We prepared the semantic core, clustered it in the directions of vacancies (loader, picker, and so on). Then the resulting groups were distributed by advertising campaigns, dividing them by regions of Russia and the CIS countries (and, of course, by advertising platforms: networks and search).
During the first month, we exceeded the plan: the average cost of the application was about 120 rubles (with a target of 150 rubles). In the second month of cooperation, we brought the cost of the application with Yandex.Direct to 80 rubles (almost two times cheaper than the customer wanted!).
For more efficient spending of funds (and higher positions in search advertising) connected Yandex.Direct and Google Adwords to the Elama automatic bid control system.
We ve worked out the list of negative keywords in order to cut off unnecessary excess traffic, and then added to this list based on new statistics from search queries.
All the ads worked so that they were as possible more accurately responded to a user's search query.
As a result, we reached a CTR of 10% to 20 % on search. By the way, an important indicator: the average position of the click. In order not to overpay and at the same time have an acceptable volume of traffic, we made sure that this indicator was in the region of 3..4 positions.
We managed to collect VERY cheap traffic from YAN. As you can see, the cost of a click barely exceeds 3 rubles: this is the minimum cost at which ads are displayed with pictures in YAN. At a lower cost per click, only text would be shown, and performance (CTR) would decrease drastically.
To increase the conversion, we conducted a technical and marketing audit site, making the necessary improvements. In cases where this is possible, we try to create landing pages ourselves.
As always, at the end of each month - a detailed analytical report with a plan for further action.
PS Some of the numbers in the case description have been changed. The essence remains the same.