I have never been a big fan of predictions. Most strike me as resulting from a few back-of-envelope calculations combined with a knowledge of what the industry is expecting to hear. And I say that with a certain authority as I was involved in calculating predications for the electronics industry for a magazine I worked for in the 1980s.Basically, we rang up as many companies as we could, asked them what they thought would happen in the next year, averaged out all the figures and gave it back to them as forecasts. And, you know what, it was mostly right.

However, occasionally somebody gets it so wrong that people start to have questions, and that happened this month when IoT connectivity firm Eseye hit the headlines when it called out some of the excessive predictions that we saw about the growth of the IoT. In particular, the company picked up on then Ericsson’s CEO Hans Vestberg who said in 2010 that his vision, remember that word, it will be important later, his vision was that there would be 50 billion connections by 2020. Cisco a year later made a similar prediction. Maybe they asked Ericsson.

Well, we are here, it is 2020, so how do those predictions stand up? Not well. Eseye says the true figure now is closer to nine billion. That is a big difference, we are talking 41 billion short of the actual figure.

I contacted Cisco and Ericsson about this. Cisco never got back to me, but Ericsson did in the form of Stephen Carson, the company’s strategic marketing manager.

First, he disputed the current nine billion figure, as he said that just referred to subscriptions, which are standing at 8.7bn. What it misses out are cellular M2M, which adds another 1.7bn, and the 12.5bn non-cellular connections. That brings the total up to nearly 23bn. A lot more respectable, but still less than half the 50bn that Ericsson was talking about a decade ago.

So still not very good really. Not so, said Carson. The 2010 figure was not a prediction, he said, but a vision. I told you to remember that word.

He said the company didn’t start making real predictions until 2015, when those predictions were basically right.

Isn’t this a bit like the weather forecast where they keep changing it as they get nearer so it always looks good?

No, in 2015 these were proper predictions, he said, based on serious industry analysis. The 2010 figure was not a prediction but a vision. Ah, now I get it. For vision, he means guess.

But why was this guess or vision so far out? Well, apparently, Africa is to blame. I kid you not. The 2010 vision was based on the assumption that just about everyone would have about ten connected devices to their home router or mobile phone. And while that is true in many developed nations, it is not the case in Africa and other less developed places. That is why the numbers are so out.

If Africa had pulled its socks up and started connecting everything to the internet, instead presumably of worrying about all their pressing economic problems, Ericsson’s vision would have come true. Carson didn’t say why Ericsson didn’t see Africa’s lack of progress in its vision. They probably didn’t ask the right people.