Wednesday, June 22, 2011

Inherent Deception: the Net Promoter Score (NPS)

It seems that these days every time you check your email or visit a vendor website or call customer service, you will be asked to take a customer satisfaction survey. Veteran marketers understand what's happening here, but most consumers apparently don't.

Customer surveys are concerned primarily with the value of the brand, not the actual quality of service.

Aren't those the same thing? Or so strongly related you can't separate them? Not at all.

Here's a diagram that shows how most people think the survey process works:

The survey process, people seem to think, is the "Ask customers if they like the new product" step. In reality, there are just two steps: "Ask customers what they think of us" (i.e., our brand) and then "Keep asking until you see the score change because they think asking a lot means we care."

In this latter process, the survey itself is the customer service improvement. It's public relations. It's hype.

Lest you think I'm just a curmudgeon on a rant, consider the metrics behind these surveys. The primary methodology in use these days is called the Net Promoter Score (NPS). This approach was developed specifically to assess the value of a brand over time. It measures what people think of your name - the polish on your image - in a series of measurements. It does not incorporate, nor was it intended to, any particular improvement process. It's all about assessment. In my experience, the functional department in the company that is charged with conducting the surveys - typically, marketing - has little or no connection  or communication with the functional area responsible for delivering goods and services.

Outrageously, in my view, big companies take the money they save by employing robots to answer phones and deliver automated reminders (usually at dinnertime) and spend it on surveys. Companies are worried about the value of their brands because that's all Wall Street cares about. The perception of value is value, or so the wise market says. Now, you might say, true value eventually will out, and to that I say, no one is looking that far ahead. Today's market price and next quarter's earnings report are about as far ahead as anyone looks.

If a company really wanted to improve customer service, it would:
  • Take the budget spent on too-frequent surveys and apply it to quality metrics that actually measure how well the product or service performs (e.g., mean time between failures, time to resolve) and process improvement.
  • Hire fewer robots and third-party call centers and replace them with live people who have personal "customer intimacy" - that is, they live in the same region as the customer, they use the same products, they are at approximately the same economic level, and so on.
  • Train those people  thoroughly before you let them deal with real people.
  • Retrain, retrain, retrain to keep them current on the changing marketplace and your continuously upgraded offerings.
  • Incentivize, rather than simply browbeat, them to care about their customers.
  • Retain and reward them when your quality assessment (and, yes, your survey) numbers improve.