Sunday, June 08, 2008

discussion: adoption, utilization and integration

How many of you create products that are completely transparent?

By transparent I mean do you really know:

  • How many installs are there of your product?
  • How many total users and active users are there of your product?
  • How are each of your products rated by your users?
  • How far does each user get through the user experience?
  • How many users take advantage of click-through links?
  • How many users complete surveys?
  • How many users are "repeat" users?
  • How many users loop in "friends and family"?

You get the idea.   

When you pile up all of this data and look at it from a great height, you see some patterns emerge.  There are essentially only three key concepts at play here: adoption, utilization and integration.  

Adoption speaks to how many of your target users have made the choice to use your product in lieu of other alternatives.  

Utilization speaks to how often (discrete usages), long (length of usage) and well (scope of usage across all feature sets) your product is used. 

Integration speaks to how well your product receives and delivers hand-offs to other elements of your user's workflow.  This includes stuff like API support but also UI integration, but goes on to cover things like datatype  and metadata consistency, and support for consolidated utilization reporting across multiple products.

Think of the metrics associated with these concepts as evidence of whether or not your customers are getting "full value" from your offering.  They're also sources of leading indicators of where your products need to be "tweaked" to maximize this value.  

When you look at the interplay between these three metrics, you can have some very interesting conversations.  I also like to use these concepts when I evaluate new products or propose enhancements to existing ones, because they are customer-centric metrics that matter and can be used to fuel ROI discussions.

If you've not built your product to throw off data, you're guessing at this stuff.  So next time you sit down with your feature backlog or sharpen your pencil for a new PRD, consider features needed to instrument your product so you can report on these measures.  Sales will thank you, and your line product managers will curse you (but privately thank you too).

No comments: