A cohort is a subgroup of users that share a common characteristic. Cohort reports allow marketers to slice and dice their users through deep segmentation. Marketers use cohort reports to uncover the common indicators of success (e.g. geography, campaign, user engagement), which cohorts are underperforming and where corrective action such as campaign optimization or user re-engagement is needed.
Early in a marketing campaign, marketers typically use cohort analysis to identify, define and refine their KPIs (key performance indicators). For example, in this case study, a highly successful mobile app used AppsFlyer’s cohort reports to determine which segments of new installs were likely to remain active, loyal users thirty days later (based on user-engagement data and in-app event tracking). Based on these insights, the marketer began sending automated postbacks whenever users met specific engagement thresholds. This allowed their ad networks to better optimize their lookalike targeting, improving the quality of their installs at remarkable scale.
Marketers commonly use cohort reports to both improve their users’ lifetime value (LTV) and their scale. AppsFlyer’s dynamic cohort reports provide marketers with maximum flexibility, so they can group and regroup their users in dozens of potential configurations. Revenue events and clear data visualizations help marketers identify performance trends, pockets of opportunity and early indicators of success. Marketers can even define their minimum cohort size (the minimum number of users needed to for a cohort), define custom date ranges and multi-select from a variety of filters, advanced filters and groupings. The entire cohort report can then be further sorted based on sessions or any measured in-app events. Furthermore, just like in the Overview Dashboard, marketers interact with the live graphics, adding and removing cohorts by clicking on their name in the legend, hovering over trend lines for additional data, clicking on column headers to sort the table by that column, or export the entire cohort dataset to CSV.
Here are a few more helpful resources to help you make the most of your cohort reports: