Multi-Store Analysis
by
All your Shopify stores and markets in one analysis: which store drives growth, where they diverge on basket, new customers and conversion, and totals that reconcile, without building a single dimension.
Shopify reports one store at a time. The moment you run a DE and a UK store, or five markets, the comparison lives in a spreadsheet someone rebuilds every month. This skill does it in one run. It finds the dimension that already separates your stores in Polar, maps each store to a label you recognise, and puts them side by side on the same definitions: net sales, orders, AOV, new customers and their share of revenue, over the last 28 days, the previous period, last year and a 12-month trend. Metrics that can't be split by store, like GA4 sessions or ad spend, are shown once at total level instead of being guessed. And before it answers, it checks that the stores add up to the total and says exactly how far off they are when they don't.
What It Does
- Finds the dimension that already separates your stores or markets, and never creates one.
- Compares stores on the same definitions: net sales, orders, AOV, new customers, new-customer share.
- Reads three periods at once: last 28 days, previous 28 days and the same days last year, weekday-aligned.
- Shows metrics that can't split by store once, at total level, instead of allocating them.
- Checks that the stores add up to the total, and prints the gap in euros when they don't.
- Follows your exact table spec when you give one: columns, order, no commentary, no totals.
Use Cases
- Spot the store that's dragging the year before the monthly review.
- Compare basket size and new-customer mix across markets in one table.
- Hand finance a monthly store-by-store table with figures that reconcile.
- Check whether a new market grows on new customers or on repeat buyers.
Example Session
/polar-multi-store Compare my stores.
Multi-store analysis · DE vs UK · last 28 days
Stores from the existing Shopify Site dimension · complete days · last year aligned by weekday · read-only
• DE is still the bigger store (€96.4k vs €71.2k), but it's down 18% on last year while UK is down 4%.
• New customers bring 31% of UK revenue against 15% in DE.
• UK's basket is €7 higher: check what DE is missing in bundles or pricing.
- KPI | DE | vs LY | UK | vs LY | Total
- Net Sales | €96.4k | −18% | €71.2k | −4% | €167.9k
- Orders | 820 | −9% | 572 | +6% | 1,393
- AOV | €117.6 | −10% | €124.5 | −9% | €120.5
- New customers | 142 | −12% | 181 | +8% | 323
- New-cust. share | 15% | n/a | 31% | n/a | 22%
Where they diverge: UK is winning new customers (+8% on last year) while DE loses them (−12%). DE's basket keeps shrinking even as its orders pick up on the previous 28 days.
QA: DE + UK = €167.6k vs total €167.9k, €0.3k of orders assigned to no store (0.2%) · new customers reconcile exactly · GA4 sessions and ad spend shown at total level only.
Install It
- Copy the command.
One line in your terminal. Free, no account needed. - Run it.
npx skills add pulls the skill from Polar's GitHub and installs it for the agents you pick: Claude Code, Cursor, Codex and more. - Connect Polar and ask.
Connect the Polar connector, then ask to compare your stores. The skill does the rest.
skills folder
Copy
npx skills add \
https://github.com/polar-analytics/claude-skills \
--skill polar-multi-storeFrequently asked questions
Why can't I do this in Shopify?
Shopify reports each store on its own. Comparing them means exporting and rebuilding a sheet, with definitions that drift between stores. Polar already holds every store in one model, so the skill reads them side by side.
Does it create a dimension or change my setup?
No. It reuses the dimension that already separates your stores, such as Shopify Site or country, and only reads. If nothing separates them, it asks one question instead of building one.
What about metrics that aren't per store, like GA4 sessions?
They're shown once, at total level, with the reason. The skill never splits a metric across stores by a ratio it made up.
Why don't my stores add up to the total exactly?
Usually orders that belong to no store, or customers counted in two. The skill computes the gap, shows it in euros and explains its direction instead of hiding it.
Can it follow my own table format?
Yes. Give it the columns, order and period, ask for no commentary or no total row, and it prints exactly that table.
Does it handle stores in different currencies?
Yes. It says whether Polar converted them to your reporting currency and puts the currency in each column header.
A store comparison is worth what your definitions are worth.
Polar models every store on the same metric definitions and currency, so DE, UK and US are compared on one net sales, not three exports.
<script type="application/ld+json">{"@context": "https://schema.org", "@type": "FAQPage", "mainEntity": [{"@type": "Question", "name": "Why can't I do this in Shopify?", "acceptedAnswer": {"@type": "Answer", "text": "Shopify reports each store on its own. Comparing them means exporting and rebuilding a sheet, with definitions that drift between stores. Polar already holds every store in one model, so the skill reads them side by side."}}, {"@type": "Question", "name": "Does it create a dimension or change my setup?", "acceptedAnswer": {"@type": "Answer", "text": "No. It reuses the dimension that already separates your stores, such as Shopify Site or country, and only reads. If nothing separates them, it asks one question instead of building one."}}, {"@type": "Question", "name": "What about metrics that aren't per store, like GA4 sessions?", "acceptedAnswer": {"@type": "Answer", "text": "They're shown once, at total level, with the reason. The skill never splits a metric across stores by a ratio it made up."}}, {"@type": "Question", "name": "Why don't my stores add up to the total exactly?", "acceptedAnswer": {"@type": "Answer", "text": "Usually orders that belong to no store, or customers counted in two. The skill computes the gap, shows it in euros and explains its direction instead of hiding it."}}, {"@type": "Question", "name": "Can it follow my own table format?", "acceptedAnswer": {"@type": "Answer", "text": "Yes. Give it the columns, order and period, ask for no commentary or no total row, and it prints exactly that table."}}, {"@type": "Question", "name": "Does it handle stores in different currencies?", "acceptedAnswer": {"@type": "Answer", "text": "Yes. It says whether Polar converted them to your reporting currency and puts the currency in each column header."}}]}</script>