The 12 Quadrillion Euro Phone Bill (and what it says about telecom arrogance)

A quadrillion is not a made-up number. It is a real number: a 1 with 15 zeroes after it. The euro is not a made-up currency. For the moment at least, it is the legal tender in 17 European countries. Phone bills, on the other hand, need have no basis in the real world. That is the only conclusion left to draw, if you feel no shame after hearing that Bouygues Telecom issued one unfortunate customer with a bill for 11.721 quadrillion euros.

Common sense should indicate that this bill was inaccurate. It should indicate a mistake was made. After all, this is what 11.721 quadrillion looks like when you print it out with all the zeros, just as Bouygues Telecom did: 11,721,000,000,000,000. I hope most people would suspect something was wrong by the twelfth zero, if not by the ninth zero. After all, this bill was sent to Solenne San Jose, an unemployed child minder, as a penalty for the early termination of her mobile contract. You might think this penalty was somewhat unjust and should be overturned. Or you might wonder how Solenne had signed up to such a terrible tariff for her mobile phone. You might also think that Bouyges Telecom must be headed for a bumper year, given that their annual revenues are normally around 5 and a half billion euros, a mere 0.00005% of the one-off charge they presented to Solenne. And you might think this was a bumper year for the French economy. The tax generated by this one bill would easily settle France’s national debt of 1.7 trillion euros. Indeed, this phone bill would make it a bumper year for the global economy, which is otherwise worth only 50 trillion euros a year. So common sense might make you think the number was wrong. But not if you work in the la-la-land of telco customer service. If you work in customer service, then you can end up behaving like a phone bill is never wrong. As Solenne explained to the journalists at Sud-Ouest:

The first time I called, I spent at least 45 minutes with one operator. He replied: “it’s automatic, I can do nothing.” Another told me that I would be contacted to arrange payment in several installments.

In contrast, Dr. Evil only dared to ask for 100 billion dollars, and they laughed at him:

So there you have it. The people working for Bouygues Telecom are even sillier than an Austin Powers comedy. But I think there are two serious points that need to be made.

First, the size of an error may have nothing to do with what causes the error. That fact is sufficient to blow a quadrillion dollar hole through any theory that says telco staff should try to forecast their transaction risk and error using some standardized cross-industry model. It is impossible to reliably predict the scale of transaction errors that your telco will make. It is silly to try to extrapolate transaction error risk from one telco to another, or from a so-called standard model to a specific telco. Forecasting the benefit of error prevention is like calculating the reduction in crime that occurs because the police and court system acts as a deterrent to crime. You cannot do it because, unless you can travel between parallel universes, you never have objective data as a basis for comparison. But that does not mean that error prevention is not worthwhile. We can reason that prevention is often better than cure, but we cannot measure the specifics of when or what kind of prevention activity will deliver benefits that are greater than their costs. By their nature, errors are caused by unpredictable oversights, unforeseen omissions, and unimagined slip-ups. Preventative steps may eliminate errors we anticipate, and also those we did not anticipate. Neither anticipated nor unanticipated errors are susceptible to reliable and objective quantification if we prevent them before they occur. Nobody can build a statistical model that tells them things where they have no relevant data for their own telco. It would be foolish to try to extrapolate from a model based on the made-up data of somebody else’s telco. Anyone who claims they can predict quantified error variances for telcos – variances that must cover both ups and downs – is welcome to examine the circumstances of this 11,721,000,000,000,000 euro variance, and then think again. I look forward to examining the ‘science’ behind a statistical model that claims to take account of the true diversity of goofs and cock-ups, and the extremely varied consequences that flow from them. Anyone offering such nonsense as an industry standard will be welcome to read my scientific retort: the number 11,721,000,000,000,000 as written on my backside!

Second, one has to bemoan the inflexibility of a mobile operator where this bill had to be escalated to the Director of Customer Relations before anyone would admit a mistake had been made. Apparently the bill should have been for €117.21. Because of the distress (and bad publicity) that had already been caused, the bill was waived completely. But why should this unlucky woman find it so difficult to get through to someone who can properly respond to this obvious mistake? Why are the customer services staff unwilling or unable to escalate this problem instantly, rather than behaving as if nothing was wrong? And what would have happened if the error was smaller, like charging €171.21 instead of €117.21? To what extent is the intransigent reply of Bouygues Telecom customer services – “it’s automatic” – used to bully other customers into accepting overcharges? Why even bother with human beings in call centres, if the answers are all “automatic”?

The maths is bad for telcos. Unlike so much nonsense which gets bandied around the industry, this quadrillion euro error is a matter of fact. When confronted with inconvenient but objective data, it gets ignored when it does not fit the prejudices of people working in the industry. Customer service staff insist that the charge is “automatic” as if that provides an explanation or an excuse. RA stooges and national regulators build so-called models to forecast how much error they think will occur in future, but they never adjust their models to account for data like this. If I asserted that telcos lose billions of dollars each year through underbilling, how many would believe me, even though I showed no data to support my assertion? But if I asserted that telcos overcharge customers by quadrillions, it would be laughed off, even though Bouygues Telecom has demonstrated that magnitude of error on a single bill. If a quadrillion-dollar error can occur on just one bill, it follows that much smaller errors could be cumulatively adding billions of dollars to telco profits each year, in much the same way that cumulative error could be subtracting billions from the bottom line. So next time you see a model that calculates the risk that the telco will be a loser as a consequence of undercharging its customers, ask yourselves about the model that calculates the risk of the telco being a winner, because it succeeded in overcharging its customers. And if there is no counter-balancing model, or if the net effect is that telcos almost always lose, and that they hardly ever win, then ask yourself about the data that justified this conclusion. Is it all the data that really exists, or only the data that people want to find?

Eric Priezkalns
Eric Priezkalns
Eric is the Editor of Commsrisk. Look here for more about the history of Commsrisk and the role played by Eric.

Eric is also the Chief Executive of the Risk & Assurance Group (RAG), a global association of professionals working in risk management and business assurance for communications providers.

Previously Eric was Director of Risk Management for Qatar Telecom and he has worked with Cable & Wireless, T‑Mobile, Sky, Worldcom and other telcos. He was lead author of Revenue Assurance: Expert Opinions for Communications Providers, published by CRC Press. He is a qualified chartered accountant, with degrees in information systems, and in mathematics and philosophy.