Pay Per ClickSplitting the catalog the way it actually sells
ServicesGoogle Ads · Feed Management · Microsoft Ads
The account was built for an earlier stage of the business. We went through the catalog product by product and grouped it by what really drives sales.
- Category, with the dresses split out from knitwear, jewelry and accessories
- Price point and sales volume, so best sellers and slower lines each got the right investment
- The dress itself: longer styles for evenings out, lighter ones for everyday wear, each with its own season
Budgets and targeting followed those groups through the year, pushing the styles the season favored and easing off the ones it didn’t.
A structure the algorithm could run
The point of the segmentation was to stop editing the account by hand. Once each group sat in the right place, our job shifted to giving the bidding algorithm the cleanest possible signal from whatever was in stock that week. The structure stayed small enough to manage, so when stock or the season shifted we could move the same day.
We also made sure the whole catalog got its turn. Earrings, bracelets and the other pieces that finish an outfit sat in the account alongside the dresses, because the brand sells a look as much as a garment.
Staying out of the wrong searches
For the foundation collection, we built negative keyword lists that kept the ads off every search tied to it. The organic traffic still came, the cause still got its attention, and the ad budget stayed on products the brand could profit from. That one move saved $4,200 in spend.
Celebrity spikes got the same daily attention. We watched what was happening on social and in the press, read the surge for what it was, and held the account steady instead of letting it chase curious clicks.
Testing in measured steps
A brand that sells well tempts everyone to try everything. We kept experiments small and tied to what the data already supported, so the account could try new campaign types as Google introduced them without putting performance at risk.