Manage 100+ SKUs without chaos
Once your catalog passes about a hundred SKUs, memory stops working and small inconsistencies pile into real problems: two products tagged three different ways, variants that don't match, reports you can't trust.
The fix is a naming system, clean product data, and a bulk-edit habit, so AI can audit and reformat the whole catalog in one pass instead of you clicking row by row. Do variants-inventory-basics first; this note assumes your variants are already set up.
Fastest path: one prompt, end to end
🤖 AI prompt — paste into ChatGPT / Claude
You are a catalog operations manager. Use MY data only.
Attach my product export (CSV from Shopify/Woo/etc.) with columns: title, SKU, type, tags, variants, price.
My category structure: [e.g. Coffee > Beans / Gear / Subscriptions]
Do this:
1. Propose a SKU code convention that encodes category, product, and variant (e.g. COF-ETH-250 = Coffee/Ethiopia/250g). Show the pattern, then re-code 5 of my real products as examples.
2. Audit my data hygiene: flag missing product types, empty tags, inconsistent casing, duplicate titles, orphan variants. Return a table of what's wrong and the exact fix.
3. Standardize my tags into a controlled list (merge synonyms like 'gift'/'gifts'), and name the 8-12 tags that actually drive storefront filtering.
4. Give me a monthly catalog audit checklist.
Do not invent product data or numbers. If a field is missing from my export, list it and ask, do not fill it in. If you can't open my file, tell me and I'll paste the rows.
Output: SKU convention + hygiene audit table + controlled tag list + audit checklist.
Or do it in 4 steps
- Set a SKU convention before you add more. A readable code like COF-ETH-250 (category, product, variant) lets you sort, search, and spot gaps at a glance. Random or auto-generated SKUs cost you every time you pick, count, or reconcile. Decide the pattern once and apply it going forward; batch-fix the back catalog later.
- Clean the data that powers filtering and reports. Product type, tags, and metafields are what your storefront filters and your exports group on. One controlled tag vocabulary (not 'gift', 'gifts', and 'Gifting' as three separate tags) makes collections and reports actually work. Fix casing and fill required fields so nothing drops out of a filter silently.
- Edit in bulk, not one product at a time. Use your platform's bulk editor (Shopify's bulk editor, a Matrixify/CSV round-trip, or Woo's Quick Edit) to change price, tag, or type across dozens of SKUs at once. Export, let AI reformat the sheet against your convention, re-import. One pass beats a hundred clicks and keeps the change consistent.
- Audit on a cadence, and know when a spreadsheet breaks. Once a month, scan for orphan variants, untagged products, and dead SKUs. When exports get slow, edits collide between people, or you sell across more than one channel, that's the signal to move from spreadsheet ops to a light PIM (product information management) so one source of truth feeds every channel.
Worked example (labeled): a store with 140 SKUs had 'sale', 'Sale', and 'on-sale' as three separate tags, so its Sale collection missed a third of the discounted items.
Exporting the catalog, having AI merge the tags to one value and re-code SKUs to CAT-PROD-VAR, then re-importing, fixed 140 rows in one round-trip and the collection finally showed everything. The fix wasn't more clicking; it was a convention plus a bulk edit.
A hundred SKUs stay manageable when the naming is systematic and the data is clean; audit monthly, and move to a PIM the day the spreadsheet starts fighting you.