What's the difference between "User Visits" and "Active Users"


(Senthil Kumaran) #1

I know that Facebook/Twitter/ and possibly others tally you into their Monthly Active User Metric, if you logged in to their platform at-least once in the last 30 days.

In the admin dashboard, we have “User Visits” which is 1.1k for the last 30 days, and in /about page, our last 30 day active user count is 260.

I’d like to know the difference (in simple terms) between “User Visits” and “Active User” measure.

Thank you,
Senthil


#2

A user visit is someone visiting your site while logged in.

An active user is someone that has performed an action (posted or liked a post).


(Senthil Kumaran) #3

Thank you!

We have “user visits” greater than the number of registered members in the site. I presume the user-visit based on session? That is, if the same user logs-out and logs-in, he/she is counted twice?

It seems that we only check if the user visited within last 30 days?
Ref: discourse/about.rb at master · discourse/discourse · GitHub

So, I am still wondering about the question of “User visits” and “Active Users”.

Thank you,
Senthil


#4

It doesn’t state that they are unique visits so (without reviewing the code) I’d assume every visit is counted.

You had logged in visitors come to your site 1.1k times in the last 30 days. 260 of those posted or liked something.


(Senthil Kumaran) #5

Hi @HAWK, by pointing to github code, I was disputing the point that " Active Users are those who posted/liked something"


(Jeff Atwood) #6

True. I think @hawk was covering the more general meaning, but in our code it means a user who visited the site as a logged in account. You can look at your post and like counts to get an idea of how many actions users are taking per week. There are also reading stats for logged in users.


#7

I was – hence the ‘without looking at the code’ clarification. Apologies for the ambiguity @Senthil .

This raises a good point though. Should we be using commonly understood terms in a not commonly understood way?


(Jeff Atwood) #8

Well if you look at the code quoted “Users 30 days” is new users who signed up in the last 30 days.

I would say visiting as a logged in user and reading topics is indeed an active user. Reading is fundamental.


#9

I agree that reading is fundamental but we’re the only platform afaik that defines the term in this way which is at best ambiguous.


(Jeff Atwood) #10

Yes well that’s why Discourse is awesome and better than the other platforms :wink:

(That said, we would indeed have to add a bit of reading to the mix to get that, right now it’s just user visited while logged on.)


#11

I don’t think I’m explaining myself well. What I’m suggesting is that we use different terminology to define ‘people that read but don’t engage in any other way’ to differentiate them from what is commonly known as ‘active members’.

I totally agree that the reading stat is valid and that those people are fundamentally important to community survival, but the fact that all other platforms that I’m aware of define ‘active members’ differently to us is confusing.


(Dean Taylor) #12

There is already a specific dictionary term for this: “Lurker”, the current definition is:

A person who lurks, in particular a user of an Internet message board or chat room who does not participate.


From the now very old Oxford English Dictionary Twelfth edition 2011:

image

The original pre-internet meaning (1) always makes me chuckle because it’s still relevant today online.


There is also the alternative term “Observer” for those fearful of the negative connotations surrounding “Lurker”.


(Alex Armstrong) #13

I’m trying to make heads or tails of the stats in the dashboard and came across this discussion. After reading it, I’m still none the wiser.

Is there a way to get statistics for:

  • The number of unique users who visited the site in a month (i.e., visitors, not visits)
  • The number of unique users who visited the site in a month and interacted with it in some way beyond reading

#14

You can get that from GA.

I get this by downloading all users into a spreadsheet, removing all but the current month, and then sorting by like and post counts. It’s a bit cumbersome but it works.

The new admin dashboard that I’ve designed will include this. We don’t have a hard delivery date on that but we’re hoping to get it underway soon.


(Alex Armstrong) #15

I’m not sure how. My community is private and therefore has no anonymous users. GA’s visitors are useful but may indicate the same user, just on different devices (since GA has no conception of Discourse’s users).

The other suggestion is a bit cumbersome, but Good Enough For Now™ :slight_smile:

Edit: On second thought, I think the other idea doesn’t work either. The like/post count values are for that a user’s entire activity. He may have a very high number in general, but none in that particular month. Or am I missing something?


#16

Ah yes, that’s fair. How many people do you think use multiple devices though? Enough to seriously corrupt the numbers?

Actually you’re right. It only gives a count of new members that have been active.

I have a data explorer query for the other. Do you have that plugin installed?


(Kane York) #17

This is what the /about page 30 day active user count is. It counts the number of accounts who logged in within the last month.


#18

That’s only helpful if you get it on the day though.
There are no custom date ranges so if you’re reporting for the past month you have a one day window.


(Alex Armstrong) #19

What @hawk said :slight_smile:

I’m not sure, but part of the community is logged in both at their work computers and on their personal devices. (Be they mobile or otherwise.) We’re a professional development non-profit. Some stuff they clearly see as work, and some they do after hours.

And yes, I have Data Explorer.


(Alex Armstrong) #20

I’m still trying to do this:

Any help would be appreciated.

@HAWK, you mentioned a Data Exporer query for the second one? Is there something similar for the former one? I wouldn’t mind running it manually for a every month – because he GA statistics seem all but useless to me.