PaspasBurada

PaspasBurada is an e-commerce brand that makes custom EVA car mats for each car's make and model. When we started working together, they were running their Google Ads account themselves. The ads were spending money every day, but nobody knew exactly how much of it turned into sales.

When we reviewed the account, we found three core problems. Conversion tracking wasn't working, so Google couldn't learn which clicks led to a sale. The bidding strategy was set for impression share rather than sales: the system tried to show the ads as often as possible instead of looking for buyers. There was only 1 negative keyword against 50 broad match keywords, and part of the budget went to competitor brand searches.

The picture was clear: every 1 lira spent on ads brought in only 1.8 lira in sales, and only 26% of orders came from ads.

Every optimization made on bad data is also wrong. So our first job wasn't touching the campaigns, it was repairing measurement. We rebuilt conversion tracking and verified it against the real orders coming into the site.

We also added our own funnel tracking to the site: page views, add to cart, checkout and orders are recorded separately. That way we can see where visitors drop off without relying on Google's report. We noticed that the gap between server time and Turkey time was putting late-night orders on the wrong day, and fixed the reports as well.

Once measurement was working correctly, we switched the bidding strategy from impression share to sales. Google was now showing the ads to searches likely to buy, not wherever they could be shown the most.

We moved broad match keywords to phrase match. We went through the search term reports one by one and added competitor brand searches and queries unrelated to car mats to the negative list. No sales were lost during this cleanup; wasted budget shifted directly to people ready to buy.

More budget doesn't always mean more sales. The data showed that past a certain point, extra spend only pushed up the cost per click, so we capped the budget there. We increased spend only where the data allowed it, not blindly.

Once the account was fed with accurate data, we turned on AI Max, Google's AI-assisted search feature. With healthy conversion data behind it, the automation started finding new, relevant searches on its own.

No matter how good the ad is, if the person who clicks lands on a slow, confusing site, there's no sale. We rebuilt PaspasBurada's old site from scratch with Next.js: a fast, simple shopping experience that's easy to use on mobile.

The heart of the site is the mat configurator. Customers choose their car's make and model, pick the base color, trim color and pattern, and see the result instantly in a preview. We prepared a car icon for each of more than 600 models across over 40 brands, so customers pick their car by recognizing it in the list.

We noticed that order confirmation emails weren't reaching Hotmail and Outlook users. This affected about 21% of orders. We moved email delivery to a reliable service and fixed the problem for good.

To catch an unexpected drop in sales the same day rather than the next, we set up an automated health check and alerting system. When something goes wrong on the site, in payments or in advertising, we know about it.

Over the 86-day period, daily orders grew 208% compared with the previous period. With ad spend up 120%, orders growing much faster shows the budget is now going to the right place.

Every 1 lira spent on ads brought in 4.9 lira in sales instead of 1.8. The conversion rate from ad click to order rose from 0.41% to 1.47%. Cost per order from ads fell 62%, and cost across all orders fell 28%. 49% of orders now come from ads, up from 26% in the previous period.