M-Pesa – a mobile-phone based money transfer and microfinancing service

First, even in your correction, you're still making a pretty useless point that is misleading to people who don't understand the nuance of the definition of GDP, don't have an understanding of MPESA (i.e. the target audience of your comment), and implies mpesa is far more significant than it actually is. i.e. it misleads, whether you like it or not. It's so misleading you probably took it as a fact when you read it the first time, and only saw the mistake when I told you it was wrong and you looked it up. Why not correct your statement with a useful quantifiable answer?

So why is it a pretty useless point? Well look, if I hand $1 to my brother, and he hands it back to me right away, and we do this 100000 trillion times on a digital money system we designed (say, a simple database with a simple algorithm to execute all these trillions of payments), we'd be able to say, 100% truthfully, that our silly payment system has a total volume of transactions that far exceeds GDP, or the entire world economy in this century so far, for that matter. It's a meaningless statistic, and it's an apples/oranges comparison because there was 0 added value in any of the transactions between me and my brother. GDP measures added value, mpesa transactions measured in volume don't. Why say 'mpesa's non added value figure is x percentage of the country's added value figure' at all?

For example, take sinopec. Chinese company with nearly $0.5 trillion in revenues. Imagine its profit margin is say about 5%, then that means that it has nearly 0.5 trillion in expenses, too. So this company processes about $1 trillion in value, every year. The chinese GDP is about $7 trillion. It'd be completely to say this single company which employs 0.02% of the Chinese population, processes money totalling about 1/7th of China's gdp, even though it's a factually true statement that's completely meaningless, when the real contribution to the 7 trillion Chinese gdp is actually just about 7 billion.

Either you express a statement of how much of added value is transacted on mpesa and take that as a percentage of total added value (GDP), or you measure the percentage of total volume that is transacted on mpesa, as a percentage of the total volume of all payment systems in the country, without getting into added value/gdp at all.

If you do so, you'll find mpesa is pretty small, the Kenyan central bank has pretty decent data. For example, in August mpesa handled about 250b kes in value, a tiny bit less than ACH, and an order of magnitude less than RTGS in Kenya which handled about 2.8 trillion. And those are just the electronic payment systems where mpesa looks to be less than 7%. Let's not forget the most recent reports are that 94% of transactions in Kenya are made with cash (which, expressed in a silly 'volume of cash transactions as a % of gdp' metric, would constitute a figure that's likely more than a tenfold of GDP, but again it's meaningless).

And I'm not even getting into the fact that a large portion of economic activity in Kenya isn't part of GDP, as the grey economy often is substantial in developing countries, or the large amount of double counting in the figure you're referring to.

Mpesa is awesome, I'm glad it's here, it'll keep on growing and become more important with time. But let's not sprinkle these bs apples/oranges metrics around on HN, the economist and the like take care of that just fine on their own. The mpesa story doesn't need this hyperbole, and its growth statistics are impressive and promising on their own.