The assumption that customers want telephone contact

Where the assumption comes from

The assumption that customers want telephone contact was not invented. It comes from a period in which the telephone was the only channel with a human on the other end who could answer immediately. Forms were slow, email was slow, and a chatbot was a menu of choices nobody understood. Whoever wanted a quick answer called. Customer satisfaction figures confirmed this for years: telephone contact scored higher than other channels, and management built on that. Contact centres were set up around call volume, waiting time became the most important KPI, and "the customer wants the phone" became a fixed phrase in every quarterly report.

The assumption was not incorrect because customers necessarily like calling. It was correct because calling was the fastest route to a good answer. That distinction is at the core of what is now shifting.

What has changed

AI is taking over part of the work that used to only go by phone: looking up a status, processing a change, answering a question that comes up the same way a hundred times a day. This is not happening everywhere at the same speed or in the same way. Three categories run through every contact process:

Fully automatic. Standard questions with an unambiguous answer — balance, delivery time, status change — are handled directly in a growing share of cases, without waiting time and without a human.
With oversight. More complex questions are prepared or answered by AI, after which an employee approves, adjusts or rejects with reason. The customer often no longer notices whether the first answer came from a system.
Remains human work. Emotionally charged conversations, negotiation, and situations in which the customer needs someone to take responsibility remain with people — not because AI cannot handle it, but because the customer is looking for something other than speed there.

Where this already works, it is often not where the technology is most advanced, but where someone took the trouble to determine per contact reason which type it is. Where the assumption still stands as self-evident, it is usually where nobody has made that breakdown.

What tells management that the assumption has expired

The assumption does not expire with an announcement. It expires in signals that are usually read separately rather than together:

Call volume drops while customer satisfaction stays the same or rises — a sign that customers have found an alternative and are not, as previously assumed, becoming more dissatisfied without the phone.
Younger or more digitally active customer segments structurally call less than older segments, suggesting that the preference for calling was a habit of one generation of customers, not a universal wish.
Competitors with less telephone availability do not lose customers because of it — a public signal that the market has already let go of the assumption.
Internally, capacity is still budgeted based on call volume, while actual contact volume via other channels is growing — a sign that planning follows the assumption, not the data.

None of these signals proves anything on its own. Together they form the picture from which management can see whether the assumption still carries the plan, or only still the budget.

What this does to capacity, not to staffing

The question underlying this is a capacity question: how much of the contact volume that now comes in by phone no longer needs to, and which part of the freed-up hours can go to conversations that do require oversight or human work. This is not a question about who keeps their job. What an employer does with that capacity falls under its own statutory requirements for personnel decisions; that is up to the employer, not a strategic check. This page describes only what happens to the work itself.

Why this is rarely an isolated assumption

The telephone assumption is often linked to other assumptions that are simultaneously under pressure. Whoever assumes that customers want to call often also assumes that the knowledge lead of their own team remains the distinguishing factor in the conversation, while AI increasingly delivers that same knowledge directly to the customer. And whoever assumes that a larger contact centre automatically performs better runs into the assumption that scale still offers protection while smaller players with fewer people switch faster. Whoever wants to understand how an assumption is actually built up and when it is due for replacement finds the basis in the explanation of what a strategic assumption is and how you test whether it still holds.

What you can measure

To know whether this assumption still holds, a fixed set of measurements is needed, not a one-off impression:

Call volume per contact reason, broken down into what AI can handle, what requires oversight, and what remains human work.
Customer satisfaction per channel, not averaged across all channels.
The share of customers who chose an alternative channel when calling did not work immediately, and whether they still called afterwards.
The age of the last time this assumption was tested against real data, not against an impression from a meeting.

The underlying question — which work in this company can genuinely be taken over by AI, and which part remains with people — is mapped per task with the work scan from FTE TO AI.

What you can do now

A management team that wants to know whether this assumption still holds can start with the free assumption check: a short round in which you name your most important assumptions and see, per assumption, when it was last confirmed. This is also the place to see whether the views within your own management team on this kind of assumption diverge, which is itself already a signal. The full strategic stress test, with sector data alongside management's own self-plot, is under construction.