In March, Polish universities opened for entry applications, and our work began.
Taking into account the previous statistics, we decided to create a separate campaign with conversion keywords.
For that, after analyzing all the keywords and search queries, we selected only those that had conversions. Then, we set up a campaign with a clear structure and a sufficient budget. It included only the keywords that previously showed a significant number of conversions. Other ones were kept in their campaigns with a minimal budget.
As a result, only this campaign generated 48 conversions in 15 weeks.
The campaign described above improved the results, but it still was not enough.
Thus, we decided to add remarketing. It brought 28 conversions over the specified period. Moreover, these conversions cost about half of those reached with regular keyword campaigns.

Then, we decided to test the Discovery campaign because it had worked well for other projects. We created one group that targeted people who entered specific keywords (we took keywords that gave conversions in other campaigns), and a second group of users who visited pages similar to competitors’ websites (data was taken from the auction insights report). Although the cost per click, in this case, was only 0.69 UAH, and it gave 1.07 thousand clicks, it did not bring conversions. We decided to stop the experiment and allocate the budget to test other options.
The next step was creating a regular display campaign with the same audiences. We excluded displays in mobile apps and selected effective marketing materials. This method gave immediate conversions. During the campaign period (May 20 – June 21), we got 18 conversions for UAH 167.

Next, we planned to test Smart GDN, but the client asked to stop the work – the study groups were full, and the advertising was paused until the next application period.