The solution
We started on Google Ads and Meta Ads together, and added Conversion Rate Optimization later. Tracking, feed work and monthly reporting ran underneath all 3 for the length of the partnership.
Pay Per ClickDefending the brand’s own name, and refusing to chase the rest
ServicesGoogle Ads · Feed Management · Tracking
Google Ads went after the people already looking for Sease. Search campaigns covered the trademark terms, shopping campaigns chased the same demand, and display remarketing brought back people who had already visited the site. Non-brand keywords kept a small allowance and never earned more, because they never produced sales.
None of it worked without a product feed, and there wasn’t one yet. We wrote the specification, every attribute Google needs and what it is for, then walked the developers through generating it out of the custom CMS. From there it went into DataFeedWatch, where we reworked titles, categories and product data without anyone touching the site.
The catalog was big enough that generation sometimes timed out on the server and the feed failed to refresh, which put feed checks into the monthly routine.
Selling in seasons
The display remarketing ran on static banners built to the brand’s look. Sease works in collections, spring summer and fall winter, and every launch rewrote the website underneath the ads, so copy, banners and shopping segments moved with each one.
Paid SocialA catalog that had to look the part
ServicesMeta Ads · Creative
Meta handled acquisition, remarketing and brand building, each with its own campaigns.
- Catalog and collection ads for the people close to buying
- Carousels behind each seasonal push
- Post engagement running the brand’s own Instagram content, swapped out every week or two
We built the carousel and collection ads around the brand’s own imagery, so every ad followed Meta’s best practice and still looked like Sease. Once that was running, we added a dynamic remarketing campaign shown only to people who had already browsed the site, which gave the account the automation Meta rewards without loosening the look.
Facebook, on the data
The brand wanted Instagram and nothing else, which is the obvious call for luxury menswear. The data disagreed. Facebook was producing sales, we brought the numbers, and Facebook stayed.
What one season taught the next shoot
Creative came from Sease’s own shoots, so our work sat at both ends of it. Before a season we sent specs: image and video dimensions, the crops each placement needs, whether a cut had to work with the sound off. After a season we pulled every ad’s data out of Meta and worked through it by hand, which only holds together with strict naming across campaigns, ad sets and assets, because the whole analysis is built on those labels.
Each campaign type was judged on its own terms, brand awareness on impressions, clicks and click-through rate, catalog campaigns on sales and return on ad spend. The report came back as slides, with the top 3 winners and the top 3 losers in each of acquisition, remarketing, brand awareness and post engagement, then the reasoning behind them. Those findings went straight into the brief for the next photoshoot: which backgrounds held attention, which crops survived a feed, which products earned the camera.
CROWatching people try to buy
ServicesConversion Audit · UX & Design
Conversion rate optimization joined later, in a lighter form than the growth service we run today: Hotjar heatmaps, session recordings and on-site surveys, read every month and turned into specific changes for the product pages and the checkout.
People hesitated on product pages, where a €3,000 jacket was not making its own case clearly enough, and they hesitated again in checkout, over extra form fields, unclear steps, and nothing on screen telling them the payment was safe. The surveys asked the question directly, and the answers came back specific.
Finding a hidden checkout issue
One of those answers was that the payment would not go through at all. We went back to the recordings and found it: on certain devices the purchase button wasn’t responding. It was easy to miss, because on every other device the checkout worked perfectly. We shared the recordings with the client and the developers, and the fix went in.
What followed was small and precise.
- Clearer value on the product page, so the price had something to stand on
- Fewer fields in the checkout
- Payment icons and reassurance at the exact point people were hesitating
Each change was measured before and after, so the account knew which one had done what.
The results
The Fall-Winter collection tripled its revenue against the same collection the year before, and Sease grew from a young ecommerce operation into a brand with the demand to carry a womenswear launch. The rest of the paid picture moved with it.
- Return on ad spend up 5%
- Average order value up 7%
Within a month of the first CRO fixes the ecommerce conversion rate doubled, and conversion value rose 288% over 7 months. The checkout fix was part of that, and so was everything that stopped being guesswork the moment recordings and surveys started running.
The feed we built for Google Shopping became the catalog Meta advertised from. The creative analysis told the brand what to shoot next, so each season’s ads were built on what the previous season’s data had shown. The CRO work repaired the page both platforms were pointing at, which raised what every euro on either of them was worth.
Reporting ran through GA4 and a Looker Studio dashboard with the creative analysis on top, so every decision came out of the same picture each month. Budgets climbed hard into November, December and January, the brand’s busiest season.
3X
Fall-Winter revenue · against the same collection a year earlier
+7%
Average order value · with ROAS up 5%
+100%
Ecommerce conversion rate · first month of CRO fixes
+288%
Conversion value · over 7 months