When working with premium goods, it is important for us that insolvent people do not click on the ads. Otherwise, the client will receive clicks and calls, but there will be no sales. To filter out the insolvent audience, we started writing the price of a sofa in the ads as “from 160-180 thousand rubles”.
The list of negative keywords was supplemented by a list of furniture factories that sell products in a lower price segment. Even if people searched by the name of the factory, they could not get to Claire's site.
Test of different visuals
We understood that people are attracted by the beauty of sofas, so we part time data constantly tested about a hundred advertisements with different photos and selected only the most effective ones:
Photo
They selected it like this: they launched an ad, gave it three days to warm up, and looked at the results — the “60 seconds” indicator and the bounce rate. If the planned indicators were met, they left the ad, if not, they tested the next one.
Adjustment rates
We segmented the audience by demographic and behavioral characteristics. For example, spent so much time on the site, visited such and such pages. For each segment, we looked at the “60 seconds” goal. If the goal was not met, we set corrective bids so that the ad was shown less often.
In addition to standard audiences, we also used look-a-like audiences, which gave us the opportunity to significantly increase our reach.
Results
During our work, we focused on two indicators: achieving the “60 seconds” goal and showroom visits. They can only be assessed together.
We not only counted how many times the "60 seconds" goal was achieved, but also calculated the cost of achieving such a goal. As a result, it decreased by 7 times, and the number of visits to the showroom increased by 9 times.