A pricing model is only as strong as what it can show its regulator
Energy and water utilities across Africa increasingly build pricing and tariff models on real-time metering and telemetry data — a genuine improvement over the estimated billing and manual reads that preceded it. The improvement only holds if that data is governed as evidence, with the lineage to survive a regulatory review, rather than simply displayed on an internal dashboard and assumed to be sufficient.
A regulator evaluating a tariff proposal is not asking whether the utility has real-time data. It is asking whether the utility can show, step by step, how metering readings became the cost allocation and loss figures the proposed price is built on — and whether that chain would hold up if challenged.
The three things a regulator will actually ask for
Metering data with a documented collection method. Not just the reading, but how it was captured, at what interval, and what happens when a meter fails to report — is the gap flagged, interpolated, or silently dropped from the calculation. Silent gaps are the single most common way a pricing model's inputs turn out to be less complete than presented.
Loss calculation logic that can be reproduced. Technical and commercial losses are estimated, not directly measured, and the estimation method matters as much as the resulting number. A regulator that cannot reproduce a utility's loss calculation from the stated methodology has no basis to trust the tariff built on top of it — and will say so.
Cost allocation that traces back to an auditable basis. When a pricing model attributes cost to a customer segment, geography, or time-of-use band, that attribution needs a documented basis a third party could check, not an internal assumption inherited from the previous tariff cycle and never revisited.
Real-time telemetry raises the bar, it does not lower it
Utilities sometimes treat the move to real-time metering as evidence in itself — more granular data, collected more often, framed as inherently more defensible than the manual reads it replaced. Granularity and defensibility are different properties. A telemetry system that reports every fifteen minutes but has no documented process for handling missing readings, sensor drift, or meter tampering produces more data without producing more trustworthy evidence.
The utilities that get the most value from real-time telemetry are the ones that pair it with the same lineage discipline that any other evidence base requires — a documented chain from sensor reading to reported cost, calibration and maintenance records that keep sensors honest over years, not just at installation, and a reconciliation process that catches discrepancies before they reach a regulatory filing rather than after.
What this means in practice
Building a pricing model regulators can approve without extended back-and-forth means treating the evidence layer as part of the tariff design process, not a compliance step added after the numbers are finalized. Utilities that do this consistently spend less time defending individual filings, because the underlying evidence base was built to survive scrutiny before the scrutiny arrived.