Systems Lab

Agent skill

customer-language-bank

Read hundreds of reviews, support tickets, and call snippets in one pass and return the exact words and themes buyers use - then map that language against your current messaging so you can see where you're talking past them.

slowingSelf-containedInstructions only494 words

Filed under Positioning and messaging.

From pmalliance/product-marketing-skills · 6 skills · 63 · pushed 2026-07-10

What it does when it runs

Read hundreds of reviews, support tickets, and call snippets in one pass and return the exact words and themes buyers use - then map that language against your current messaging so you can see where you're talking past them. Use when refreshing messaging, writing copy that should sound like the market rather than the boardroom, building personas from real voice-of-customer, or when you want the phrases customers actually say instead of the ones the team assumes. Triggers: "voice of customer," "how do customers describe us," "mine our reviews," "what words do buyers use," "make the copy sound like our customers."

Read from the skill and the 0 files bundled beside it. A skill’s own description is written to be selected by an agent, so it describes the job and not the dependencies.

Keys and connectors you must supply
None found.
Hosts it reaches
No third-party host appears in the skill or its bundled files.
Tool permissions it declares
No allowed-tools in the frontmatter. It only issues instructions, so there is nothing to bound.
Actions present in the files
None. Instructions only.

Ask about customer-language-bank

Opens your assistant with this page's verified links already in the prompt.

Is this safe to install?ClaudeChatGPT
Adapt it to my stackClaudeChatGPT
What else do I need for it to workClaudeChatGPT
Rather ask a human? Talk to Cheetah
git clone --depth 1 --filter=blob:none --sparse https://github.com/pmalliance/product-marketing-skills.git /tmp/product-marketing-skills
git -C /tmp/product-marketing-skills sparse-checkout set "pmm-signal-pack/skills/customer-language-bank"
mkdir -p ~/.claude/skills/customer-language-bank
cp -R "/tmp/product-marketing-skills/pmm-signal-pack/skills/customer-language-bank/." ~/.claude/skills/customer-language-bank/

Picked up without a restart. A project skill of the same name is shadowed by your personal one. For one repository only, swap ~/.claude/skills for .claude/skills. Claude Code docs ↗

Or take the whole library

This repo ships a .claude-plugin manifest, so Claude Code can install all 6 skills at once. Plugin skills are invoked as /<plugin>:<skill>, so they never collide with your own.

/plugin marketplace add pmalliance/product-marketing-skills
/plugin

The folder is the same in every client that implements the format — 46 of them — so if yours is not above, only the destination changes.

Reproduced in full from pmalliance/product-marketing-skills/blob/2ed9bdeacfa276a05271f38c7a97c066c36d1a19/pmm-signal-pack/skills/customer-language-bank/SKILL.md, which is licensed MIT (repository). 494 words, 7 headings.

Customer Language Bank

What this does

Reads your voice-of-customer material in bulk and returns the language buyers actually use - the words for their problem, the phrases they use for the outcome they want, the terms they use for you and for alternatives. It groups this into themes ranked by how often each shows up, then holds it up against your current messaging to show where your words and their words have parted ways.

Copy written in the team's internal language quietly loses buyers who describe their world differently. This gives you their vocabulary, weighted by frequency, so the messaging can meet them where they are.

Why this is a Fable-recommended job

The signal is in the volume - a phrase matters because many customers use it, and you only see that by reading a lot of them together. Claude Fable 5 can take hundreds of reviews, tickets, and snippets in a single run and rank the recurring language, which a sample of a dozen would miss.

If Fable is not available or the budget is spent: run in batches of a few hundred items, produce a themed phrase list per batch, then merge and re-rank the lists in a final pass.

Evidence to gather first

  • Product reviews (G2, Capterra, app stores, Trustpilot)
  • Support tickets and chat logs
  • Sales and onboarding call snippets where customers describe their problem or results
  • Community posts, survey free-text, churn and cancellation reasons
  • Social mentions, if available

Read .agents/product-marketing-context.md first if it exists, so the language can be checked against your positioning and personas.

Process

  1. Confirm the sources and roughly how many items are included. The report will state the volume.
  2. Read all of it. Pull the recurring language for: the problem, the trigger to look for a solution, the desired outcome, how they describe your product, and how they describe alternatives.
  3. Cluster into themes and count frequency. Keep the raw phrasings as examples.
  4. Compare the top themes against your current messaging. Note where you use a different word than the market does for the same thing.
  5. Produce the language bank.

Output format

  • Summary - items analyzed, sources, the dominant themes.
  • Language table - theme | what buyers say (real phrases with source) | how often it appears | the word your messaging currently uses | gap.
  • Swap list - specific places your copy could adopt the buyer's word.
  • Persona signals - differences in language by segment or role, if the material supports splitting them out.

The evidence standard

Every phrase in the bank must be a real quote with its source noted (review, ticket, call), not a paraphrase or an invented "customers say" line. Frequency counts must come from the material; if you can only estimate, label it an estimate. Do not clean up buyer language into marketing language - the unpolished wording is the whole point.

Other skills for the same job

Different authors, same problem. Matched on the words in the skill name, across every library in the catalogue except this one.

Need help setting it up?

This page tells you what customer-language-bank does and what it needs. Cheetah builds the agent setup it runs inside: data, CRM, sequencing and the guardrails.

Book a call →

The directory stays free. There is nothing gated behind this.