The thing is that it is better not to consider advertising in a vacuum, because all campaigns influence each other in one way or another. For example, by launching an advertising campaign for branded queries in Yandex search, you take traffic and leads from organic branded traffic. By launching advertising campaigns for queries for which you are at the top of search engines, you take leads from branded search traffic.
So the client's situation was that the company did not count how costa rica cell phone number list many additional leads and sales the advertising brought. Only the total number was counted and the results before and after the launch of the campaign were not taken into account. Therefore, the client, despite the fact that he launched high-quality advertising, had no idea how it worked and how effective it was.
In order to correct the situation, it was necessary to calculate the incremental effect, or in other words, the supplementary effect of advertising campaigns.
Where Millions Go - Image 3
This happened in the following way. The business had clustering, that is, users were divided into clusters, and advertising campaigns into categories. We conducted an analysis and saw that brand advertising in St. Petersburg and Moscow works better.
The Cost of Analytics Mistakes: Where Millions Go - Image 4
Then we looked at the effect of advertising campaigns together with conversions and calculated the incremental effect, i.e. the squeezed value from each advertisement, using this data. It turned out that the brand campaign in Moscow gives a negative ROMI. Most people who came through this advertisement were ready to buy the product anyway.
Thus, calculating the incremental effect helped find an answer to the question of whether the business was making money on the ads it launched. As a result, campaigns that were producing good results cannibalized other campaigns, as well as other acquisition channels. We automated the incremental analysis, which allowed us to increase additional monthly income from 1.7 million rubles to 3.6 million rubles.