Passware Boosts Google Ads Sales
Net Revenue By +24.75%

Our Approach

1. We used tROAS portfolio bidding strategy across all generic campaigns and a shared budget to give machine learning consolidated data. The hypothesis was that Google would start learning and setting up bids to reach the set target ROAS across all generic campaigns.
2. Started consolidating ad groups for a more simplified account structure. This was done to feed Google more data into ad groups. If possible, the ultimate goal was to get at least 3,000 impressions weekly per ad group.
3. Added all available assets to RSAs. Left one ETA per ad group in case RSAs would get disapproved to avoid losing traffic.
4. Used keyword insertion and pinned it to the first headline. Because Google’s RSAs learn slowly, it takes time for the machine to determine which asset combinations are winners. This was a purely experience-based decision, as we know that the probability of getting a click while having the keyword in the first headline is the highest. We later removed the headline pinning. 
5. Reviewed all keywords and ensured they would look good being inserted in the 1st headline. Some of the best-performing keywords needed to be paused, but the search terms were still being captured by other keywords in the same ad group.
6. We make sure the Search Lost IS (budget) metric is at 0%, as long as we deliver our return of investment level.