Here's an article I wrote explaining a few of the reasons customers should care about the new community services we've built.

Microsoft should care, and does, for a variety of reasons. Most of them having to do with customers being happier, better connected, and more productive. But there are other reasons as well and below we discuss one of them -- one that is not generally considered.

Outrageous claim: if Microsoft customers were connected in rich social networks our support costs would drop dramatically.

Here's why:

  • Social networks are also trusted communications channels -- cheap, effective, and fast communications networks
  • When customers are connected in an online social network (actually online or offline, but we're focusing here on the online experience) and one customer has a problem, they all know about it when it happens, and they all know about it when it gets fixed.
  • A customer only needs to call support, or visit a support forum, if their social network has failed.
  • Therefore, support calls drop dramatically through social network connection

Consider the following illustration:

This bell curve represents a typical product support incident call pattern. Here's the scenario. At point "B" a problem is discovered in a newly released product and the first customer support call is registered. As the product is adopted, the problem effects a growing number of customers and the call volumes increase. At point "C" a fix is found. Nevertheless, the call volume continues up for some time before leveling off and eventually trailing off. That is the case today.

The next illustration compares call volumes when social networks are fully realized (a theoretical best case):

In this case, shortly after point "B" the potentially effected social network is notified of the problem. Word spreads, but not necessarily fast enough to immediately curtail the increase in call volume. However, as word spreads (and communications across on-line weak-tie networks can be very fast) everyone -- ideally -- in the potentially effected community knows about the problem. Call volumes drop off because there is no point to calling. When the solution is found, the network will effectively spread word of the solution. The area within the curve and between the dashed lines represents the cost of social network failure in terms of call volume.

Overstated? Clearly. The extreme case was illustrated to make the point. How extreme is it? Hmm. No good way to know. Consider this, satisfaction rates among opensource users seems quite high, and social networks are a principle means by which information is shared across that community.