Almost daily there’s a press release or marketing deck touting some analyst firm’s massive figures for anticipated growth in this or that market sector ($X billion, $Y zillion). Most of these are built on assumptions only loosely grounded in reality.
These numbers are almost always inflated because nobody gets media attention with moderate growth. Accordingly, there are many inherent biases in the business of forecasting.
For these and other reasons — a market is too young or speculative or some necessary development doesn’t happen — forecasts more often than not turn out to be wrong. Nobody’s really keeping score so we only see the numbers as they’re released but almost never in hindsight.
However I came across a fascinating example of a forecast that was way off other day: the 2011 IDC smartphone forecast for global markets.
Global smartphone market share 2011 – 2015
Source: IDC smartphone forecast 2011 – 2015
The firm estimated that by 2015 Windows Phones would overtake the iPhone in global market share. It also assumed that BlackBerry would still be going strong in 2015.
Windows Phones were supposed to have 20.3% global share by 2015 and BlackBerry was going to suffer only modest declines, from 14.2% in 2011 to 13.4% share by 2015. Instead BlackBerry is dead and Windows Phones are all but dead. Globally Microsoft’s mobile share is well below 2 percent and CEO Satya Nadella has said that the company’s current mobile market position is “unsustainable.”
Indeed, Microsoft is rumored to be introducing a “Surface Phone” aimed at creating a new smartphone brand for enterprise customers. In 2015 Microsoft wrote off its Nokia acquisition to the tune of $7.6 billion.
Notwithstanding any impending Surface Phone there are two operating systems in the smartphone market: iOS and various flavors of Android. It won’t always be this way of course but for now that’s the market. Whatever assumptions IDC made about enterprise-driven smartphone purchases or other assumptions proved to be wildly inaccurate.
In fairness to IDC I’ve created forecasts myself that were way off (an early PPCall forecast comes to mind, which assumed Google would roll it out). But this is an interesting example of getting something very wrong in more than one way.
I suspect if we went back and exhumed numerous projections and forecasts we’d find more than 75% of them would be off in some significant way. These forecasts take on a life of their own as marketers pick up those that suit them, pop numbers in decks and give them credibility through repetition.
The only problem is that most of these numbers often have little relationship to reality.