When Net Promoter Scores Don’t Make Sense
There are plenty of good reasons to use Net Promoter Scores (NPS). They are based on sound measurement, are simple to use, are conceptually compelling, and are widely known by business executives outside of market research.
But you know something is amiss when your bank starts asking how likely you are to recommend the online tool you are using to a family member or friend. Recommend it for what? So my friends can transfer money among my accounts? Should I share my password so they can check it out? NPS seems like an odd way to measure my satisfaction here.
Turns out that a lot of respondents think so, too. At a recent conference in Baltimore, LIMRA presented research showing that many consumers would never recommend certain products, services, or even certain types of companies no matter how ecstatically satisfied they felt. When it comes to banks, for example, just two out of five (39%) consumers said they would be very likely to recommend a bank, even if their experiences were extremely favorable. For brokerage and mutual fund companies, only one in four (27%) would be very likely to recommend.
It reminded me of in-depth interviews we did last year to explore what was driving certain low NPS scores our client was getting. I asked an HR administrator in charge of benefits for the company, “When you completed the survey online, why did you give a score of 5 on likelihood to recommend?” Her answer: “It’s not that I was saying I wouldn’t recommend them. It’s that I’m an HR administrator, and I have to be neutral, so I can’t recommend these benefits over any others offered.”
So before using an NPS measure for everything, and before dropping other measures of satisfaction or loyalty, think hard about whether it makes sense in the context you want to use it. Will customers think it’s a weird question? Would they ever make the kind of recommendation you’re asking about? Do a little research first (hey, give us a call for that!) so you don’t end up with an NPS that makes no sense.