Creating a comprehensive tool to explain user behaviour in the product and considering how the tool helps in finding insights
As your product evolves, additional concerns arise about whether a certain user behaviour is related to their chance of returning or purchasing anything. Questions from numerous stakeholders at various levels include: would ARPU be higher if we compare users who have visited more than ten pages with a product to users who have visited fewer than ten? Is it more probable for a user to return to a product after placing more than two orders in the first week? Do consumers who utilize a particular functionality more than five times have a greater lifetime? Or, even better, let’s focus on the functions that affect the main metric… Product analysts who have completed this level of tasks and analysis will be able to predict the answer to such questions immediately…
All of the above questions, as well as others of a similar type, may be answered with confidence. Every action taken by the user increases the probability of progressing to the next step (conversion to a target action, return to a service, etc.). This is exactly what engagement means: the readiness of a user at his current level of development to take actions regarding the product, such as selecting a goods/a service/some content offered in UI, passing the main product funnel, and continuing to engage with the product after completing the target action. Therefore, a unified method for assessing a user in comparison to other users may be implemented by developing an engagement model.
An engagement model analyzes each individual user’s activities regarding a product, indicating the user’s desire to engage with the current product.
The user with id 13 is the most involved here, while the user with id 10 is the least.
As a result of such an analysis, we may categorize users into those who know and understand how to use the product, those who are interested in learning more about the product in its current state, and those who accessed the product and soon departed without discovering anything substantial after performing simple actions.
Eventually, as demonstrated in the second part of the article, segmentation by level of engagement allows connecting product metrics such as CR, CTR, Retention, Stickiness, and business metrics such as ARPU, ARPPU, LTV.
In general Engagement model helps in the following areas:
Continue reading: https://towardsdatascience.com/product-analytics-engagement-model-22d53c96d169?source=rss—-7f60cf5620c9—4