Get a free consultation with a marketer marketolog@adbest.ru>

Segmenting consumers by website behavior versus segmenting by CRM

This is another innovative development for you! We've developed a special way to process contextual advertising statistics using cluster analysis to divide all website visitors into groups based on their interest in a specific product group. This may reveal patterns you never even suspected. In particular, it may turn out that a company has a record-breaking product that deserves its own brand. In any case, after this research, you'll have no problem understanding which products are searched for together.

Consumer Segmentation Training

The most interesting thing, of course, is to create a profile of the typical visitor for each segment. After all, the study allows us not only to group pages visited by similar visitors, but also to consider this information in conjunction with data on socio-demographic characteristics, solvency level, device type used, and regions of display. The ratio of segment sizes is also interesting.

Description of the cluster analysis method

Next, we'll segment consumers using the same cluster analysis method, but using the CRM database. We'll compare the results, and you'll see that customer interests expressed on the website may differ from the actual segments determined by actual purchases from the CRM. This information will help you focus on the most in-demand products and understand market trends.

Стоимость сегментирования

The consumer segmentation method discussed above is a non-response method. In addition to cluster analysis, responsive segmentation methods are also used, which divide consumers into two or more groups based on the performance of a specific action, such as whether they purchased or did not purchase, visited or did not visit a website, visited a specific number of pages, etc. The most important responsive segmentation methods are discriminant analysis and the construction of decision trees to identify nodes that influence the outcome (the CHAID method).

Description of the discriminant analysis method

We also widely use these methods to set up contextual advertising to understand which factors have the greatest impact on the desired result specified in the conversion rule. Such factors can include audience demographics, the content of the pages themselves, and quantitative parameters such as cost per click, budget, and number of impressions.

Using segmentation for contextual advertising