Khoros Atlas Logo

%3CLINGO-SUB%20id%3D%22lingo-sub-556924%22%20slang%3D%22en-US%22%3ECalculating%20WAU%2C%20retention%2C%20and%20churn%3C%2FLINGO-SUB%3E%3CLINGO-BODY%20id%3D%22lingo-body-556924%22%20slang%3D%22en-US%22%3E%3CP%3EI'd%20like%20to%20calculate%20WAU%20for%20my%20community%2C%20but%20unfortunately%20I'm%20struggling%20to%20do%20that%20with%20the%20Khoros%20data%20not%20matching%20itself.%20I%20would%20love%20to%20know%20if%20others%20of%20you%20are%20calculating%20the%20follow%3C%2FP%3E%3CP%3E%3CSTRONG%3EWeekly%20Active%20Users%3C%2FSTRONG%3E%20-%20Total%20number%20of%20users%20who%20visited%20the%20community%26nbsp%3B%3C%2FP%3E%3CP%3E%3CSTRONG%3ENew%20Users%3C%2FSTRONG%3E%20-%20Number%20of%20users%20who%20visited%20the%20community%20for%20the%20first%20time.%3C%2FP%3E%3CP%3E%3CSTRONG%3EReturning%20Users%3C%2FSTRONG%3E%20-%20Users%20who%20visited%20the%20community%20this%20week%2C%20not%20for%20the%20first%20time%20and%20not%20having%20churned%20in%20the%20past%20week%3C%2FP%3E%3CP%3E%3CSTRONG%3EChurned%20Users%3C%2FSTRONG%3E%20-%20Users%20who%20have%20not%20visited%20this%20week%3C%2FP%3E%3CP%3E%3CSTRONG%3EResurrected%20Users%3C%2FSTRONG%3E%20-%20Users%20who%20have%20visited%20after%20previously%20churning%20the%20week%20before.%3C%2FP%3E%3CHR%20%2F%3E%3CP%3EAs%20a%20note%2C%20I%20have%20done%20some%20exploring%2C%20but%20haven't%20found%20a%20great%20way%20to%20calculate%20these.%20When%20I%20look%20at%20overall%20metrics%20it%20just%20tells%20me%20the%20number%20of%20unique%20visitors%20(so%20WAU)%2C%20but%20not%20on%20an%20individual%20basis%20so%20there's%20no%20data%20that%20points%20to%20who%20visited%20each%20week%20and%20what%20retention%20looked%20like.%3C%2FP%3E%3CP%3E%3CBR%20%2F%3EI%20have%20also%20pulled%20the%20user%20reports%20where%20the%20number%26nbsp%3Bof%20users%20pulled%20for%20a%20week%20don't%20match%20what's%20in%20the%20community%20analytics.%26nbsp%3B%20In%20addition%20to%20that%2C%20when%20I%20look%20at%20the%20user%20report%20I%20will%20have%20instances%20where%20a%20user's%20most%20recent%20sign%20in%20is%20on%20Dec%203%2C%20but%20when%20I%20pull%20data%20for%20the%20week%20of%20Dec%201-%20Dec%207%20it%20says%20that%20user%20has%200%20total%20sign%20ins.%26nbsp%3B%3C%2FP%3E%3CP%3EI'm%20not%20sure%20why%20this%20doesn't%20match%20up%2C%20but%20am%20hoping%20someone%20could%20help%20me%20here%20and%20explain%20why%20these%20don't%20equal.%26nbsp%3B%3C%2FP%3E%3C%2FLINGO-BODY%3E%3CLINGO-SUB%20id%3D%22lingo-sub-556972%22%20slang%3D%22en-US%22%3ERe%3A%20Calculating%20WAU%2C%20retention%2C%20and%20churn%3C%2FLINGO-SUB%3E%3CLINGO-BODY%20id%3D%22lingo-body-556972%22%20slang%3D%22en-US%22%3E%3CP%3EThere%20could%20be%20all%20kinds%20of%20things%20going%20on%2C%20so%20I'll%20start%20backwards.%26nbsp%3B%20Keep%20in%20mind%20Khoros%20data%20isn't%20real-time%2C%20any%20reporting%20you%20pull%20is%20generally%2024-48%20hours%20behind%2C%20so%20if%20you%20are%20attempting%20to%20pull%20it%20same%20day%2Fnext%20day%2C%20it%20likely%20won't%20be%20accurate.%3CBR%20%2F%3E%3CBR%20%2F%3ENext%20up%2C%20Analytics%2FLSI%20isn't%20the%20same%20dataset%20as%20what%20you%20get%20out%20of%20the%20Admin%2FUser%20reports.%26nbsp%3B%20It's%20a%20pain%20point%2C%20but%20ultimately%20it%20is%20what%20it%20is%2C%20they%20have%20different%20filtering%20in%20place%20before%20the%20end%20data%20ultimately%20gets%20to%20us.%26nbsp%3B%26nbsp%3B%3CBR%20%2F%3E%3CBR%20%2F%3EIf%20after%20a%20few%20days%20you%20still%20have%20the%20inconsistency%20you%20see%20for%20user%20data%2C%20I%20would%20suggest%20opening%20a%20ticket%2C%20it%20could%20be%20something%20off%20that%20they%20need%20to%20take%20a%20look%20at.%3CBR%20%2F%3E%3CBR%20%2F%3EAs%20for%20the%20metrics%2C%20you%20are%20going%20to%20be%20stuck%20manually%20calculating%20these%20yourself%20in%20excel%20each%20week%2C%20having%20to%20maintain%20a%20user%20dataset%20to%20compare.%26nbsp%3B%20Generally%20speaking%2C%20export%20the%20user%20report%20each%20week%2C%20compare%20the%20differences%2C%20and%20build%20out%20a%20summary%20of%20the%20data%20using%20that%20data%20set%20to%20keep%20all%20your%20user%20data%20aligned%3CBR%20%2F%3E%3CBR%20%2F%3E%3C%2FP%3E%3CP%3E%3CSTRONG%3EWeekly%20Active%20Users%3C%2FSTRONG%3E%3CSPAN%3E%26nbsp%3B%3C%2FSPAN%3E-%20Export%20the%20user%20report%20for%20those%20who%20logged%20in%2C%20count%20it%20up.%3C%2FP%3E%3CP%3E%3CSTRONG%3ENew%20Users%3C%2FSTRONG%3E%3CSPAN%3E%26nbsp%3B%3C%2FSPAN%3E-%20User%20report%2C%20count%20up%20users%20based%20on%20registration%20date%20-%20I%20think%20LSI%20also%20shows%20this%20though.%3C%2FP%3E%3CP%3E%3CSTRONG%3EReturning%20Users%3C%2FSTRONG%3E%3CSPAN%3E%26nbsp%3B%3C%2FSPAN%3E-%20This%20is%20something%20you%20would%20have%20to%20calculate%20in%20excel%20yourself%20each%20week%2C%20comparing%20which%20users%20previously%20visited%20vs.%20the%20new%20week%3C%2FP%3E%3CP%3E%3CSTRONG%3EChurned%20Users%3C%2FSTRONG%3E%3CSPAN%3E%26nbsp%3B%3C%2FSPAN%3E-%20Ditto%2C%20manually%20calculate%20in%20excel%3C%2FP%3E%3CP%3E%3CSTRONG%3EResurrected%20Users%3C%2FSTRONG%3E%3CSPAN%3E%26nbsp%3B%3C%2FSPAN%3E-%20Ditto%2C%20manually%20calculate%20in%20excel%3C%2FP%3E%3C%2FLINGO-BODY%3E
New Commentator

Calculating WAU, retention, and churn

I'd like to calculate WAU for my community, but unfortunately I'm struggling to do that with the Khoros data not matching itself. I would love to know if others of you are calculating the follow

Weekly Active Users - Total number of users who visited the community 

New Users - Number of users who visited the community for the first time.

Returning Users - Users who visited the community this week, not for the first time and not having churned in the past week

Churned Users - Users who have not visited this week

Resurrected Users - Users who have visited after previously churning the week before.


As a note, I have done some exploring, but haven't found a great way to calculate these. When I look at overall metrics it just tells me the number of unique visitors (so WAU), but not on an individual basis so there's no data that points to who visited each week and what retention looked like.


I have also pulled the user reports where the number of users pulled for a week don't match what's in the community analytics.  In addition to that, when I look at the user report I will have instances where a user's most recent sign in is on Dec 3, but when I pull data for the week of Dec 1- Dec 7 it says that user has 0 total sign ins. 

I'm not sure why this doesn't match up, but am hoping someone could help me here and explain why these don't equal. 

1 Reply 1
Highlighted
Esteemed Contributor
Esteemed Contributor

Re: Calculating WAU, retention, and churn

There could be all kinds of things going on, so I'll start backwards.  Keep in mind Khoros data isn't real-time, any reporting you pull is generally 24-48 hours behind, so if you are attempting to pull it same day/next day, it likely won't be accurate.

Next up, Analytics/LSI isn't the same dataset as what you get out of the Admin/User reports.  It's a pain point, but ultimately it is what it is, they have different filtering in place before the end data ultimately gets to us.  

If after a few days you still have the inconsistency you see for user data, I would suggest opening a ticket, it could be something off that they need to take a look at.

As for the metrics, you are going to be stuck manually calculating these yourself in excel each week, having to maintain a user dataset to compare.  Generally speaking, export the user report each week, compare the differences, and build out a summary of the data using that data set to keep all your user data aligned

Weekly Active Users - Export the user report for those who logged in, count it up.

New Users - User report, count up users based on registration date - I think LSI also shows this though.

Returning Users - This is something you would have to calculate in excel yourself each week, comparing which users previously visited vs. the new week

Churned Users - Ditto, manually calculate in excel

Resurrected Users - Ditto, manually calculate in excel

0 Kudos
Reply
Loading...