CX Methods
Never leave a harmful KPI in place 'temporarily'
A KPI you know is harmful does damage every day it stays live. Swap in an imperfect proxy now and refine it later.
The slow fix sounds responsible: change the KPIs when the new platform is ready. Meanwhile the harmful metric keeps steering hundreds of daily decisions the wrong way. Deferral is not neutral; it is ongoing damage.
Every organization that diagnoses a bad metric hears the same reasonable-sounding plan: we will fix the KPIs when the new measurement platform lands next year. Meanwhile the harmful number stays on dashboards and in bonus plans, and every day it stays live it keeps steering teams toward behavior everyone has already agreed is wrong. Deferral is not neutral. It is ongoing damage with a project plan attached.
Why it matters to the business
Metrics steer harder than strategy documents, especially with money attached. Fred Reichheld, who created NPS, warned in Harvard Business Review that linking scores to frontline bonuses breeds score-chasing, pleading, and gaming; the same logic applies to any KPI that rewards the wrong behavior. And misdirection compounds quietly: Bain's 'Closing the Delivery Gap' found 80% of companies believed they delivered a superior experience while only 8% of customers agreed. That is what sustained optimization of a self-serving number looks like from the outside. Every quarter a harmful KPI survives, it shapes hiring, coaching, roadmaps, and thousands of small frontline choices, and unwinding those habits costs far more than replacing the number.
How to use it
- The week a KPI is diagnosed as harmful, pull it from targets and dashboards.
- Stand up an imperfect proxy that points the right direction, blending surveys with behavioral analytics.
- Decouple the old metric from bonuses immediately, before the next payout cycle.
- Set a review cadence to refine the proxy as better data arrives.
- Document what the old KPI incentivized, and test the replacement against that failure list.
Where teams get it wrong
The failure is perfectionism. Teams treat measurement change as a platform project with a roadmap and a go-live date, because an imperfect proxy feels unrigorous. But rigor is not the issue; direction is. The frontline keeps optimizing whatever is live today, so a rough measure pointing the right way beats a polished one pointing the wrong way, starting immediately.
Ask your team
- Which of our current KPIs would we not choose again today, and why are they still live?
- If we discovered this morning that a KPI was harming customers, what is our process for pulling it by Friday?
- What is bonus pay currently attached to, and how could each of those numbers be gamed?
An imperfect proxy that points the right direction beats a precise metric that points the wrong one.
Apply this
Reading about never leave a harmful kpi in place 'temporarily' is one thing. Seeing where it applies in your journey is the useful part.