prioritization call — resist the temptation to declare every attractive segment a priority.
  8. Write one paragraph per priority segment naming the job, the trigger event that makes them buy now, and the objection that stops them.

Inputs

  • Customer interview or survey data, ideally including purchase trigger and objections
  • Usage/behavioral data segmented by job proxy where available (use case, workflow, trigger event)
  • Current product/channel/cost-structure capabilities, to assess right-to-win honestly

Output format

MECE segment set defined by job-to-be-done; size and attractiveness score per segment; right-to-win score per segment; attractiveness × right-to-win 2x2; one-paragraph profile per priority segment (job, trigger, objection).

Example

A project-management tool initially segments by company size (SMB/mid-market/enterprise). Interviews reveal the real driver is the job: "coordinate a distributed team with no single owner of the calendar" shows up across company sizes and predicts purchase far better than headcount does. Re-segmenting by job surfaces a previously invisible high-attractiveness, high-right-to-win segment — distributed agencies — that the size-based segmentation had split across three different buckets.

Common pitfalls

  • Segmenting by attributes that are easy to measure (age, industry, size) instead of the job that actually predicts behavior.
  • Scoring only attractiveness and skipping right-to-win, leading to a target list the company can't actually win.
  • Producing overlapping segments that aren't MECE, which makes messaging and channel decisions ambiguous.

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.