agentleFS
buildwithreefbuildwithreef/claude-skills/skills/website-conversion-audit/SKILL.md

website-conversion-audit

Skill0 starsChanged 4 months agoMITView on GitHub

Runs a full conversion-rate (CRO) audit on a small-business website — homepage clarity, calls-to-action, forms, phone paths, trust, speed, mobile, and analytics — and returns fixes ranked by impact and effort. Use when the user says their website gets traffic but no leads, asks for a conversion audit or CRO review, or says "my site doesn't generate calls" or "audit my website".

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 4 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Every stumble is a finding.

    line 26
  2. Very specific:

    Every finding gets Impact (H/M/L: how many conversions it touches) and Effort (H/M/L: hours vs.

    line 41
  3. Distinctive:

    Don't recommend a redesign when the phone number being unclickable is the actual problem.

    line 70
  4. Distinctive:

    Don't report Lighthouse scores as findings.

    line 71
  5. Distinctive:

    Don't audit against big-brand patterns (mega-menus, chatbots, popups) — a small business wins on clarity and speed-to-human.

    line 72

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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

Website Conversion Audit

A landing-page critique judges one page; this audits the whole machine — every path a visitor can take from arrival to becoming a lead, and every place those paths leak.

Step 1 — Establish the facts

  • How does a customer convert today? (call, form, booking tool, walk-in) — rank them by revenue
  • Roughly how much traffic, from where? (if they don't know, that's finding #1)
  • What pages exist? (map the nav: home, services, about, contact, anything else)
  • Is anything being measured? (GA4, call tracking, form notifications that actually arrive)
  • Who's the customer and what do they want within 10 seconds of landing?

Step 2 — Walk the paths

Play the customer three times, on a phone:

  1. The caller: land on the homepage → find the phone number → is it clickable, in the header, on every page? How many seconds and taps?
  2. The form-filler: find the form → count fields → submit a test → does a confirmation appear, does the notification reach a human, does anything say when they'll hear back?
  3. The researcher: pick a service → is there a real page for it (not a bullet on a list)? Does that page answer cost, process, and "why you," and end with a CTA?

Every stumble is a finding. The full area-by-area checklist (30+ items across homepage, nav, service pages, forms, trust, speed, mobile, analytics) is in references/checklist.md — load it and work through it.

Step 3 — The audit areas, weighted by what usually leaks most

  1. Above-the-fold clarity (home + top pages): what/for whom/do-what-next, five-second rule.
  2. Phone path: clickable header number sitewide. For local businesses this is routinely half of all conversions.
  3. Forms: fields ≤4 unless each extra one earns its place; instant confirmation; notification tested end-to-end (broken form notifications are shockingly common and invisible).
  4. Service pages: one per money service, each self-sufficient (offer, proof, price signal, CTA). "Services" list pages don't convert; they route.
  5. Trust: reviews on-site (not just "find us on Google"), real photos, license numbers, guarantees, years/counts.
  6. CTA consistency: same primary action everywhere; footer of every page ends with one.
  7. Speed & mobile: PageSpeed Insights mobile run — report only the 2–3 fixes worth doing (usually image compression and unused scripts), not the whole waterfall.
  8. Analytics: GA4 present, call clicks and form submits tracked as events. If nothing's measured, every future decision is a guess.

Step 4 — Prioritize with the impact/effort matrix

Every finding gets Impact (H/M/L: how many conversions it touches) and Effort (H/M/L: hours vs. days vs. rebuild).

  • H impact / L-M effort → Do now (this list should be 3–7 items)
  • H/H → Schedule (usually service-page buildout or redesign)
  • L impact → Skip — say so explicitly; permission to ignore small stuff is part of the audit's value

Output format

# Conversion Audit — [site]

**Verdict:** [2–3 sentences: where the machine leaks most]

## Do now (this week)
| Fix | Where | Impact | Effort | How |
|---|---|---|---|---|

## Schedule (this month/quarter)
[...]

## Skip (looked at it, not worth it)
[...]

## What to measure after
[the 2–3 numbers that will show whether the fixes worked]

What not to do

  • Don't recommend a redesign when the phone number being unclickable is the actual problem. Exhaust cheap fixes first.
  • Don't report Lighthouse scores as findings; report the two changes that move them.
  • Don't audit against big-brand patterns (mega-menus, chatbots, popups) — a small business wins on clarity and speed-to-human.

Built by Reef — AI systems for small business → https://buildwithreef.com

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.