Google’s Service Problem

January 15, 2010

The NYT recently described Google’s customer service supporting its new Android phone.  The article described how customers can’t call Google for help and that it may take up to 48 hours for the company to respond to email messages.  The article referenced one customer who has been trying for a week to talk to a live person. Still no luck getting through.

This article made the front page of the NYT business section because it’s a story about Google, not because it’s a particularly rare or surprising service phenomenon.  When a service model designed for one type of customer needs (e.g., using web-based adwords) is then used to serve an entirely different set of needs (e.g., after-sales service for a new type of phone), there is predictable turmoil.

The interesting part will be to watch how Google responds to the realization that it’s trying to meet wildly divergent service needs — what I call distinct operating segments — with a single operating model.  In my experience, the company has some choices to make if it wants to deliver excellent or even adequate service.  Option A is to serve one operating segment well and essentially ignore the other.  That’s not as bad as it sounds.  Southwest Airlines optimizes its service model for low-maintenance travelers. When high-maintenance travelers come along, it doesn’t turn them away, but it also doesn’t work very hard to accommodate them.  Southwest serves one segment with excellence and asks everyone else to adjust.  This is the definition of focused service.

Option B is to create a distinct service model for each operating segment.  If Southwest is an example of focused service, I call this approach multi-focused service.  To execute well on a multi-focused structure, Google must convince itself that multiple service models are better off under the same corporate roof.  I’ll talk about that more in later posts, but the key is shared services.  While shared services are often appealing at first glance, the model can be very difficult to pull off.  (Here’s a link to an HBR article where I touched on the concept of shared services briefly at the end.)

There are many examples of excellent organizations operating with either focused or multi-focused service models. But there are far more examples of organizations doing neither.  Instead, these organizations work hard not to disappoint either operating segment too much, which ensures a limit to the anger and outrage, but also ensures mediocrity.  It’s the path of least resistance because customers complain more about bad service than they do about the absence of excellence.  A hard truth about service is that you often have to disappoint some people in order to delight others.

The responses by two of Google’s employees seem to foreshadow the direction it’s taking:

Katie Watson, a Google spokeswoman, said no one was available to speak about the service problems. But in an e-mail statement, she said, “Solving customer support issues is extremely important to us.”

Andy Rubin, Google vice president for engineering in charge of Android technology, gave a similar response, indicating that its challenge was to reduce its email response time from days to a couple of hours.  These statements suggest that Google’s still committed to using its existing service model to serve an entirely new operating segment.  I’m hopeful that abandoning this fantasy is an outcome of these initial service difficulties.

Tom Brokaw, Heretic or Hero?

April 21, 2009


Have we moved beyond these cumbersome divisions we call states? In a recent Op-Ed Tom Brokaw proposed that local governments blur their administrative lines and work together to deliver services more efficiently. In discussing North and South Dakota’s 17 colleges and universities, he made the following blasphemous comment:

I know this is heresy, but couldn’t the two states get a bigger bang for their higher education buck if they consolidated their smaller institutions into, say, the Dakota Territory College System, with satellite campuses but a common administration and shared standards?

Jefferson and Madison may be turning in their graves, but Brokaw makes a legitimate case for consolidating costly and overlapping public services. The economics are clearly in favor of pursuing such a system, but politics often get in the way of this kind of progress. As Brokaw points out, parochial interests will be the biggest hurdle to making these changes a reality.

The same dynamics play out in companies. Even when there are known advantages to centralizing activities — even when it makes things cheaper and better — the self-interest of individuals and business units can undermine a centralization campaign. Firms that overcome this tension usually do a few things right. First, they put someone in charge of “shared services” who has the leadership skills to bring a skeptical organization along. Second, they focus on the better as much as the cheaper, on the upside of leveraging learning and best practices across the entire organization.

It’s not only that the Dakotas’ 17 colleges and universities can buy chalk for less when they combine some activities, but also that the Dakota Territory College System can use the knowledge now embedded in each institution to improve the education being offered by all of them. And designed correctly, realizing these “economies of experience” doesn’t have to come at the price of innovation or agility or even customization. Indeed, done correctly, these changes can free up the time and resources for an organization to deliver unprecedented quality to its constituents.