Data management

Compatibility data isn’t a “set and forget” thing. Products change, listings get deleted, manufacturers release new models, and your catalog evolves over the months and years. ModelMatch gives you four tools for keeping your data clean as that happens: the Records page for managing what’s already in, auto-sync for keeping records aligned with Shopify changes as they happen, orphan detection for catching records that point at products you’ve deleted, and the Activity log for seeing what’s happened recently. Most of this runs automatically once configured. Your job is to know what each tool does so you can troubleshoot and tune as needed.

The Records page

What it does

The Records page is your main interface for managing compatibility data. Every record in your database appears here. From this page you can search, filter, edit, bulk-act, and run health checks — it’s where most of the day-to-day operating happens after your initial import. If you’re going to bookmark one ModelMatch admin page, this is the one.

[Screenshot: Records page showing the health strip at the top, search/filter bar, and a populated records table.]

The health strip

Across the top of the Records page is a row of five stat cards that summarize the health of your data at a glance:

• Total records — Count of every compatibility record in your shop’s database. This is your raw catalog size.
• Mapped — Records that are linked to a Shopify product that still exists. These give shoppers the best experience: product images appear in finder results, add-to-cart works directly from the finder, and the product page widget can rely on the link being live.
• Unmapped — Records with no Shopify product link at all. Compatibility lookups still work — a shopper can verify “yes, this fits your equipment” — but they can’t go directly to a product from the finder. Some Unmapped is normal and fine; a high count is a signal to add product mappings where you can.
• Orphaned — Records that were linked to a Shopify product, but that product has since been deleted. The card is highlighted in amber when the count is above zero. Always worth investigating; the orphan section below covers what to do with these.
• Last sync — When the last orphan check ran. Shows “Just now,” “X minutes ago,” “X hours ago,” or the actual date for older runs. Tells you whether the orphan count above is fresh or stale.

Search, filter, and edit

For day-to-day operations:

• The search bar matches against brand, model, alias, part SKU, and part name. Type any of those and you’ll find the matching records.
• Filters narrow the table by category, brand, or status (Mapped / Unmapped / Orphaned). Combine search and filter freely.
• Click any record to open an edit drawer with all fields visible. Make a change, save, and the table updates without a full page refresh.
• Click “Remove” on a row to delete a single record. A confirmation prompt prevents accidental deletes.

Pagination

The Records page shows 50 rows per page with numbered pagination at the bottom. The page count adapts to your active filter — searching for a specific brand might narrow you to one page even if your full database has 50,000 records. This means the table stays fast no matter how big your catalog gets.

Bulk operations

Each row has a checkbox; the header has a “select all” checkbox. Once you’ve selected one or more records, the bulk action bar appears with two options:

• Bulk edit — Apply the same change across selected records. Useful for things like renaming a category that drifted (“Replacement Filter” → “Filter”) across hundreds of records at once.
• Bulk delete — Permanently remove selected records. A confirmation modal shows the count before the action runs.

Bulk actions respect your active filters. If you’ve filtered to a specific brand and click “select all,” you’re selecting all matching records for that filter — not all records in your database. This is what you want and probably matches your mental model, but it’s worth knowing so you don’t accidentally over-select.

Run check now

The “Check now” button in the Records page header triggers an orphan detection sweep on demand. Useful right after you’ve deleted Shopify products in bulk, or any time you want to confirm that the orphan count and “Last sync” timestamp on the health strip are current. The check runs synchronously for typical catalogs and takes under 30 seconds.

Connect Data

Three import paths

Connect Data is your import hub. Three tabs, each for a different starting point. (For a deeper walkthrough of each, see Your first records.)

• CSV Upload — The most common starting point. Upload a spreadsheet of compatibility data. Three modes: Update existing + add new (the default and recommended for most uses), Add new records only (skips matching rows, only adds new), and Replace all records (destructive — wipes the database first; requires typing “REPLACE” to confirm).
• Import from Shopify — For merchants without prepared compatibility data. AI mapping reads your product catalog and suggests records based on web research. Available on every plan, with monthly usage caps that scale with the tier.
• Manual Entry — A single-record entry form. Best for one-offs, testing how the storefront handles a specific scenario, or filling individual gaps in your data.

Auto-sync settings

Below the three import tabs is the auto-sync toggle. When enabled, ModelMatch listens for changes to your Shopify products and updates records automatically as those changes happen. Available on Growth and Professional plans.

What auto-sync does:

• Product created — If a new Shopify product matches an existing record by SKU or product handle, the record’s product mapping fields auto-populate. New products that don’t match any existing record do nothing on their own — you still have to add the compatibility data manually or via AI mapping.
• Product updated — If a product’s variant ID, handle, or URL changes, every record linked to that product gets the updated values automatically. This means renaming a product or restructuring your collection URLs doesn’t break your records.
• Product deleted — Records linked to the deleted product are handled per the deletion behavior described below.

What happens when a product is deleted

The behavior on Shopify product deletion is determined by your auto-sync setting — there’s no separate toggle for it.

When auto-sync is on AND a Shopify product gets deleted:

• ModelMatch hard-deletes any compatibility records that were linked to that product.
• This is appropriate when products are deleted because they’re truly retired and the compatibility data is no longer relevant.

When auto-sync is off AND a product gets deleted:

• Records linked to the deleted product are flagged as Orphaned instead of being deleted.
• You can review them in the Orphans tab in Insights and decide whether to unmap, delete, or leave them.
• This is the safer default if you’re unsure about the lifecycle of your products, or if you sometimes delete products in error and need to be able to recover the compatibility data.

[Screenshot: Connect Data page with the three tabs visible and the auto-sync toggle.]

Orphan detection

What an orphan is

An orphaned record is one that points at a Shopify product that no longer exists. The compatibility data — brand, model, category, alias — is still there, but the product link is broken. Shoppers won’t see orphaned records on the storefront (ModelMatch hides them from finder results so the experience never breaks), but the data lingers in your database and continues to count against your plan limits until you act on it.

Orphans typically happen when:

• You delete a Shopify product, intentionally or accidentally.
• You have auto-sync off and don’t manually clean up after deletions.
• A product gets deleted in bulk by a third-party app — an inventory sync, a vendor changeover, a migration tool — that you didn’t anticipate.

Automatic detection

ModelMatch checks for orphans on a schedule that depends on your plan:

• Starter — No automatic checks. Run “Check now” from the Records or Insights page when you want to detect orphans manually.
• Growth — Daily automated check. Detects orphans within 24 hours of products being deleted.
• Professional — Hourly automated check. Detects orphans within 1 hour.

The automatic check runs against Shopify’s Admin API to verify each linked product still exists. Records pointing at deleted products are flagged. Records pointing at existing products are confirmed healthy and the “Last sync” timestamp on the health strip updates.

Manual checks

Anyone on any plan can trigger an orphan check on demand:

1. Go to the Records page or Insights > Orphans tab.
2. Click Check now.
3. Wait for the sweep to complete — usually under 30 seconds for typical catalogs.
4. A toast notification shows the result, and the orphan count on the health strip updates.

For Growth and Professional, manual checks are useful right after big product changes — bulk deletions, catalog migrations, vendor changeovers — when you don’t want to wait for the next scheduled sweep. For Starter, manual checks are the only way to detect orphans, so building a manual check into your workflow after any product changes really matters.

Managing orphans

Orphaned records show up in two places:

1. Records page — The Status column shows “Orphaned” with an amber warning icon. You can act on individual orphans here using the row’s Edit or Remove actions.
2. Insights > Orphans tab — A dedicated view showing all orphans together with bulk action options. This is the right place when you have more than a handful to deal with.

In the Orphans tab, two actions are available per record (and as bulk actions across multiple selected records):

• Unmap (recommended default) — Clears the productVariantId, productHandle, productUrl, and imageUrl fields. The compatibility data — brand, model, category, alias — is preserved. The record becomes Unmapped instead of Orphaned. This is the right choice when you want to keep the compatibility relationship and possibly remap to a different product later.
• Delete — Permanently removes the record. Use this when the compatibility data itself is no longer needed (the product is fully retired, the model line is discontinued, etc.).

Bulk operations on the Orphans tab are available on Growth and Professional plans. Starter merchants can act on orphans one at a time, with a clear upgrade prompt visible when bulk would help.

[Screenshot: Insights > Orphans tab showing several orphaned records with Unmap and Delete buttons per row.]

AI mapping operations

What it does and where to find it

AI mapping lives inside Connect Data’s Import from Shopify tab. Brief recap (the full walkthrough is in Your first records):

• AI mapping reads your Shopify product titles, vendor, type, and SKU.
• It suggests compatibility records based on web research.
• Each suggestion comes back with a confidence level (Verified, Likely, Unverified).
• You review every suggestion before anything is saved — nothing is auto-committed.

Available on every plan, with monthly caps that vary by tier.

Monthly limits and tracking

AI mapping has monthly usage caps that scale with your plan:

• Starter — 100 mappings per month
• Growth — 500 mappings per month
• Professional — 2000 mappings per month

A “mapping” counts each accepted record, not each product analyzed. The count only goes up when you click “Save records” with AI suggestions selected. Analyzing 100 products and reviewing the suggestions does not use any of your allowance; saving five records per product does (500 mappings used).

Limits reset on the first of each calendar month. You can see your current usage on the Connect Data page so you know how much of your monthly allotment is left.

If you reach your monthly cap mid-month, you can still:

• Use CSV upload — no AI involvement, no cap.
• Use manual entry — also unaffected by the AI cap.
• Wait for the first of the next month for the reset.
• Upgrade to a higher tier for a larger cap.

When to use it (and when not to)

AI mapping is most useful:

• During initial setup if you don’t have prepared compatibility data and you do have a populated Shopify catalog.
• When adding new product lines or expanding into brands you don’t have data for yet.
• When you want a starting set of compatibility records that you’ll review and refine, rather than having to research every part-to-equipment relationship by hand.

AI mapping is less useful:

• When you already have a CSV ready — use CSV upload instead. It’s deterministic, faster, and doesn’t burn your monthly cap.
• For products with ambiguous or generic titles where the AI has little to work with. (“Universal filter — pack of 4” gives the AI nothing to research.)
• For very specialized parts where compatibility info isn’t published anywhere on the open web.

Reviewing AI suggestions

Every AI suggestion has two pieces of metadata to help you decide whether to accept it:

• Confidence label — Verified (sourced from a manufacturer or major retailer listing), Likely (extracted from product attributes or pattern-matched), or Unverified (the AI couldn’t validate against an authoritative source).
• Source attribution — Where the suggestion came from. Common labels: “Manufacturer official listing,” “Retailer listing,” “Compatibility chart,” “Product title — size-based match,” “Product title — extracted.”

Even Verified suggestions should be reviewed. AI is helpful but not infallible — see the warning callout in Your first records for the full caveat.

Activity log

What it shows

The Activity Log tab in Insights shows recent events affecting your data, merged into a single chronological feed:

• Search events — Shopper searches against your data, with the brand, model, and result status (match / no match) for each.
• Import events — CSV imports, manual entries, and Shopify imports, with row counts and import kind for each.

A filter dropdown at the top lets you switch between All activity, Searches only, and Imports only, depending on what you’re investigating.

What to look for

Activity log is most useful when:

• You’re trying to confirm what just happened (“Did that import actually run? How many rows did it add?”).
• You’re debugging unexpected changes (“Why did my record count drop overnight?” — likely an auto-sync deletion or an orphan cleanup).
• You’re verifying that auto-sync activity is happening as expected after enabling it.

The log’s window is bounded by your plan’s analytics history:

• Starter — 30 days
• Growth — 90 days
• Professional — 365 days

Older events scroll off the bottom of the window. If you need a longer historical record, export periodically or upgrade your plan.

[Screenshot: Insights > Activity Log tab showing both search and import events in chronological order with the type filter visible.]

A few practical tips

• Run a manual orphan check after any bulk product change. Bulk deletes, catalog migrations, vendor changeovers — don’t wait for the automatic sweep, just hit “Check now” so the orphan count is fresh before you make decisions.
• Use auto-sync on Growth and Professional unless you have a specific reason not to. The cleanup work it saves is significant, and the deletion behavior is sensible for almost every store.
• For most merchants, “Update existing + add new” is the right CSV import mode forever. Replace all is destructive — only use it when you’re truly starting over.
• Keep an eye on the Unmapped count in the health strip. A high Unmapped count means most records aren’t linked to products, which limits the storefront experience (no images in finder results, no inline add-to-cart). Mapping coverage is one of the highest-leverage things to improve.
• The Activity log is your friend when something unexpected happens. Check there before contacting support — it usually answers “what happened?” faster than a support email round-trip would.
• Don’t aim for perfect data on day one. Records improve over time as you watch what shoppers actually search for and where the no-matches cluster. Insights tells you where to focus.

With your data clean and connected, the next thing to learn is how to read shopper behavior. See Understanding insights for how to interpret the analytics surface.