Users in South Africa may experience sudden loss of connection, for instance, late on a Tuesday evening, without the operator providing any notification regarding the cause or expected restoration time. After restarting equipment and checking the application, nothing changes. Contacting the call center only results in receiving a ticket number and a promise of a callback that never materializes. Two days later, the service is restored as quietly as it disappeared.
The Reason Behind Operator Silence
Although network failures are inevitable in a country with uneven power supply, the failure itself is not an indicator of the operator's quality. However, silence during such outages can become a decisive factor for the customer. The problem lies not in management ignoring the situation, but in the structural characteristics of the operators themselves, which impede timely communication.
Structural Deficiencies
The first deficiency relates to call centers. The industry has long sought to minimize costs associated with customer interaction by outsourcing this function. Outsourced centers are oriented towards contract metrics that encourage processing speed, contact cost, and call closure, rather than problem resolution. Consequently, the agent merely logs the failure, provides a number, and closes the ticket, creating the impression that the customer was processed rather than helped.
The second structural barrier is the organizational structure. Specialists who know about a tower malfunction and employees who interact with customers are in different departments, report to different managers, and use different systems. Network maintenance operations are evaluated based on uptime, whereas customer communications (under marketing or care) are assessed based on campaigns and sentiment. Failures occur at the intersection of these two spheres, and no one is responsible for informing customers about the breakdown and its expected resolution time.
Technical Communication Limitations
The third factor is the technical infrastructure. Even operators striving to provide proactive updates often cannot do so because the fault detection systems were not initially integrated with customer communication systems. Fault management platforms notify technicians and track repairs, but they are not designed for mass messaging to all affected users. Integrating these systems competes for budget with projects that yield more obvious revenue. Thus, scalable communication during an outage is absent for most operators.
The combination of these factors transforms silence from a sign of negligence into a predictable consequence of the architecture: a call center built for cost savings, no owner for the customer problem, and no pathway from logging a failure to informing the user. This system reliably generates silence, regardless of management's sincere claims about customer care.
Comparing Network Metrics and Sentiment
There is a way to test this hypothesis: if complaints accounted for network quality, the operator with the most negative network interaction would have the lowest measured network quality. However, the situation is the opposite. According to Opensignal's 2025 report on South Africa, based on billions of phone measurements, MTN won 11 out of 15 awards, including all overall experience awards. Rain received recognition only in the voice communication category. Ookla recorded MTN's median mobile download speed at 74.76 Mbps, the highest figure in the country, and Icasa registered national 4G population coverage at 99.5%. This indicates that the bottleneck is not the network itself, and MTN maintains leadership across various benchmarks.
However, when compared to customer perception, the picture changes drastically. The operational customer experience score from DataEQ is almost directly opposed to network quality ratings. Rain, which ranks last in network quality, shows the highest sentiment among operators—+16%. MTN, leading in the network, shows a score of -17%. The better-measured network lags behind the worst in how its customers feel. Cell C also has a positive score of +11%, and here the comparison is cleanest: Cell C does not own its own network but uses MTN and Vodacom networks, so its customers primarily use MTN towers.
When the same towers yield opposite results depending on the name on the bill, it means the measured metric does not reflect the towers themselves. This gap needs careful examination, as the two datasets measure different things. Benchmarks assess speed and coverage via devices, while DataEQ measures user feedback, which includes billing, pricing, fixed-line services, and repair difficulties.
What Truly Influences Satisfaction
Therefore, customer sentiment reflects not the network itself, but all the surrounding factors. The report indicates exactly which aspects are causing damage: network coverage is rated at -27%, which is poor but tolerable, while customer service is -85%, account administration is -80%, and the call center is the worst channel in the industry with a rating of -84%, with only 39% of urgent messages receiving any response.
The network is not the reason for decreased satisfaction; it is the service surrounding it. It should be noted that Cell C's score could have been inflated by a coverage campaign, which is narrative control, not service improvement. Rain demonstrates a clearer signal, as its rating is based on the ability of agents to resolve issues the same day. Customers value the entire network interaction experience: getting a response, being informed, and having the issue resolved.
Ways to Improve Service
Solving this problem turns out to be cheaper than it seems. If the system generates silence, the system itself must be changed, not just the messages: a single communication owner must be assigned during an outage with authority over all operations and service areas; the contact function metrics must be changed from focusing on throughput to focusing on resolution; and the failure systems must be integrated with customer systems so that status updates happen automatically, rather than requiring heroic effort. None of this requires new spectrum or a complete network overhaul. Communication during an outage must be viewed as an integral part of the product. Operators who continue to view this as a secondary issue do not lack goodwill; they lack ownership, incentive, and necessary connectivity.