Mobile App Retention: The Critical Difference Developers Miss

the most common metric is day one retention it's your first signal one of the first signals that helps you identify if that's a good source or not and in my opinion it's better when the number uh the retention number is conservative and because the absolute is always higher than relative I would always be relative retention hi everyone welcome to another edition of tension ay11 this is Roman and today I'm joined by Jaz from our product team hey guys I am Jaz um or JRE but Everyone likes to call me Jazz I'm a senior product manager at tenin and I happy to be here today Roman and it is exciting topic we'll talk about absolute relative absolute and relative cohorts

or absolute and relative retention most of the mobile Publishers don't know the difference they see a retention and they assume they know what it is but there is a huge difference and once you know the difference you do much smarter business decisions whether it's published user acquisition or just analytics let's just jump into it uh We've prepared some beautiful slides uh towards the end JZ will also show us how it works in Tangent so here is a absolute cohort uh just you want to take a step of what it is sure absolute cohort um or you can call it a calendar under date cohort so before I go into

that I can talk a little bit about what a cohort is so when we talk about cohort especially at tenine we mean um when a user installs the app and then we build cohorts based on that now when we say absolute cohort we say um when did a user install the app and by absolute we mean calendar date um and in our tangent terminology we use um UTC time zone right now so let me give you an example so let's say that you know your UTC time ends at 7 p.m. local time as just as an example and if um a user in your app installs the app before 6:00 p.m. or 7 p.m. um then they are day Zero or when they do it after that um um they will be day um zero will start after 7 p.m. and

then this changes you know the retention also a little bit um so the the way retention changes here is um when the user comes back let's say the user installed the app at on 600 p.m. so they became a day Zero user and then they come back at 800m so they're retention um they're a retained user on day one so I I just talked a lot about that but in simpler terms absolute is just a description for um the way we cohort and the strategy we use there is just calendar date yes and as you can see here like even if a user was acquired on a specific day like like let's say even 11:59 and he comes back 2 minutes later at 1201 he'll be considered to be retained

if we're talking about retention cohort right yes and this is sort of a false positive signal whether you're doing analytics or you're doing UA because uh I I'm not sure if this user is really retain right you just came back in two minutes yeah 100% that that that makes sense but sometimes if you want to do more of a calendar date um analytics and for other use cases that we can talk about later then maybe using absolute cohort might make sense for your use case yeah yeah yeah absolutely there are no like good or bad retention there are different cases right and again yeah we we have a slide for that too yes let's and another one is a relative one yes so I talked about what a cohort is and you know it is basically an installation date

that's what our cohorts are based off and absolute is calendar date while relative describes um a 2 24hour period so when we say relative it means relative to their installation date so let's say Roman you install an app right um so you install an app on I don't know at 2 p.m your time so it doesn't matter if it is a UTC time zone or um Us New York time zone um it is relative to your installation time so that's what we mean by relative and it is um 24 hours so what we mean by that is you install the app at 100 pm. and then you need to come back you know um in 24 hours to be called as a retained user so that's the core difference between what a relative cohort is versus an absolute cohort where we use a UT UTC time zone

makes sense um now let's see the different uh use cases right so for the relative I can start just based on the my experience in UA uh the most common metric is day one retention it's your first signal or one of the first signal that helps you identify if that's a good source or not and in my opinion it's better when the number uh the retention number is conservative and because the absolute is always higher than relative I would always speak relative retention so that's the T yeah I agree with that if you want to analyze your users and if you want to run campaigns and do a lot of like us user based

cohort analysis and you want you know minimum disruptions um from different time zones so using relative retention will give you the best answers to how your users are performing and how your campaigns are working so I agree with that now to the comparing with analytics providers and for like I I I think the disclaimer here is that we're using this terminology absolute versus relative but as analytics providers might use different uh labels for that basically so my advice is always to check what sort of methodology they use before you compare the data yeah and uh based on our research uh we've seen that most of the analytics providers so for example Firebase is using absolute ret

right and uh yeah you you you just once you know that you just save yourself a lot of time when you start comparing the data between Firebase and Tangen or Firebase or any other MMP for example right 100% so that's the biggest use case also of one of the biggest use cases of using absolute cohort strategy if you want if your use case is to compare different analytics provider or different you know Data Systems in general so using the absolute cohort strategy will be the best for you because it will give you similar results and you can have a comparison because a lot of these analytics providers work on a calendar date like analytics um side of thing they care more about calendar date um numbers um which is why if your use cases to compare you know different providers um

then yes absolutely use absolute cohort and the third use case is when you're working with Publisher make sure that you know what R type of retention they use and uh make sure that you are reporting on the like retention that is basically like higher right it's in your interest and that's again absolute is always higher than relative retention I've seen so many cases when uh developer starts working with a publisher and then they see that the numbers are different and they have no idea what it is and they think there is some like technical issue with it but in reality it's a different methodology again right yeah 100% I think the the first you know the initial days of your

campaigns or the initial cohorts um relative is lower um right than absolute um so but as you go over time the numbers sort of will match like your more older cohorts um but yeah Roman I do agree with you you know you need to know that for sure yeah very good point so this is mostly again guys day one retention day seven it's going to be are kind of the same but you want to find the best users as soon as possible and spend as little money as possible on not so good users so this is uh why we think it's important and this is why we're doing this video B yeah and and Roman and we know what our users want right like we know our users have different use cases for a you know running their campaigns um working with Publishers and comparing

with different analytics providers which is why we offer you both these retention cohorts or cohort strategies that you can basically toggle on our Dash show us Jess yes I will I'll stop sharing uh while just preparing to share the screen give us a like if you learned something new today leave us a comment if you have any comments about these things or you want us to cover more topics like that and subscribe for the next episodes uhuh I see you see it awesome so this is our demo account and um this is our beautiful Dash you can go under my account you can go under manage user so remember the cohort strategy is you know for your user so it's a preference that is set at the user level so it's

not an orc level setting so every user can pick their own strategy and again like Roman said these are different use cases right maybe a UA manager wants to use relative maybe a Dev wants to use um absolute or UTC that we call so you can pick your own um toggle on our dash for your own user and and it's very simple you pick cohort strategy you pick relative or use UTC so relative is the 24 hours that I talked about UTC is just absolute and then you can read you know all the cool details that describes the differences between relative and UTC time zones and then you just hit update and all done and then you just go back to all reports and um everything will now all your cohorts not just retention but Roi LTV roas all your cohorts will now use the cohort strategy that you U picked under my yeah and the coolest

thing and I was kind of surprised that it all calculates on the fly so switch uh absolute one numbers switch back other numbers nothing is lost you can always uh find the best approach and don't lose the data yeah and have something for yourself and you see Roman how fast this was this is and I do want to talk a little bit bit about this uh this is because of our really cool Tech Team um and I don't know if you know um the the users who are watching this they probably have seen amazing improvements on how fast our Dash has become and that's because of this re architecture uh that we did a while ago that leverages click house and we are dynamically generating queries you know with aggregated tables that are immutable and then they are able to pick different strategies and run different

queries on the fly so it is super super fast and you know with very minimal changes so we were able to make this change happen with like some really cool rear Innovative re architecture that we did with our um tech systems and the features are so easy to release now like uh our December newsletter like if you're not subscrib subscribe but it add so many new features yeah yeah basically with this you know we helped made you your cohort strategy simple yeah that's a good end for this episode uh we have another video uh about the retention we invited like a game designer and he explained everything that you need to know about the retention in general uh we'll link it somewhere here in this video so you can drill down into the this Rabbit Hole

of retention if you want um thanks a lot J thank you Roman for having me have a good one cheers youo bye bye guys [Music]