"If you don't give us a 9 or 10 it means we have failed." That's score-begging at a hotel checkout, and it's the visible symptom of badly set frontline NPS targets. Skip the targets entirely and NPS gets treated as the soft metric. Here's how to give frontline staff NPS targets without breaking the survey, covering punitive use, gaming, data variability, and line of sight.
Is the NPS scale a Likert scale? It's a question that comes up in survey-design arguments more often than you'd think, and the answer matters for how you analyse the data. Here's the short version, with the bit most articles get wrong.
"We tried Net Promoter and it didn't work." Almost every time someone says that to me, it comes back to a small number of fixable implementation mistakes — no C-level buy-in, treated as market research, no root cause analysis, score collected and nothing done. Here are the seven, and how to correct each one without restarting the program.
![How to Analyse Survey Data in Excel [+Video]](/_astro/how-to-analyze-surveys-in-exce-opt-e1607575803448.ZxA-rJo-.jpg)
Statistical packages are powerful, expensive and overkill for most customer feedback work. Excel can do almost everything they can do, and it's already on your desktop. Here's a step-by-step walk-through of analysing survey data in Excel, simple stats, error bars, histograms, with a video to follow along to.
There's a recurring critique that NPS doesn't actually measure anything real. It's a fair question, and it deserves a proper answer rather than a defensive one. Here's a pragmatic case for using a methodology that works, even when no one can fully explain why.
Customer feedback isn't a thermometer reading, it comes back tangled with halo effects, three-year-old memories and irrelevant tangents. That doesn't make it useless. Here's why messy data is still rich data, and how to stop "the data isn't clean enough" being used as a reason not to act.