They are spoken of less often these days, but we still keep hearing about those pesky ‘estimates’ of ‘average’ leakage. RA software vendors, obsessed with the profit (or loss) they will be reporting to investors each quarter, like to exaggerate the prevalence of leaks to help secure a sale, much like a burgular alarm salesman is going to exaggerate the number of robberies if it helps him increase his commission. Exaggeration can generate a sale in the short run, though in the longer run, RA vendors inevitably get undermined by the observation that industry estimates never go down. ‘Estimates’ of leakage remain steady or go higher every year, no matter how many hundreds of millions of dollars is spent on RA software in the meantime. That rather implies that the net benefit of RA is negative: every year the telco industry spends more on RA, yet every year the estimates of leakage get worse!
We all know the reasons why the numbers are not reliable in the way presented, but my favourite bugbear is the idea that there is an “average leakage” for intercarrier traffic. Think about it. Take every telco in the world. Estimate the net error when they pay each other for intercarrier services. What is the mean average leakage? Zero. We know that as a fact. We know it without estimates or data or guesswork or case studies or anything. If the purpose of the estimate was to get an average across all telcos, and we are only talking about intercarrier, then we know that one telco’s loss is equal to another telco’s gain. In short, errors in intercarrier billing and settlement are a zero sum game. If one telco is being charged too much, another is earning too much revenue. If one telco receives less than it should, another is paying for less than it used. Whatever gibberish you spout, there is no way to identify an average ‘leakage’ without also admitting to an average ‘gain’ that is equal and opposite. Yet vendors are so addicted to marketing spiel, they cannot stop themselves. Bang goes their credibility as they try to tell us a mathematical impossibility has been proven thanks to their very unscientific (and very convenient) observation. I just found this gem, courtesy of a cVidya-Stratecast report:
In the wholesale market, inter-carrier revenue leakage represents approximately 3 to 5 per cent of service provider’s total costs for traditional voice products, and in the range of 7 to 11 per cent for broadband products.
Perhaps cVidya are trying to say that service providers are always on the losing end of intercarrier errors. I bet they do not say that to their customers who sell traffic on a wholesale basis! To be fair to cVidya, they are not alone in falling into this trap. I once blogged about similar interconnect leakage hyperbole from Subex, though I am heartened to say I have never caught them repeating that mistake. And I know that cVidya’s people regularly read talkRA (IP works both ways – allowing us to tell where the readers are). So hello to everyone at cVidya, and please stop pretending that intercarrier error is all losers and no winners…