4 ms·
Ask HN: How do you calculate the # of active users for your web app?
This question came up in a conversation recently and we couldn't agree on a good way to do it.
For example, let's say you launched an app six months ago and have had 5000 sign ups and are growing at a rate of 100 new users/day.
One way to calculate the number of active users would be to count the number of users who have performed a certain action within the last month. The problem is that that also includes everyone who signed up in the last month, many of whom will never use your service once and never come back. It doesn't seem correct to count them towards your active users.
Another way would be to look at all of the users who signed up more than a month ago and who have used your application in the last month, but that seems more complicated than it needs to be.
How do you do it? How do you calculate the number of active users for your web app?
- _ud4a 15y agoi would calculate the difference between their registration date and their last login. then i would add an extra criteria to make sure out of those the registration date is at least a month old so you would get the number of active users in the past month who are not new registers. it's a pretty simple sql query.
- spullara 15y agoWe calculate monthly actives as those who returned this month who didn't sign up this month. We are also now tracking cohorts of users who joined in certain week to see how many of them come back week 1, 2, 3, etc. That will let us see if engagement features we've implemented are working.
- neuromancer2600 15y agoOnce you simplify your second option it becomes actually pretty easy: count the number of people who logged in sometime in the last 30 days and subtract from that the number of people who were signing up during that period. It should give you some sense how people are using your app at least 30 days into it. That would be the most basic MAU (monthly active users). I guess it boils down to what your goal is. Stickiness (repeatable action) is something very important to understand what really drives interaction. So, going in with some hypothesis on what you want users to do and then figure out if that actually holds true is a vital part at all stages.
- namank 15y agoYou could start logging the IP of every client and do some analysis based off of that. Should paint a fairly accurate picture.