LP Login

Think Big. Move Fast.

Growth without retention leads to the dreaded sharksfin curve. If you’re not retaining your users, then it doesn’t matter how quickly you add new users to your product. You have a leaky bucket.  Sooner or later, when you saturate your market of users and your new user growth slows, the rate of your existing user loss will outpace your new user growth and you will see overall usage fall.

Conversely, as long as you have good retention, if you add good growth you will rapidly see overall usage increase.

The problem with measuring and understanding retention is that by definition it takes time to measure. You have to watch multiple user cohorts over manye months to be able to really understand who users will use your site or app over time.

As a proxy for retention, you can measure engagement instead. If a user is going to stay a user for a long time, you need to create habituation, and habituation comes from repetition. So the more often someone uses your product, the more likely they will be to stick around. This isn’t a perfect predictor, but it is pretty good within a category. So within social games for example, all else equal, a game with a higher DAU/MAU ratio will have greater retention than a game with a lower DAU/MAU ratio. But a game with a higher DAU/MAU ratio may or may not have greater retention than a communications app or a media app with a lower DAU/MAU ratio. Because usage of apps varies by category, you need to be careful of comparing across categories.

This is one of the reasons that Facebook watches the “6 of 7″ day metric very closely. This is the proportion of total users that have used your product 6 or more days out of the last 7. Clearly, highly engaged users. For a new startup if you were above 20% on this metric you would be very, very happy.

I always look at engagement for fast growing social apps to see if the growth is underpinned by strong retention. When we invested in Snapchat in the seed round and Whisper in the A round, we saw extremely strong engagement in both cases, and both companies have seen tremendous growth since then.

 

  • Ranah Edelin

    Great post Jeremy, definitely agree! What do you think healthy DAU / MAU ratios are for various types of services (media, games, communications, etc)? I think it’ll be interesting to see how those rates differ for web-based services vs mobile services too.

  • plus8star

    Nice piece.

    I’d recommend also looking into the “layered cake” of weekly or monthly cohort contributions to daily DAU and revenue.

    To be really clean you might need to normalize by using $/cohort size and DAU/cohort size.

  • Mads Aarøe Mathiesen

    Great post. We have just (soft)launched a classifieds site in a confined 75K inhabitants geography on the other side of the country and have attracted 2.1K unique users in less than 3 weeks of which 1.1K have registered (fb login). The DAU/MAU is +20%, also when discounting new users in DAU. We THINK it looks promising, but basically haven’t got a clue if we’ve found a good product/market fit…can you help out? Any sort of bench marks we should look at?

  • Mads Aarøe Mathiesen

    Great post. We have just (soft)launched a classifieds site in a confined 75K inhabitants geography on the other side of the country and have attracted 2.1K unique users in less than 3 weeks of which 1.1K have registered (fb login). The DAU/MAU is +20%, also when discounting new users in DAU. We THINK it looks promising, but basically haven’t got a clue if we’ve found a good product/market fit…can you help out? Any sort of bench marks we should look at?