Interoperator tariff negotiations determine the wholesale rates that operators pay each other for roaming traffic. They run on multi-year cycles, involve significant sums, and have direct margin implications for both retail and wholesale roaming revenue. Most of them happen without a working model of how the proposed rate changes will actually affect the operator's financial position. The reason is not a lack of interest in data — it is that the data exists in a form that most operators' systems were never designed to analyse for this purpose.
The data problem at the heart of IOT negotiations
The raw material for IOT negotiation intelligence exists in every operator's environment: TAP files recording roaming usage, NRTRDE data for near-real-time traffic monitoring, bilateral agreement terms, and the traffic distribution data that shows what mix of services subscribers actually use when roaming on a given partner network.
Correlating these datasets to answer the question that actually matters in a negotiation — "if we move this rate by X%, what happens to our margin across our current traffic profile?" — requires joining data sources that were designed for different operational purposes. TAP files are optimised for settlement. NRTRDE is optimised for fraud detection. Neither was built as a negotiation intelligence tool, and most BSS environments have no native capability to bridge them for commercial analysis.
What operators typically go into negotiations with
Most roaming teams prepare for IOT negotiations with historical aggregates — total inbound and outbound traffic volumes by partner, approximate revenue per unit, and a qualitative sense of which partners generate favourable traffic mix. What they lack is a forward-looking margin model that projects the impact of specific rate proposals across their actual subscriber usage patterns.
The consequence is negotiations conducted on the basis of experience and relationship rather than quantitative modelling. That is not inherently irrational — experienced roaming managers develop good intuitions about partner markets. But it means that the outcome of a negotiation depends significantly on who is in the room, rather than on which operator has the more accurate model of the economics. In a sector where IOT-driven margin differences can run to millions of dollars annually per major partner, that is a significant structural disadvantage.
Why the BSS was never built for this
Business support systems in the roaming context were architected around the operational functions they needed to perform: TAP file processing, clearing, settlement, and billing. The data structures and processing pipelines that make BSS effective for those functions are not the ones that make it useful for commercial negotiation analysis.
Volume is part of the problem. A major operator processing billions of roaming CDRs annually cannot run ad hoc analytical queries against live TAP processing infrastructure without affecting operational performance. The analytical work needs to happen in a separate environment, against a data model that is designed for slicing and scenario analysis — not against the settlement database.
What data-driven IOT negotiation looks like
Operators with mature roaming analytics capabilities approach IOT negotiations by building a margin model for each significant partner relationship before negotiations open. The model ingests the historical TAP record for that partner, segments traffic by service type and subscriber category, and then applies the proposed rate changes to the actual usage distribution to project the margin impact.
Scenario modelling adds a second layer: what happens to the overall roaming margin if the rate change shifts traffic patterns — if, for instance, a more favourable outbound rate on data drives higher data roaming usage on that partner network? These are the questions that experienced roaming managers ask intuitively. A working analytical model lets them answer them quantitatively, and adjust their negotiating position in real time as proposals evolve.
What an analytics platform needs to deliver for IOT
The capabilities that distinguish roaming analytics platforms built for IOT negotiation support from those built only for operational reporting are well-defined.
- TAP and NRTRDE ingestion at sufficient granularity to reconstruct traffic distribution by partner, service type, subscriber segment, and time period.
- IOT rate modelling that applies proposed tariff changes to actual historical traffic and projects margin impact at the CDR level, not just at aggregate volume level.
- Multi-scenario comparison — the ability to run and save multiple rate proposals simultaneously and compare their projected margin outcomes.
- Commitment modelling that accounts for volume commitments and their impact on effective rate under proposed agreement structures.
- Forecast capability that extrapolates historical traffic trends into the period covered by a proposed agreement, so that the margin model reflects what the operator expects to generate, not just what it generated in the past.
