Same plan, same ARPU, very different cost.
In telecom the plan is the same for thousands of subscribers, but the cost of serving each one is not. The network capacity they drive at peak, the support contacts they make, the channel they came through and the provisioning they need all vary, and a blended cost-per-subscriber averages every one of those differences away. The subscriber who looks like every other on the plan can be the one quietly losing money.
Cost and Profitability Consulting · 150+ models since 2010 · TDABC
Cost-to-serve in telecom is driven by peak network consumption, support contacts, channel and provisioning, not by the plan. Industry research shows cost-to-serve varies two to three times between customers who look identical, and studies consistently find most organisations do not measure the cost of unused capacity. TDABC assigns network capacity (peak-weighted) and service cost to each subscriber, channel and service, so a blended ARPU stops hiding who actually pays. We apply this with transversal evidence and the method, not an invented telecom benchmark.
Consumption decides, not the price card.
A plan is a revenue label, not a cost. Two subscribers paying the same monthly fee can sit at opposite ends of the cost distribution: one streams at peak through a congested cell, calls support every month and was won through an expensive retail channel; the other uses little at peak, never calls, and self-served online. The plan charges them identically. The cost of serving them is two to three times apart, and the blended ARPU reports them as the same customer.
Peak network share
The capacity a subscriber drives at the busy hour is the real network cost, because peak is what sizes the network. Average volume understates the heavy peak user and overstates the off-peak one.
Support intensity
Contacts, complaints and truck-rolls are time on expensive resources. A high-touch subscriber consumes far more than a self-sufficient one on the same plan.
Channel of acquisition and service
Retail, call centre, reseller and digital have very different costs. The channel a subscriber uses is part of their cost-to-serve, and it never appears on the plan.
Provisioning and churn drag
Activations, plan changes and re-provisioning consume cost. A subscriber who churns through the catalogue costs more than a stable one, regardless of ARPU.
PEAK CAPACITY, NOT AVERAGE VOLUME
Illustrative analogue from a large IT organisation, not a telecom benchmark. Cost follows the peak the customer drives, not the average they report.
Cost follows the subscriber, contact by contact.
The subscriber's cost is built from what they consume: a peak-weighted share of network capacity, the minutes of support they take, the field events they trigger, the channel cost they carry and the provisioning overhead they generate. Multiply by the capacity cost rate of each resource and the cost lands on the subscriber that caused it, not on the plan average.
Cost to serve a subscriber = network capacity share (peak-weighted) + support contacts x minutes x call-centre rate + field / truck-roll events x crew time + channel acquisition and servicing cost + provisioning and billing overhead
Illustrative structure, not a measured benchmark. The peak share and the support term are where same-plan subscribers pull apart.
Peak-capacity pricing, from an adjacent sector.
A large IT organisation priced shared IT capacity using peak-capacity pricing across several time zones, charging internal consumers for the capacity they drove at the busy hour rather than a flat average. The result was that heavy peak users finally carried their true cost and off-peak users stopped subsidising them. Telecom's network is the same kind of shared, peak-driven asset, so the logic transfers directly: charge cost to the peak each subscriber drives, not to the bytes they average. This is an illustrative analogue from an adjacent sector, anonymised and with adjusted figures, not a telecom benchmark.
Frequently asked questions
- What is cost-to-serve in telecom?
- The full cost of serving a subscriber: peak-weighted network capacity, support contacts, field events, channel cost and provisioning, before the revenue. Industry research shows it varies two to three times between customers on the same plan.
- Why do two subscribers on the same plan cost differently?
- Because they consume different network capacity at peak, make different numbers of support contacts, were acquired through different channels and need different provisioning. The plan is the same; the consumption is not.
- How is channel cost-to-serve measured?
- By assigning the real cost of each channel, a retail store, a call centre, a reseller, self-serve digital, to the customers who use it, rather than spreading a blended cost across all of them.
Find the subscriber cost your ARPU is hiding.
The Profit Check takes five minutes and no data upload. It points to where your cost-to-serve and your plan pricing are most likely out of line, and what closing the gap is worth.