Vanity Resource: A Clean Feed Won Shopping

The Setup

Vanity Resource sells high-end bathroom vanities direct to consumer. The catalog was solid, the brand had margin, and the team had been relying on organic search with results that were not scaling.

The issue wasn’t demand. The issue was that bigger competitors with cleaner feeds were winning impression share on the exact searches Vanity Resource needed to own.

Google Shopping does not read your ad copy, because there isn’t any. It reads your product data and decides which queries you deserve to appear on. A vanity is a considered purchase with a specific vocabulary: size in inches, sink configuration, top material, finish. A buyer types those words. If the feed does not carry them, the auction never happens.

What Was Broken

1. The feed was incomplete. Missing GTINs on a meaningful slice of products, generic product titles that didn’t include the searched-for attributes, and no use of custom labels. A title reading “Ashwood Vanity” competes for nothing. The same product titled with width, sink count, and top material competes for the query a buyer with a tape measure types.
2. One campaign, one priority, one bid strategy. No separation between branded queries, generic queries, and high-intent long-tail queries. Everything was getting bid on the same way, which means the clearance item and the flagship vanity were fighting each other for the same budget on the same terms.
3. No exclusions. Shopping was matching queries from people researching DIY projects, looking for repair services, or shopping at radically different price points. That was measurable in the behavior, not only in the search terms: paid sessions were bouncing on the same product pages organic visitors engaged with cleanly, which is the signature of an audience problem rather than a page problem.

What I Did

Cleaned the feed at the source. Worked through the WooCommerce product attributes to fix titles, descriptions, GTINs, and Google product categories. Fixing it at the source rather than in a feed rule matters, because the corrected attributes then serve the storefront, the sitemap, and any future channel instead of living inside one Merchant Center transformation nobody documents. Added a supplemental feed for the gaps the native feed couldn’t fill.

Tagged products with custom labels. Margin tier, seasonal flag, and best-seller status. These became the foundation for splitting Shopping by priority. Custom labels are the only lever that lets a retailer tell Google which products deserve which budget, and most catalogs I audit leave all five slots empty.

Built a priority structure. Three Shopping campaigns running at high / medium / low priority with shared negatives, so high-intent queries got routed to the campaign with the strongest bids and clearance products got routed to the campaign with the tightest budget.

Built the negative list. Negatives layered in across the structure, with a weekly review cadence to keep the noise out.

The Result

• Revenue more than doubled within the first 60 days, climbing well over 100 percent versus monthly averages
• Shopping ROAS lifted 2.4x on the rebuilt structure
• Impression share on the highest-intent queries grew 38 percent on the new Shopping structure
• Engaged sessions rose 88.8 percent against the prior period
• 22 percent of the monthly budget that had been going to wrong-fit queries got reallocated to working ones

The engaged-sessions number is the one worth sitting with. The product pages never changed. The same pages that were bouncing paid visitors started holding them, because the visitors arriving were the ones the pages were written for.

The owner’s words: “Wonderful! Explains it thoroughly so I understand. More importantly, seeing great results in such a short time period.”  ·  Olivia Streeton

What I Did Not Claim

I did not run a holdout. The feed cleanup, the priority structure, and the negatives shipped inside the same 60-day window, and the revenue lift is the combined effect of all three against the prior period. Seasonality on a bathroom-remodel catalog is real, and a single 60-day comparison cannot fully separate it from the work. What the numbers support is a directional read on a rebuilt structure, not an incrementality measurement.

I also did not touch the site. No new photography, no page redesign, no offer change. That is the constraint that makes the engaged-sessions number readable at all.

The Takeaway

In Google Shopping, the campaign settings are downstream of the feed. You can have the perfect bid strategy and the right negatives, but if your titles don’t match the queries buyers are typing, you’ll lose to the brand that does match.

Most Shopping audits I run start with a feed audit. That’s where 70% of the wins live.

If you sell considered-purchase products with measurable attributes, the diagnostic takes an afternoon. Export your feed. Read fifty product titles as if you were the buyer. Count how many carry the dimension, material, and configuration words a buyer would type. Then pull ninety days of search terms and mark the share that came from someone who was never going to buy at your price point. Those two numbers tell you whether your problem is the feed or the bids, and in my experience it is the feed. The same work runs inside the Imagery and Shopping program for home and furniture brands.

Running Shopping or PMax and not sure if your feed is the bottleneck? Let’s look at it together.