The UK comms regulator and the firms that audit the accuracy of British phone bills need help.
It is increasingly important for communications providers to complete regular testing of new equipment and devices to ensure that call duration and other events continue to be measured within the standards required by the Metering and Billing Direction. It is similarly important for communications providers to ensure they are appropriately monitoring the accuracy of existing equipment and devices that customers are already using. We therefore request that, to the extent it is not already, the testing of call duration and other event measurements are treated as areas of focus when changes are going to be made, and new equipment is to be introduced.
This comes from a recent public letter issued to comms providers by Ofcom, the UK regulator. It is a roundabout way of admitting there is uncertainty about how to cope with new technology, both from the regulator and from the small number of businesses that have earned a steady income by performing regulatory audits of British phone bills over the last quarter. The challenge they now face was somewhat predictable. These were people who thought that knowing about Ericsson AXE10 switches represented the height of technological expertise in this domain, but who then applied the mindset of ISO9000 quality auditors to every other aspect of how data is conveyed and processed from switch to bill. Seasoned revenue assurance veterans hardly need to be told why switch to bill reconciliations are less important than they were: the amount charged to customers is far less influenced by the duration of calls than previously. Now the old switch-to-bill paradigm for these audits is on the brink of collapse as the UK approaches the imminent switch-off of all its PSTN networks.
Most other countries are not facing the same challenge as the UK. Partly this is because they do not have a hard deadline for switching off PSTN. Mostly it is because they never adopted the unusual phone bill audit regime mandated by the UK. There is a strong argument for claiming the UK’s Metering and Billing Direction, as the regulation is now called, was an important influence during the 1990’s on the worldwide development of revenue assurance as a recognizable discipline. Like revenue assurance, it is now struggling to understand its place as the world around it changes. Revenue assurance is dying; these billing accuracy audits could die too, although their demise will likely take longer because a regulator is involved. Too much of the assurance work has been associated with specific facts about specific technologies, such as determining whether there is any drift in the internal clocks of particular makes of switches and how those clocks are then periodically reset by operators to align with an external time reference. Questions like that become academic if the rates applied to calls are the same at any time of day.
The business model of most comms providers is increasingly oriented around exploiting software, not hardware. This reduces costs but it also creates new revenue possibilities too. Tariffs have generally become more simple, but new capabilities like network slicing also introduce the potential for new complexity. To use an analogy, quality auditing serves a clear purpose when applied to the manufacturing line for motor cars, where there is supposed to be very high consistency in the end product. Production lines do not change much but the throughput makes it worthwhile to spend money on making changes that eliminate errors. It is harder to apply the same methodology to the many software updates that are applied to Teslas and other modern cars. Software has become a far more important component of cars than before, and the same is true of communications. Quality auditors will disagree, but it is very difficult to do quality audits of software. Different techniques are needed to find and minimize bugs. Software has more influence over the price charged to a customer than ever before, and will continue to grow in importance as the ‘metering’ side of ‘metering and billing’ systems become as grounded in software as every other aspect of the business.
Earlier this year, the UK government announced a new approach to regulation that is meant to support growth. They reasoned that:
Over the past few decades, an excess of poorly designed regulations has resulted in unnecessarily complex and burdensome requirements for businesses and investors. Even if all these regulations had been optimally designed, their sheer cumulative impact — and the associated increase in regulatory activity — has produced an additional layer of burden which adds to this complexity. Whilst each regulatory intervention has been rationalised in its own terms, the unintended consequences of the cumulative effects have not been properly analysed.
Sadly, Ofcom did not treat this as an opportunity to repeal its Metering and Billing Direction and to start again with new rules that are more fit for purpose. The argument for repeal is straightforward: most other countries do not have the mandatory audits that UK comms providers must pay for, and they do not have more complaints about the accuracy of their bills. The data used to justify the UK’s audits has always been weak and has consciously ignored major audit failings, including one which allowed the systematic overcharging of customers over a period of 15 years. There is a need for consumer protection, but not in the way that the UK is trying to protect consumers. This is easiest to see when considering how the methods used by fraud managers are increasingly diverging from those used by revenue assurance.
There was always a straightforward rationale for telcos combining their revenue assurance and fraud management teams into combined RAFM units. Both teams look for anomalies in the same data. There are some differences in the character of the anomalies they look for, but those differences are not so great to override the significant cost savings by using a common provider of RAFM systems. It is notable that fraud management — including the anti-scam requirements imposed by regulators — is increasingly pinning its hopes on using artificial intelligence to find data anomalies indicative of crime. Meanwhile, revenue assurance is being left behind, and the regulatory audit of UK phone bills persists with a philosophy that originated in the 1990’s, when telco employees would literally read meters to verify the accuracy of recorded durations. The UK’s Metering and Billing Direction was the successor to a regulation that only covered metering, and this latest letter reveals that it remains fixated on metering as an activity, even though the notion of testing a literal meter no longer applies in practice.
None of this means that consumers should not be protected from inaccurate charging. The point is rather that audit work needs to be efficient and effective if the public is to be properly protected. Asking comms providers to have more ‘focus’ is a sign that the people paid to write the regulations and do the audit work lack the precise conceptual tools needed to make their work more efficient and effective. Almost 25 years ago, decision-makers in the UK openly stated that they expected the UK’s regulatory audits of phone bill accuracy would be copied elsewhere. Their hopes came to nothing. There is some irony in the fact that a German-owned business does most of this audit work in the UK but that the German regulator chose not to copy the UK’s regulations. Some nation may potentially adopt a radically different approach to consumer protection, perhaps by using AI to link the search for errors with the search for illegal traffic. There are countries which are currently investing a lot of faith in putting software on phones to stop crime. That is more of a technological challenge than installing software to prevent a phone user being overcharged for the services they actually consumed. As stated above, one reason to like software is that it can reduce costs while improving performance. The same logic can be applied to consumer protection audits too.
Ofcom’s open letter asking for a ‘focus’ on testing call durations and other event measurements can be seen here.
