The client was dissatisfied with the actions of the previous marketing agency. At the very beginning, two sales conversion sources were connected to the account: one directly from the Google Ads account and the other from Google Analytics. Although it is not a bad setup, in this case, one of the conversions needed to be set as "additional." This way, Google would not consider those data in the statistics. With both conversions counted together, the sales results on the account were calculated 2-3 times, causing significant issues in optimizing the campaigns effectively. Additionally, at the beginning of March, a decrease in clicks was observed due to the disconnection of the Google Merchant Center account by the previous agency from the Google Ads account.
“Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.”
Therefore, the first thing we focused on in mid-March was creating a new Google Merchant Center account for this Google Ads account. Then, we had to set up new Performance Max campaigns. For this purpose, we used a division where one Performance Max campaign was dedicated to a single product category.
We checked the currently set conversions. Unfortunately, the previous sales-oriented Google Ads were set up incorrectly, so we created a new one. After a week, when we were sure that the conversion was performing well, it was set as the primary conversion. The remaining sales-oriented conversions from Google Analytics were set as additional conversions. From that point on, the view on conversions was clear. Additionally, we removed text campaigns that were not bringing in conversions and were mainly generating losses.
In this quarter, we managed to achieve a ROAS (Return on Advertising Spend) of over 350% in each month, with an overall average of more than 402%.
During these months, the optimization process involved testing campaign behaviors after various optimization moves. For underperforming campaigns, we experimented with increasing or decreasing the ROAS, changing ad displays in major cities in Poland, and excluding products that generated a lot of clicks but had no sales.
If a campaign still didn't yield satisfactory results, it was paused.
In other campaigns where sales were doing well, I tested the impact of increasing and decreasing the ROAS. Then, I examined products that had the highest number of clicks and sales. If there were products that were not selling well, I created separate campaigns for them. This helped us determine whether the products were not selling because Google's algorithm chose other products to display or if they were genuinely weak in performance.
After a month of testing, if a campaign continued to underperform, it was paused.
The next optimization step involved increasing the budget for successful campaigns and reducing the budget for those that were performing poorly.