Customer segmentation and the need for good data

By February 1, 2013Consumer Internet

There was a good post on Gamesbrief a couple of days back about customer segmentation – i.e. dividing the customer base into groups based on behavioural characteristics and targeting each group with different communication campaigns. The article starts by saying that most game developers intuitively understand that player segmentation is important, and goes on to say that getting started can be daunting, but is actually easier than many people realise. My immediate thought was that these points apply to consumer internet generallly, not just games, and that consumer internet developers may be further back on the learning curve than games developers.

Our former portfolio comapny Lovefilm had a very effective customer segmentation programme which divided customer into groups depending on how engaged they were with the service – added no films to the list, added 1 to X films to their list, added more than X films to their list, regularly added new films to their list, and so on – and each of these groups got different emails designed to push them on to the next level of engagement with the overall goal of reducing churn. I haven’t seen many consumer internet companies get as sophisticated as that, whilst most either don’t do anything or maybe have separate emails for newbies (a set of welcome emails), and/or run the odd campaign designed to reactivate dormant members.

Some good use cases for segmented messages:

  • Welcome programmes which help users get familiar with the basics of a service – starts with the sign up flow, should increase the percentage of new users returning in the first week or two
  • Tips for more experienced users – should increase engagement amongst the heavy users and inrease downstream retention
  • Gifts for the best customers
  • Reactivation campaigns

Like most areas of consumer internet execution getting this right isn’t rocket science, but it does require good data, clear thinking, and hard work over an extended period of time to design the campaigns, measure their effectiveness, and then optimise. Reading this post back, the requirement for good data stands out as the biggest stumbling block for most companies. Good data pays dividends in so many areas and if I had one piece of advice it would be to bear than in mind from day one. Too many companies end up with services that are having some early success but struggle to optimise because it is too difficult to get good data out and they don’t have the time and/or resources to re-write their code and fix the problem.