A little less than two years ago on the sascom voices blog, I argued that even social media enthusiasts should be listening more to social media curmudgeons like Tom Davenport. Tom is best known in SAS circles for co-authoring the books Competing on Analytics: The New Science of Winning and Analytics at Work. He also maintains a blog at Harvard Business Review, where he writes about IT, analytics and enterprise technology.

In his most recent post, Web 2.0 Versus Service 1.0, Tom argues that, in many cases, he'd prefer to see companies focus on good customer service via traditional channels before moving on to meet customer needs through social media channels like Twitter. He says:

I, for one, would rather work with a company that truly cares through a single communications channel — even snail mail — than one that shows its lack of caring through all the latest Web 2.0 tools.

It's a good point, but Frank Eliason from Comcast also makes a good counterpoint in the comments of Tom's post:

70% or more of the people we help via Twitter never contacted us before. Twitter is an early warning system and we see things there before other communications channels. This helps reduce overall costs and create better Customer experiences ... Since we started our social media work we have seen strong improvements in Customer Satisfaction, and we look to continue to improve them.

I recently spent some time talking with customer service for my digital phone and Internet service, and I was left wondering if I might have received better, more informed service on Twitter. I was dropping calls in my home office and losing internet connection at least a couple times a day. Usually when this happens, I just reset my modem and go on about my day. But after five days of repeated lost calls, I contacted a service rep through online chat.

I told the support person the history of problems I was experiencing and let her know the status of my connection at that moment: my internet was working but my digital phone was not. The first mistake she made was to assume my complaint was only about the immediate problem. It was not. She wanted to walk me through the process of resetting my modem, which I’d already been doing all week all by myself. I reset it again as we were chatting, and she typed, “Great! So we’re all set?!”

No! My complaint was not about the fact that my service was out at that moment. The problem was ongoing, and that’s why I contacted customer support. I explained that, and the support person said, “Well, let us know if the problem persists, and we’ll set up a service call.”

“No,” I said again. “It is already a persistent problem. I want a service call set up now.”

So the next day I hear from the service technician and learn that three main servers were out that week, and another power spike or some such thing at the local cable office was also causing problems. She came out to my house anyway and tested my modem, and of course I haven’t had any problems since then.

For customer care specialists: How many things could have been handled differently here? What else could the original support specialist have done to help keep my business? At what point does true customer service that aims to keep customers look beyond the immediate situation and address the larger problem? And what data or information is needed to address the larger problem?

On the one hand, to support Tom's point: the customer service through the traditional Internet site could have been improved. But to support Frank's point: I do wonder if I would have received better, more informed service - or even notified the provider about the problems - if I had mentioned my connection difficulties on Twitter sooner.

I'm not typically one to complain on Twitter about my problems, but when you change your viewpoint and start to see customer complaints as useful data points, as opposed to random bellyaching, you can start to see the business benefit of mining Twitter for brand sentiment.