The End Of The Install Game

You have to hand it to the people over at Facebook. They pay attention to what’s going on in their community and when something is out of hand, they fix it.

Everyone knows that the race to get the most installs had some bad effects. It led to application spamming, cost per install advertising, and the net result was lots of installs and little use.

So Facebook made some changes recently to the way they rank applications when you browse the service for new ones. Installs no longer matter. Usage, real usage, matters. Here’s the default tab when you start browsing for apps (the best real estate you can get).


Top Friends is still on top, but their 13mm installs don’t matter anymore, its just the 2.7mm people that actually used the Top Friends app recently that counts. As it should be.

Back in early July, I wrote a post comparing Facebook stats to FeedBurner stats. In that post, I said:

But Appsaholic needs more data to be truly useful. Like FeedBurner
did, Appsaholic needs to get beyond the "subscriber/user" number and
get into what is actually getting used.

In more good news, Appsaholic has adopted the new methodology too. So focusing on installs is over. The game is now daily usage. And that’s exactly what every really cares about at the end of the day.

I gotta hand it to Facebook. It seems that they are doing most things right these days, but most importantly, when they see something is wrong, they fix it. And that’s how you build a great company.

#VC & Technology

Comments (Archived):

  1. Uri L.

    Yet another great example why implicit data is always more valuable and crucial for sustainability.

  2. Steven Kane

    i noticed this change. big applause!but, how do they measure “daily active users”?today i added “iRead” app and added a few books. it “loads” on my profile page every time i open my profile page. does that count fo iRead? or only if i add another book?i looked around but couldnt find any explanation of how FB is calculating stats…

  3. Tony Wright

    Hrm… Thank goodness for this. But this seems to favor applications that require ongoing interaction rather than “chrome” FB apps– apps that are designed to enhance your profile but don’t require (or benefit from) daily interaction. Still, it’s very nice to see.

    1. narnia

      I concur.

  4. ziomek

    too much of that fb applications crap everywhere…

  5. mcnodoze

    It’s a very good, much-needed first shift, but much more needs to come (btw, Adonomics just launched their analytics; I haven’t seen it yet, but I imagine it will be Google Analytics-like, minus the polish initially).The problem is still what Steve was asking–how they measure “use” is very important. It’s defined in the developers group on Facebook for those who are interested, and it has interesting implications for how apps will be designed to game the metric (WGMGD–What Gets Measured Gets Done has been thrown around the developers board a lot recently).Right now, DAU can still be gamed by growth; if I have 1,000 people adding my app everyday, those users count in the DAU. So as long as my app keeps growing (and everyone is still in growth mode), their DAU will be high relative to others. Separating out new user usage vs. established user usage will prob be the next step–that is where you will see a bigger difference between Zombies and iLike.

  6. Jeremiah Owyang

    I think they got their terms of the measurement attributes confused, as they’re really measuring Interaction, not engagement.details here:http://www.web-strategist.c

  7. Mark

    good post (that I just posted on my tumblelog).

  8. Mark

    I understand that maybe the way Facebook hd it wasn’t perfect, but it was MUCH easier to understand (how the apps were ranked).