Publishing a Metric Is Good. The Right One Is Better

A contact-center giant deserves real credit here: it publishes containment and efficiency figures. Most vendors refuse. But containment rate measures whether a call avoided a human, not whether the customer's problem got solved. For an appointment-driven business, the metric that predicts revenue is completion. Did the caller get booked, rescheduled, or answered correctly? That matters. Whether the call stayed inside the automated system does not.

Why is publishing any metric better than publishing none?

Transparency is a rare discipline. The contact-center giant's own reporting of case-study containment figures gives buyers something concrete to interrogate. Most vendors in this category publish marketing copy and nothing that can be checked against a source. Putting a number in front of analysts and press is a meaningfully more honest posture than silence. We should state that plainly before arguing the number itself is the wrong target for call- and appointment-driven businesses. We laid the foundation for this claim in the Resolution Scorecard buying standard.

Why can a published metric still be the wrong metric?

This is where Goodhart's law becomes operational. Anthropologist Marilyn Strathern's 1997 paper on the "audit explosion" formalized the idea now commonly summarized as Goodhart's law: "when a measure becomes a target, it ceases to be a good measure" (Strathern, 1997, European Review). Containment rate is a target metric for a contact center trying to reduce headcount cost. It rewards keeping the caller inside automation, regardless of the outcome. That makes sense for high-volume, low-stakes transactions like a balance inquiry. It is the wrong target for call- and appointment-driven businesses. When a caller wants a specific slot, a specific advisor, or an answer the automated system cannot provide, forcing containment traps them. The system holds the caller hostage instead of routing them to booking.

Gartner's research on Customer Effort Score makes the same point from the customer side. High-effort service interactions predict disloyalty better than satisfaction scores do. Low effort, not fewer live-agent transfers, is what operators should optimize for. A contained call that leaves the customer without a booked appointment is a high-effort call wearing a low-cost label.

What's the counterargument, and does it hold up?

The counterargument is fair. Containment matters because live agents are expensive. Every case-study containment figure represents real cost avoided for that specific deployment. That is true, and it explains why vendors should label these figures per case study rather than generalizing them as a company-wide rate. The contact-center giant's published numbers are case-specific results, not fleet-wide averages. The problem is not that containment is meaningless. The problem is that it answers a cost question, not a completion question. Appointment-driven businesses live or die on completion. We apply this exact logic one level deeper here: what call volume figures prove.

What should an appointment-driven business measure instead?

A tire shop, a dealership service lane, or a furniture delivery desk should measure the third row, not the first. Containment can go up while completions go down. A scorecard that only reports containment will miss that failure every single time.

Publishing a metric at all is a discipline worth crediting. See our full breakdown on our comparison of the alternatives of what that vendor reports and what it leaves out. For how this same gap shows up in queue design, read queues versus booked appointments. You can also see how AiPRL's Analytics reporting is built entirely around completion, not containment.