Customer Focus

The Facebook fracas of last week is a cautionary tale on customer relations. On the surface, Facebook is successful with a loyal user base, with many millions of users, primarily students, preferring its network to other options. And, it had the ostensibly unbeatable benefit of a founder CEO who is a user too - he built something that he knew he and his friends could use when in college.

A sketchy recap on the issue. On the face of it (sorry!), it seems likely and logical that a user would want to know about updates to profiles of friends on his network automatically without having to seek them, and that the said friends wouldn't mind. But vast numbers of users got agitated, they didn't like their updates triggering auto feeds and felt it was a violation of privacy, much to Facebook's surprise. Techcrunch had a post supporting the usefulness of the new feature, noting all that people had to do if they didn't want the updates distributed was to set the privacy controls appropriately. That didn't soften the user response. The CEO, Mark Zuckerberg, put out a plea for calm, the users' concerns were being heard, but that didn't lessen the noise - nothing made a difference until the message went out that privacy controls will be enhanced and the feature tweaked to let the user determine what should be shared or not. Facebook received many kudos for having listened to the users, and being a role model for startups.

It is not surprising that Facebook listened to its users - in this kind of business, there is no alternative (though that could arguably be true for all businesses). The cautionary part is how easy it is to think you know your customer/user and be totally wrong. This is particularly true when you think you know how a user will react because hey, you're one yourself. The reality is that you (and your design staff) are not typical users. You are informed (ok, tainted) by the business you're building and the roles you're playing in building it, and will not have the same perspective or experiences of 'real' users. And somehow 'logical' assumptions on user behavior seem singularly fragile.

In the early stages of a startup, extrapolating user behavior is standard ops as there's not enough bandwidth to run every feature by a focus group. Making assumptions is unavoidable, but they must be acted upon with caution. Internal tests are good for bug testing, not for gauging customer response. Embrace the beta release instead and actively solicit feedback. And if something slips past your vigilance and kindles user outrage, don't try to explain your position - publish a fast mea culpa and retreat. You are not the customer, and the customer is always right.