CX Philosophy
The only hard data that matters is yours
Benchmarks describe other companies. Hold every idea to one evidentiary standard and let your own customer data make the call.
Selective rigor is the anti-pattern: demanding proof to block ideas you dislike while accepting external benchmarks unquestioned. What works for Google is evidence about Google. Your context, and your own joined-up data, decide what works for you.
Every organization has a version of this meeting. Someone proposes a change; a skeptic demands hard data. Someone else cites what Apple does, or a famous industry benchmark; heads nod, and no data is demanded. That is selective rigor: proof required for ideas you dislike, a free pass for ideas that arrive with a logo attached. External benchmarks strip out context. Different customers, different constraints, different economics. What works for Google is evidence about Google, nothing more, and popular is not the same as effective.
Why it matters to the business
Even celebrated industry numbers can fail the transfer test. When Keiningham and colleagues re-examined Net Promoter's claimed superiority across 21 firms in the Journal of Marketing, it predicted revenue growth no better than the older ACSI satisfaction index. The lesson is not that any one metric is useless; it is that borrowed conclusions need local validation before they steer your money.
Your own data, joined properly, is far more decisive. Peter Kriss's analysis in Harvard Business Review linked each customer's experience scores to their later behavior: customers with the best past experiences spent 140% more than those with the worst, and great experience lifted one subscription firm's one-year retention from 43% to 74%. Numbers like that, drawn from your own ledger, end debates that benchmarks only prolong.
How to use it
- Join experience data to revenue: connect each customer's feedback and behavior to their later spend and churn in your own systems, then price journey fixes by dollars at stake.
- Weigh a customer complaint above an internal 'it works for me'. A complaint is field evidence; an opinion is not.
- Combine quantitative metrics with qualitative signals: complaints, emotions, and observed behavior carry information dashboards miss.
- When someone cites a benchmark, ask what would have to be true of your customers for it to transfer, then test exactly that.
- Apply the same evidentiary bar to ideas you love and ideas you oppose, and say so out loud in the meeting.
Where teams get it wrong
The failure mode is benchmark worship: copying a competitor's checkout flow, survey program, or org chart and expecting their results. Imitation imports the visible practice and none of the invisible context that made it work. Teams then measure success against the industry average instead of against their own customers' behavior, which is the only scoreboard that pays.
Ask your team
- Which of our current initiatives exists mainly because a competitor or benchmark does it, and what local evidence supports it?
- Can we trace this quarter's experience scores to actual spend and churn for the same customers?
- When did a customer complaint last override an internal opinion?
What works for Google is evidence about Google.
Apply this
Reading about the only hard data that matters is yours is one thing. Seeing where it applies in your journey is the useful part.