claude skills / Claude Skills Pro / store-growth-audit / en français

Store growth audit.

Sales are down: Claude finds which lever moved, why, and what to do this month.

See what it produces
  • Shopify
  • WooCommerce
  • PrestaShop
  • Magento
  • Google Analytics 4
  • Google Sheets
3factors isolated: sessions, conversion, average order value
10 maxranked actions, including 3 to launch this month
100 %of figures tied to a file or a visible formula

orders of magnitude, depending on your volume

the problem

What gets in the way today.

  • 01Revenue is falling while traffic is rising, and nobody can explain the gap.
  • 02Dashboards are full of numbers, but none of them tells you where to start.
  • 03Every meeting produces 25 ideas, with no owner, no date and no metric.

input, output

You give this. You get this.

What you give Claude: store exports for two periods of the same length, and the context

Store: Harrow Lane Home (Shopify, EUR)
Period: 2026-09-01 to 2026-10-31 vs same period 2025
Files: sales_over_time.csv, conversion_over_time.csv, sessions_by_referrer.csv,
sales_by_product.csv, orders_export.csv
Context: newsletter paused since August, paid social budget raised in September,
best seller out of stock 19 days in October

$ claude "Audit my store with these exports: why did revenue drop, and what should we do first?"

Store growth audit, report excerptexample

Store growth audit: Harrow Lane Home

Period: 2026-09-01 to 2026-10-31 vs 2025-09-01 to 2025-10-31 | Shopify | EUR

SUMMARY

Net revenue fell by 7,254 EUR (minus 11.5%) while sessions rose 7.7%. The drop comes from conversion (2.10% to 1.68%), and most of that is a change in traffic mix: paid social sessions more than doubled at a low conversion rate while email sessions fell by more than half after the newsletter paused.

REVENUE DECOMPOSITION

Sessions: 51,900 vs 48,200 (+7.7%)

Conversion rate: 1.68% vs 2.10% (minus 20.0%)

AOV: 64.10 vs 62.40 (+2.7%)

Contribution (log method): sessions +4,397 EUR, conversion minus 13,248 EUR, AOV +1,598 EUR

FINDINGS

1. Fact: email orders fell from 231 to 124 (sessions_by_referrer)

2. Fact: best seller out of stock 19 days, its revenue minus 34% (sales_by_product)

3. Inference: most of the conversion drop is mix, not site quality

4. Hypothesis: the new shipping app slows mobile checkout. Test: compare 2 weeks before and after install

ACTIONS

1. Restart the newsletter, 1 send per week. Moves: email orders. Score 100. Marketing, 2026-11-25

2. Restock best seller, low-stock alert at 3 weeks of sales. Score 80. Operations, 2026-11-30

3. Keep only the 2 paid social ad sets with CR above 1%. Score 80. Marketing, 2026-11-22

NEXT REVIEW

2026-12-15. Check first: email orders, conversion by channel, mobile checkout completion.

the method

What Claude does, in order.

  1. step 1Check the data

    Claude lists each file, its period and its rows, reconciles the totals and removes test orders and outliers.

  2. step 2Break down revenue

    Sessions, conversion rate and average order value are compared across both periods, with each one's contribution to the gap.

  3. step 3Read channels, funnel, customers, products

    Revenue per session and per channel, the weak funnel step, share of returning customers, products viewed a lot but rarely bought.

  4. step 4Rank the actions

    Impact, confidence and effort are scored, then the top 10 actions get an owner, a date and a metric.

the film / 44 seconds, with sound

The pack film.

the pack film

In 8 seconds.

the guardrails

What the skill never does.

  • Store data only: a missing export is named, along with where to get it.
  • No invented industry benchmarks: the store is compared with its own history.
  • Every finding is labeled fact, inference or hypothesis, and every hypothesis comes with its test.
  • Low-volume segments are flagged, never interpreted.

Questions

Which exports do I need to provide?

Sales over time, sessions and funnel, sales by channel, sales by product and the orders export, for the period under review and a period of the same length. The skill explains where to find each one.

Does the skill compare my store with the industry average?

No, unless you provide a sourced and dated figure. The comparison is first with the same period last year, then with the previous period.

Do the actions promise a specific gain?

No. Each action states the metric it should move and in which direction, never a promised amount.

Can we use it for a monthly review?

Yes, that is what it is built for: monthly or quarterly reviews, pre-season planning or explaining a drop in sales.

the pack

In the pack: Claude Skills Pro.