The first stage: audience clustering. This step was supposed to answer two questions:

Why do users come to the site? What difficulties do they face there?

To do this, the company needed statistics for 9 months and a thorough analysis of user behavior patterns. We took into account all their actions on the site: from which channels and devices they come, how much time they spend on the site, which sections they get acquainted with and in what sequence, whether they carefully study the composition of drugs, how they react to promotions, what is the age of users. As a result, it was possible to identify 5 clusters and analyze them.

The second stage:
The strategy of placement and optimization of the site. Analysis of user behavior helped to discover several interesting patterns:

Users under the age of 30 were often interested in products for people from 55 and above. This means that they are not always looking online for information for themselves. This insight allowed us to adjust the advertising strategy and take into account such audiences in future campaigns.

A cluster consisting of 27% of users from the entire audience goes to the site from non-advertising traffic. These people view the most pages per session, but do not perform conversion actions. It seemed that such users lost interest in the information on the site, but at the same time they continue to come here even without advertising. This made me think about the difficulties that can occur on the way of users who go to the site from mobile devices, visit mainly the catalog pages and leave the site because they did not find anything they needed. They should be given special attention, and the company decided to test the optimization option of the search bar on the site. This was supposed to help users quickly find what they didn't find during the session in the directory.

The cluster, which goes to the site from advertising traffic and accounts for 31% of all users, carefully studies drugs, but is not interested in information about offline pharmacies. There were many new users among this audience. Realweb conducted an A/B test in order to tell the audience on the main page where to buy the right product offline. To do this, the button "Buy at the nearest pharmacy "Evalar"" was moved to a more prominent place. We decided to test the banner in height — placed on the main page, it could annoy mobile users.

The third stage: launching A/B tests in Google Optimize
Test #1: "Optimization of the site search bar".
The hypothesis sounded like this: if you add search fields to the catalog page, user engagement will increase. This assumption was tested by the following indicators:
Increasing the audience's interest in the list of pharmacies
Increasing the number of registrations on the site

If users quickly and easily find the right product, they will ask where they can buy it offline. If the audience finds it on the site, its loyalty increases — as a result, subscriptions to news and newsletters grow.

Test results via Google Optimize:
User interaction with the search bar has increased by 70%

The number of registrations (subscriptions to news and newsletters) increased by 18%

Mobile users turned to the form with questions 24% more often

The conversion rate for the "Viewing addresses offline" goal increased by 3%

Test #2: "Adding the button "Buy at the nearest pharmacy".
The task here was to find out: if the audience understands where to buy goods offline, will this increase its involvement online?

User engagement was measured by:
Increasing the audience's interest in the list of pharmacies

Increase the duration of the session
Reducing the number of failures

If the user understands where he can buy the product offline, then before buying, he will study the information about the drug and possible discounts on it. When he becomes interested, he will most likely want to subscribe to the company's news and newsletters in order to be aware of such information.

Results:
-User interaction with pharmacy contact information increased by 66%
-The number of page views with promotions increased by 35%
-The duration of the session increased 1.6 times
-The number of refusals decreased by 1.3 times
-The number of registrations increased by 1.7 times