The complaint did not stop at the call centre. It went to the regulator.
A billing dispute, a data bundle that vanished, a deactivation the customer says was wrong — most die in the contact centre. The ones that escalate land at ICASA's complaints committee or the Communications Authority's consumer unit, where the operator's written answer is read by someone with the authority to act.
Every operator runs a complaints funnel, and most of it never leaves the building. A customer disputes a charge, queries a vanished bundle, or contests a service decision; the contact centre resolves it or it lapses. The small fraction that does not resolve is the part that matters, because it does not simply disappear — it escalates to the regulator's consumer machinery, where the operator becomes the respondent in a proceeding.
The machinery is real and structured. In South Africa, ICASA runs a consumer complaints process that moves from the operator's own resolution attempt to an alternative dispute resolution stage and, if unresolved, to the Complaints and Compliance Committee for adjudication. The Communications Authority of Kenya is mandated to receive and handle consumer complaints — including the fraud complaints that increasingly dominate the channel — as part of its regulation of the sector. At each escalation the operator must produce a written account of what happened and why.
That written account is the artefact that decides the matter, and it carries a risk most contact centres are not structured for. The regulator is not reading the operator's intentions; it is reading the consistency between the operator's answer and its own records. If the response to the regulator says one thing and the interaction history, the billing record, or the deactivation log says another, the inconsistency does not merely lose that complaint — it invites the regulator to look harder at the operator's conduct generally.
The operational difficulty is that the response is usually assembled under time pressure by someone reconstructing events from fragmented systems — a billing platform, a CRM, a network log, a fraud case file — none of which were designed to produce a coherent narrative for a regulator. The person drafting the response is making judgements about what the records show, and those judgements become the operator's official position whether or not they are fully supported by the underlying data.
The complaints that escalate are also, increasingly, the ones with the highest stakes attached. A SIM-swap fraud complaint that reaches the Communications Authority sits against the backdrop of an active class action; a billing or deactivation complaint to ICASA's committee can establish a pattern the regulator then pursues. The market-conduct function is, in effect, drafting documents that may be read in the context of regulatory enforcement, and it needs them grounded in the record rather than in a hurried reconstruction.
The first-line reality makes this harder than it sounds. The contact-centre agent who first handles a dispute is working at speed, across systems that were never designed to present a single view of the customer, and the note they leave becomes part of the record the operator may later have to defend. An inaccurate or incomplete first-line note does not stay contained; it propagates into the escalated response and into the regulator's file. The quality of the conduct defence is set long before the matter reaches the market-conduct function — in the quality of the record the front line was able to create in the first place.
The pattern holds across the continent's regulators. Nigeria's NCC runs its own consumer-affairs machinery, and in every regime the escalated complaint becomes a written record the regulator can place alongside others. That cumulative reading is the real exposure: a single billing complaint is a nuisance, but a sequence the regulator can characterise as systematic — unfair deactivations, mishandled fraud claims, opaque charges — is the basis for a market-conduct finding against the operator as a whole. The function is not defending individual complaints so much as the shape of the picture they form, and that picture is built from the consistency of the operator's own answers over time.
The regulator is not reading your intentions. It is reading the gap between your answer and your own records.
Where each sits.
Akki holds the full interaction history as a governed substrate — the billing events, the service decisions, the contact-centre record, the relevant network and fraud logs — so that a complaint can be reconstructed from one coherent record rather than stitched together from systems that do not talk to each other. The account the operator gives the regulator is built on what actually happened, traceably.
Solva drafts the response to the regulator grounded in that record, and refuses to assert facts the record does not support. Where the operator's position would require a claim the evidence does not back, Solva surfaces the gap rather than papering over it — which is exactly the discipline that prevents the damaging inconsistency between the operator's answer and its own data. Underneath sits the audit trail the regulator's committee may ask to see.
Complaints handling is full of personal data — the complainant's identity, their account, often their financial detail — and it crosses into the regulator's hands. SyniSense keeps that identifiable detail inside the perimeter, allowing the matter to be reasoned over and the response drafted without exposing the complainant's data beyond where it lawfully needs to go. The conduct response stays compliant with the data-protection regime even as it answers the conduct regulator.
For the market-conduct function, the response to the regulator is grounded by construction. It is drafted from a coherent record rather than reconstructed under pressure, and it declines to assert what the record does not support, so the dangerous inconsistency that turns one complaint into a wider inquiry is engineered out.
For the contact centre, the same record that answers the regulator improves the first-line resolution. An agent working from a coherent interaction history resolves more disputes before they escalate, which shrinks the funnel reaching the regulator in the first place — the cheapest complaint to handle is the one that never leaves the building.
For the customer, the resolution is faster and better grounded, because the operator is reasoning over a complete record rather than a partial one. In a market where complaints are increasingly about fraud and the fear of it, a conduct process that visibly takes the record seriously is part of rebuilding trust.
For the regulator, the operator presents as one whose conduct responses are consistent, evidenced, and traceable. That standing matters: a regulator's view of an operator's market conduct is cumulative, and a function that reliably produces grounded answers shapes that view in the operator's favour long before any single complaint is decided.
For the operator's regulatory standing, the cumulative picture turns favourable rather than adverse. When each escalated response is grounded and consistent, the pattern the regulator assembles is one of an operator that handles disputes properly — the strongest position to hold before the body that decides whether a sequence of complaints amounts to a conduct problem worth pursuing.