{"id":285,"date":"2020-12-03T20:09:13","date_gmt":"2020-12-03T20:09:13","guid":{"rendered":"https:\/\/oqtacore.com\/app\/?p=285"},"modified":"2022-11-23T12:39:41","modified_gmt":"2022-11-23T12:39:41","slug":"segmentation-of-target-audience","status":"publish","type":"post","link":"https:\/\/oqtacore.com\/blog\/segmentation-of-target-audience\/","title":{"rendered":"Segmentation of the target audience for a travel agency"},"content":{"rendered":"

Purpose:\u00a0<\/strong>to personalize the offers of tours in accordance by applying segmentation of customers. Increase ad clicks on open emails<\/p>\n

Analytical report on the solution of the problem<\/p>\n

<\/span>The problem of segmentation<\/strong><\/span><\/h2>\n

Task:<\/strong><\/p>\n

    \n
  1. Segment your target audience<\/li>\n
  2. Identify preferences of different segments<\/li>\n
  3. To formulate recommendations on the distribution of tour offers corresponding to segment preferences ..<\/li>\n<\/ol>\n

    Assessment of the quality of the solution to the problem<\/strong><\/p>\n

    An increase in the ratio of the number of scanned letters to the number of all sent tour offers acts as a measure of the quality of solving the problem<\/p>\n

    <\/span>The segmentation approach<\/strong><\/span><\/h2>\n

    To highlight the segments of CA based on statistical data on customer activity when viewing the site, cluster data analysis algorithms were applied.<\/p>\n

    To identify segment preferences, clusters were compared with tags for the proposed tours, using text mining ( text mining ) methods .<\/p>\n

    Manually, none of the tasks could be adequately handled due to the volume of statistics and the number of related parameters. The analysis used data from more than 80,000 clients, for each of which there were more than 500 features and more than a hundred different offers of tours.<\/p>\n

    All the email list subscribers emails were clustered based on their behavior and given their personal offers<\/strong><\/p>\n

    <\/span>Results of segmentation<\/strong><\/span><\/h2>\n

    1. Target segment and their preferences<\/h3>\n

    According to customer behavior data on the site (page views), the clustering algorithms allocated 5 clusters:<\/p>\n

      \n
    1. Cluster 1 is the main one, it includes clients with the most typical behavior: they are very interested in beach vacations and summer vacations, they are interested in memorable excursions no less than summer vacations in general, the idea of \u200b\u200brelaxing with children is not too attractive to them, although they often relax with them than the two of you. Special offers are not interested.<\/li>\n
    2. Cluster 2 as a whole is interested in travel (except special it dry redlozheny ) more than anyone, but mainly rest (unlike the cluster 4). Briefly \u2013 this is a more active version of cluster 1 with the only differences that excursions catch it much less in comparison with the theme of beach holidays than cluster 1.<\/li>\n
    3. Cluster 3 is generally very little interested in traveling, but, unlike cluster 5, it is still realistic to attract offers of beach vacations in the summer, but special. suggestions are useless. It seems that this cluster includes people who are deeply immersed in work, who rarely travel and sometimes remember about vacation, because \u201ceveryone goes to the sea in the summer.\u201d In other seasons, vacation to cluster 3 is not interesting.<\/li>\n
    4. Cluster 4 demonstrates typical behavior, as does cluster 1 (with the same level of general interest in tours) with one feature: interest in special offers is extremely pronounced. This topic \u201ccatches\u201d customers from cluster 4 as much as the theme of summer holidays in general.<\/li>\n
    5. Cluster 5 is only interested in special offers.<\/li>\n<\/ol>\n
      \"Diagram<\/figure>\n
      \"Diagram<\/figure>\n

      2. General comments on target audience segmentation<\/p>\n

      Based on the information provided, both common patterns of behavior for all clients and specific for each of the identified clusters were identified. The presence of general patterns explains the selection of one main segment and several smaller segments.<\/p>\n

      You can either focus on the main segment of Central Asia, or personalize proposals for segments.<\/p>\n

      3. General observations<\/p>\n