Microsoft Ads Product Explorer: Fix Feed Issues Fast

Microsoft Ads Product Explorer is a catalog-diagnostics view that Microsoft Advertising rolled out inside Microsoft Merchant Center around June 12–13, 2026, giving advertisers a single searchable, filterable lens over an entire product catalog — paired with trailing-30-day performance data — instead of bouncing between disconnected reports. At launch it is available to U.S. advertisers with fewer than 100,000 SKUs.

That scope matters. The under-100,000-SKU launch window points the tool squarely at mid-market retailers — the Shopify, WooCommerce, and Magento stores where feed errors tend to be most common and tooling has historically been thinnest — rather than at enterprise catalogs. And the timing is not incidental: with Microsoft pushing feeds into Copilot-era agentic shopping, a clean catalog is becoming a prerequisite for discovery, not just a Shopping-campaign hygiene task.

This guide covers what Product Explorer actually exposes, how to read its product states, and the closed-loop fix workflow that connects it to Recommended Actions and supplemental feeds. We add two proprietary references — a triage priority matrix and a Microsoft-versus-Google feed-attribute comparison — and finish with a weekly routine you can run in well under an hour. Feature specifics below are vendor-stated via Microsoft’s Ads Liaison, Navah Hopkins, as reported by trade press; verify behavior in your own account before relying on it.

Product Explorer is a new interface inside Microsoft Merchant Center (MMC), not a feature bolted onto the campaign UI. It gives advertisers a searchable, filterable view of their entire product catalog in one place, with performance metrics attached to each product. According to Microsoft’s Ads Liaison, the tool was built explicitly in response to advertiser feedback that managing feeds in Microsoft meant hopping between multiple reports and workflows.

Trade-press coverage of the launch appeared on June 12, 2026 (Search Engine Land) and June 13, 2026 (PPC News Feed), and Search Engine Journal carried a direct quote from Microsoft on the rationale. Note the sourcing carefully: at the time of writing there was no corresponding post on Microsoft’s official product blog. The details below are vendor-stated via the Ads Liaison and reported by trade press, which is how Microsoft Advertising routinely communicates — treat them as announced rather than documentation-confirmed.

Product Explorer’s value comes from combining two layers in one view: the feed attributes you filter on, and the performance metrics you evaluate. The reported filterable attributes are Title, Product ID, Brand, GTIN, Product Type, and Custom Labels. The reported performance metrics are Impressions, Clicks, Conversions, Spend, CTR, and Conversion Rate, drawn from a trailing-30-day window directly inside the interface. Filtered lists can also be exported for offline analysis, though the file format was not specified at launch.

The Custom Labels dimension is the quiet workhorse here. Custom Labels 0–4 in Merchant Center let you segment products by business-defined criteria — margin tier, bestseller status, seasonality, clearance, or priority level — and those labels show up as filterable dimensions in Product Explorer. That means you can slice catalog health by what the business actually cares about, not just by raw attributes. For the underlying labelling discipline, our feed optimization decision matrix walks through which attributes to prioritize first.

Product Explorer surfaces three product states: serving (active and eligible), limited (constrained by a feed issue), and rejected (withheld from campaigns). That framing maps onto the four historical Merchant Center statuses — Published, Rejected, Warnings, and Pending — with the “limited” label reframing warning-level issues as something actionable rather than a passive caution.

The distinction worth internalizing is that limited is the dangerous middle. Rejected SKUs are obvious: they are not running, and the loss is visible. Limited SKUs still serve, so they rarely trigger an alarm, yet they are quietly leaving impressions and conversions on the table because a non-compliant attribute is suppressing their reach. Product Explorer’s value is making that silent middle tier easy to isolate and fix.

The reason Product Explorer is more than a prettier report is that it sits at the front of a complete fix loop. Product Explorer finds the problem SKUs, Microsoft Advertising’s Recommended Actions feature tells you what to fix with issue-specific guidance, and supplemental feeds let you patch the offending attributes without re-uploading your entire catalog. Microsoft made supplemental feeds globally available on September 17, 2025, specifically so merchants could update individual attributes — title, description, custom labels, size, color, promotion IDs — instead of re-submitting everything.

That three-step loop is the operational core of this post. Most published coverage stops at “Product Explorer shows you rejected products.” The leverage is in connecting it forward: diagnose in Product Explorer, get the fix from Recommended Actions, and apply the fix through a supplemental feed so the change is fast, surgical, and reversible.

As Microsoft framed the supplemental-feeds problem it solved, merchants previously “needed to reupload entire product catalogs when making changes to individual product attributes.” That is exactly the friction this loop removes. The same feed discipline pays off across platforms — the attribute-level fixes you make here are the same ones described in our product feed strategy for Google Shopping, and they feed directly into Performance Max asset testing, since PMax draws from the same Merchant Center catalog.

Below is our operational runbook for turning Product Explorer’s states into an ordered set of actions. It maps each product state to the filter you apply, the fix mechanism, an urgency level, and a realistic posture on how long it takes a product to return to serving after a fix. Treat the recovery timing as a planning expectation, not a guarantee — actual re-review timing depends on Microsoft’s feed processing and is not a published SLA.

The ordering principle is simple: sort the worklist by business impact, not by alphabetical SKU. Filtering rejected and limited states and then sorting by Spend or Conversions puts the products that are both broken and valuable at the top, so the first thirty minutes of triage recover the most reach.

Most feed-health coverage is Google-centric, which leaves a real blind spot: a feed that passes cleanly in Google Merchant Center can still get products limited or rejected on Microsoft, because the platforms apply some attribute rules differently. The reference below collects the Microsoft-side requirements we have sourced — and flags where Microsoft’s specifics, rather than Google’s, are most likely to be what Product Explorer surfaces. Where a Google value is not separately sourced here, we mark it as broadly comparable rather than asserting a specific Google figure.

Product Explorer arrives in the middle of a deliberate Microsoft push to make Merchant Center feeds the substrate for Copilot-era agentic commerce. On April 21, 2026, Microsoft announced UCP-ready (Universal Commerce Protocol) feed support for U.S. Merchant Center, enabling product data to power agentic purchasing flows via Copilot. As Microsoft frames the split, feeds “primarily enable discovery (products showing up in AI responses), while UCP enables transactions (checkout and post-purchase actions within AI experiences like Copilot).”

Read forward, that reframes what feed health is for. A constrained or rejected product is not only missing from a Shopping auction — it is increasingly at risk of being invisible to AI-assisted discovery as well. Microsoft reports that Copilot Checkout has expanded to more than 500,000 merchants and that some top Shopify merchants saw roughly 90% growth in Copilot impression share through real-time feeds; both figures are Microsoft-stated and independently unverified, so treat them as directional signals of intent rather than guarantees. The strategic point stands regardless of the exact numbers: the cost of a dirty feed is rising as more surfaces consume it.

Microsoft Advertising’s own positioning makes the practical case plainly: “A cleaner, more complete feed can help your ads match more relevant shopping searches.” Product Explorer is the instrument that makes “cleaner and more complete” a thing you can actually measure and act on each week.

The point of a diagnostics tool is the cadence it enables. Here is the repeatable routine we run for Microsoft Shopping accounts using Product Explorer — designed to fit in well under an hour and to recover the most reach first.

Run that loop weekly and the catalog trends toward fully serving instead of accumulating silent rejections. For teams that would rather have this operated for them — wired into Google Shopping and Performance Max in parallel — our paid media management and ecommerce growth engagements build exactly this kind of cross-platform feed-health cadence, with the measurement to prove it is working.