Funnel Reports - Median Time from First Step

Related products: PX Analytics

Hi everyone,

We think that it would be really useful to have the option to see “Median Time from First Step” and “Median Time from Previous Step” in our funnel reports. We would even dare to say that this would be more useful to have then the currently shown “Average Time from First Step” and “Average Time from Previous Step” due to the nature of our products and our users.

Hi @dorian_cudina,

Thank you for your post. Could you share some more context into median being more useful than average for your usecases?


Hi @Chandu, sure;

Since some of our products are available in Self Service mode, various types of users with various reasons and motivations are signing up for them.

For example, some of the users end up creating an account because they didn’t quite understand what are we offering, while others just want to check out how does this product look like - these are the users which will make a few clicks inside the product and maybe log in again in a week or in a month. There are also users that will create an account, but they will then go to lunch and forget about exploring the product for several hours or even for a day or two.

Now, seeing only the “Average time between steps” (doesn’t matter if the steps are some closely related critical events or onboarding steps or something else) isn’t so relevant to consider when exploring the behavior of users which entered the product with the real intent of using it right away. For example, let’s imagine that there is 80% of such users and 20% of users described above. The first group of users takes 20 minutes on average to perform two critical and related actions, while for the second group it takes 10 hours on average to do the same. In this situation, I would get a lot more useful information by seeing the “Median time between steps” instead of the “Average time between steps”.


Hey @Chandu can you please share with us if there has been any development related to this idea?


Hi @dorian_cudina,
Firstly, thank you for explaining the use case in detail. This is added to our backlog and will be prioritized appropriately. Will keep you posted on this request. 
 


No StatusAcknowledged

Also upvoting. For funnels with longer timeouts, several long outliers can drastically skew the data. The median would be more resilient against these effects.
Honestly, would be nice to see a histogram of the data… but maybe that’s another idea post :)