American Airlines was getting three times more on-hold Twitter complaints than its nearest competitor, even after adjusting for passenger miles. Could public crowd data predict NPS without running a survey? The numbers say yes, and that has interesting implications for benchmarking, brand monitoring, and how loudly hold times leak into customer loyalty.
I've spent years telling people to ignore external NPS benchmarks. So when Bain's Rob Markey, the guy who literally wrote the book, pushed back, I expected an argument. We ended up in violent agreement. Here's the narrow set of cases where benchmarks actually do help.
Transactional NPS is just two questions, but the way you collect the responses changes the response rate by 5x. Outbound phone, inbound IVR, email, web; each has its case and its trap. Here's the comparison across the main collection channels, and the criteria (timely, frequent, low-touch) that pick the winner for your touchpoint.
Every couple of years someone proposes a new metric that's about to make NPS obsolete. Net Revenue Score, pitched in the Wall Street Journal, was the latest. It misunderstands what NPS actually measures, and the response is worth reading even if you don't care about the metric.
Online customer advisory panels are useful — fast feedback, large sample sizes — but the ones run by market research companies build a relationship with the researcher, not you. There's a more valuable version: a dedicated panel that belongs to your brand. Here's how one ran for an insurance client launching a new direct brand, and what it returned that an outsourced panel couldn't.
Most new product development is done by guessing what customers want, then surveying them at the end to confirm. Customer co-creation flips that. After 18 months working with a financial services client that ran an online conversational community alongside focus groups, here are four ways to involve customers in NPD that lower risk and lift relevance.