Call volume is a fact of nature, like the monsoon: you forecast it, you staff for it, you cope. That belief is comfortable and mostly wrong. A large part of what arrives at a contact centre exists because something upstream failed: a bill nobody could read, a delivery date that slipped silently, a form that rejected a valid entry, an answer that did not solve the problem the first time. Calls like these are not demand. They are failure demand, a term from the work of John Seddon, who contrasted it with value demand, where a customer wants something the organisation is actually there to provide.
Bill Price and David Jaffe built a whole book on the consequence, titled The Best Service Is No Service. Their idea is simple to say and hard to do: find why customers contact you, and remove the reasons. The agent who handles the call brilliantly is a second-best outcome. The best outcome is that the call never happened.
Repeat calls are the easiest place to start
The word is slightly off, actually. I should say repeat contacts, because customers also chat, email and tweet. A repeat contact is the same customer coming back about the same issue within a set window, often 7 days, sometimes 24 or 72 hours, depending on the client’s definition. That window is a policy choice. Change it and your repeat rate changes with nothing on the floor changing, so always ask what window a quoted figure uses.
Repeat contacts matter for two reasons. They cost money twice, and they signal that the first contact failed. First contact resolution (FCR) is the flip side: the share of issues solved without a second attempt. They are not perfectly complementary, since FCR measured by survey and repeat rate measured from call logs can disagree, but they point at the same wound.
A worked example
Take an imaginary telecom process receiving 100,000 contacts a month. Analysis of the logs shows 18% are repeats within seven days: 18,000 contacts. If a fully loaded contact costs ₹55 (a made-up average, include agent, supervision, seat and technology), those repeats cost ₹9.9 lakh a month.
The team digs in. Tagging the repeats shows a third come from one cause: customers told “your refund will be processed” with no timeline, who then call to ask where it is. A single change, a refund status SMS with an expected date, removes most of those calls. Say it prevents a third of all repeats: 6,000 contacts a month. Saving: 6,000 times ₹55, which is ₹3.3 lakh a month, about ₹39.6 lakh a year. No hiring, no coaching programme, one SMS template and an integration.
Observe what did the work. Not agent training. A process fix at the source.
Finding the avoidable share
You need three things. A decent set of disposition codes, so that you know why people called (and if yours say “other” for a quarter of calls, start there). A way of linking contacts from the same customer, which needs a customer ID across channels. And someone willing to sort reasons into groups: customer asking for something we offer, customer asking about something we failed at, customer unable to self-serve something we could have made easy.
Then rank causes by volume, as in a Pareto chart, and look at the top three. Questions worth asking of each: could proactive information have prevented it? Is a policy or a form generating confusion? Can the customer do it alone in an app, and if not, why not? Is the agent unable to fix it because of a permission, so it bounces to another team?
The uncomfortable bit for a BPO
Here is the tension nobody likes to mention. If your client pays per FTE or per call, removing calls can reduce your revenue. A BPO that cuts volume by 10% may see its headcount, and billing, shrink by about as much unless the contract has other terms. So the business model matters. Under per-seat pricing, the incentive to reduce avoidable contacts is weak. Under outcome-based or gain-share models, where the saving is shared with the vendor, it is aligned. This is one reason many clients want volume reduction on the agenda while vendors hesitate, and a sensible account lead raises it with the client rather than waiting. A vendor who helps cut failure demand and gets redeployed to higher-value work usually builds a longer relationship than one who defends volume. That is my judgement, not a law.
Where the idea has limits
Not all contact is waste. Some customers call because they want a human, to feel reassured, to negotiate, to complain properly. Pushing them to self-service against their will leads to a different problem: abandoned journeys and later, angrier calls. A drop in contacts is also not proof of improvement. If you make the phone number hard to find, calls fall and customers go silent for a while, then leave. Track whether the underlying problem disappeared, through resolution on first contact, complaint volumes, churn and customer effort. And when a fix reduces calls, check the average handle time. The easy calls vanish first, so what remains is harder, and AHT often rises even though the process improved. Report it that way, or the improvement gets read as a deterioration.
Next time someone reviews the monthly volume forecast, try one more question alongside “how many will come?”: how many of them did we cause?
Figures and names in the examples are illustrative.
Leave a Reply