Operational metrics: honest servants, terrible masters

Not every number on the dashboard has to be about the customer. Cost per acquisition, cost per contact, and operational efficiency are legitimate business measures, and they have a rare virtue: they are hard to game. A cost is a cost. The trouble starts when these honest servants become the boss, and teams begin meeting cost targets by making the customer's experience worse.

Why it matters to the business

Savings that damage experience are false economies. PwC's research across 15,000 consumers found 32% will walk away from a brand they love after a single bad experience, and Bain research cited in Harvard Business Review puts the cost of acquiring a new customer at 5 to 25 times that of retaining one. Cut support quality to save a few dollars per contact and you trade cheap contacts for expensive replacements.

Cost-cutting can even fail on its own terms. Gartner's channel research found customers adopt new self-service channels without abandoning the expensive live ones, so channels bolted on to cut cost often add contacts instead.

How to use it

• Publish cost metrics and experience metrics on the same page so trade-offs stay visible.
• Require every cost-reduction proposal to state its predicted effect on customer effort, resolution, and churn.
• Measure cost per resolution, not cost per contact; a cheap contact that fails is not cheap.
• Adopt a standing rule: no operational target may be met by degrading the service experience.
• Review quarterly whether recent cost savings were followed by rising repeat contacts or churn.

Where teams get it wrong

The failure mode is inversion. The cost number becomes the goal, so teams shorten calls, deflect contacts, and bury the phone number. Handle time falls, the dashboard glows, and the cost reappears with interest as repeat contacts, escalations, and quiet defection. Lagging cost indicators are acceptable precisely because they resist manipulation; they are never a license to manipulate the customer.

Ask your team

• Which decisions this year were made to hit a cost target, and what happened to repeat contacts afterward?
• Do we track cost per contact or cost per resolution, and what would change if we switched?
• If support cost fell sharply next quarter, how would we know whether we got more efficient or just harder to reach?

The boundary matters more than the metric.