Bilateral IOT renegotiation has historically been a commercial process with a relatively limited evidential basis: each party knows its own revenue from the relationship and its aggregate network costs, but detailed per-partner, per-RAT cost attribution is rarely available. The party seeking higher rates makes an assertion; the counterpart accepts, rejects, or compromises. The outcome reflects bargaining position more than economic reality. That model is changing — and the operators changing it are achieving better results.
What cost attribution actually requires
Per-partner IOT margin modelling requires three inputs joined at the per-session level: settlement records (what the partner's inbound roamers consumed, by service type and RAT), network cost allocation (what it cost to carry that traffic, broken down by access, core, and transport), and the current IOT schedule (what is charged per unit for each service type).
The network cost allocation is typically the hardest component. Telco cost ledgers are not designed to allocate costs at the per-session, per-RAT level — they are usually structured by network layer, equipment group, or geographic region. Building the allocation model requires engineering input on the cost-per-GB-per-RAT at the access layer, core processing cost per session type, and any transport or interconnect costs associated with delivering the intercepted traffic.
The output: a per-partner contribution margin
The output of the modelling exercise is a contribution margin by partner: for each bilateral relationship, what is the difference between the revenue generated under the current IOT and the fully allocated cost of carrying that traffic. Partners are ranked from most to least profitable, and the analysis identifies which IOT elements — data rate, voice rate, SMS rate, NR premium — are driving the misalignment in under-performing relationships.
This output reframes the renegotiation conversation. Instead of "we want higher rates," the commercial team can present: "here is the cost of serving your subscribers on our network, here is what we are currently recovering under the IOT, and here is the specific rate adjustment that would bring the relationship to a reasonable margin level." Most counterparts, when presented with this analysis, engage more constructively than they do with a generic rate request.
Operationalising the model
The initial margin model is typically a one-off analytical exercise to build the renegotiation evidence base. Operators that find value in the exercise usually move to operationalise it: building a recurring join of settlement data and cost ledger that updates the margin picture on a monthly or quarterly basis, creating a commercial dashboard that tracks each bilateral relationship's margin trajectory, and embedding the model in the IOT negotiation preparation process.
The recurring model is more valuable than the one-off: it allows the commercial team to track whether a renegotiated IOT is actually producing the expected margin improvement as traffic volumes evolve, and to identify new pressure points before they become significant. In a roaming landscape where NR traffic share is growing quarter over quarter, a model that was accurate in Q1 may need updating by Q3.
