Measure friction, not praise

Most customer surveys fish for praise. They ask people to rate their satisfaction or how likely they are to recommend, and the answers cluster near the top of the scale while the real problems go unreported. Friction measurement inverts the question. Instead of asking customers whether they are happy, it intercepts them during or right after a task and asks them to score negative dimensions: was this confusing, frustrating, slow, too many steps? Scores run from zero to one hundred, where higher means fewer bad experiences. A companion move works one level up: an anonymous, cross-functional assessment that asks product, engineering, and CX teams how customer-centric the organization actually behaves, not how customer-centric it claims to be.

Why it matters to the business

Praise-seeking surveys feed corporate self-delusion. Bain's Closing the Delivery Gap study found 80% of companies believed they delivered a superior experience while only 8% of their customers agreed. Friction, by contrast, predicts real money: CEB research published in Harvard Business Review found effort-based measures were 1.8 times more predictive of loyalty than satisfaction scores, and that 96% of customers who had high-effort experiences became more disloyal, against 9% after low-effort ones. There is also a cost to asking badly. Qualtrics XM Institute data shows survey response rates have fallen 7 to 8 percentage points since 2021, and Gartner finds only 16% of customers strongly believe their feedback drives change. Every wasted question spends goodwill.

How to use it

• Add a short friction question set to your three highest-volume tasks this quarter: confusing, frustrating, slow, too many steps.
• Trigger the questions at the moment of completion or abandonment, not in a monthly email blast.
• Report friction scores by task, and give each of the three worst tasks a named owner and a fix deadline.
• Run an anonymous cross-team assessment of customer-centricity once a year and compare the results with what leadership says publicly.
• Pair every friction score with a behavioral signal, such as repeat contacts or abandonment, so self-report never stands alone.

Where teams get it wrong

The classic failure is collecting friction data and treating it like satisfaction data: it lands on a dashboard, gets averaged into a single number, and nobody owns the repairs. Friction scores only earn their keep when each low score maps to a specific task, a specific owner, and a scheduled fix.

Ask your team

• Which of our surveys lets a customer tell us something is broken, and who acts on that answer within a week?
• What are the three highest-friction tasks in our experience right now? If nobody can name them, we are fishing for praise.

You learn more from asking what hurts than from asking whether people are happy.