PSA: Don’t Trust Analyst Estimates More Than A Year Out

Oh, analysts. When will we ever hear the end of your silly claims? The latest round of analyst shenanigans comes from Pyramid Research who estimates that Windows Phone 7 will be the top smartphone operating system by 2015, displacing Android as the top-selling smartphone platform. Now, for all we know, that could be the case. On the other hand, shall we take a stroll down memory lane to see how well other analysts’ expectations have fared?

October 2009: Analysts from Gartner estimate that, by 2012, Android will rank second in smartphone sales behind Symbian. In reality, Android took the lead for smartphone sales at the end of 2010, while Symbian has been abandoned by Nokia in favor of WP7 in the long term. Gartner also estimated that “[they] expect Android to run on nearly 40 different handsets by 2012.” Today we call 40 Android handsets a slow start to CES. It’s also worth noting that Ken Dulaney of Gartner pointed to the Motorola CLIQ as an example of how well Android performs. Ugh. I mean, I know the Droid hadn’t come out yet, but come on. At least look at the Hero or something.

July 2009: Research firm Generator Research predicts nearly two years ago that Nokia’s marketshare will be reduced to roughly half of what it was then by 2013. This would, in turn, lead to Apple finally passing Nokia’s marketshare by 2011. To be fair, this may yet be the case as Nokia has abandoned Symbian and who knows how long it will take for the die hards to die off, however it’s yet another example of an event analysts simply can’t foresee. For what it’s worth, they also predict that in 2011, Apple will sell 77 million iPhones. In Q1 of 2011, Apple sold 16.2m iPhones which means, in order to reach 77 million units by the end of the year, they’ll have to either sell 20m units every quarter for the rest of the year, or sell two million more units each quarter than the previous one. This is not entirely impossible, of course, but we’ll still have to wait and see.

August 2009: MSN Money analyst estimates that Apple stock will hit $335/share by 2012. Apple’s stock price today is $347 a share. Of course, this is the kind of thing that an investor likes to see an analyst get wrong. Still, wrong is wrong and it could just as easily go the other way.

December 2010: It’s nice, occasionally, to see some folks at least see analysts for what they really are. Here, Business Insider notes that some analysts were crazy to predict that the Verizon iPhone wouldn’t help Apple sell any new iPhones. Of course, equally crazy was the notion we all heard that the Verizon iPhone would be the death of AT&T, the death of Android, and the death of basically anything not a Verizon iPhone. Turns out, ultimately, some people just like a Verizon iPhone. Maybe a few more than like an AT&T iPhone, but really, both carriers were activating roughly the same number of iPhones (BGR pegs the number at 277k per week for AT&T and 314k per week for Verizon during Q1 2011). It’s also worth noting that, prior to the advent of the Verizon iPhone, Android was outselling the iPhone domestically by a factor of roughly 5:1*. U.S. iPhone sales roughly doubled after the Verizon iPhone bringing the ratio to about 2.5:1, with no expected effect on the global market. Oh, yes, and AT&T, by some counts, is larger than Verizon now. So, it doesn’t look like anyone died.


Ultimately, analysts are generally pretty bad at predicting the future, especially when that future is more than a few months out. The shorter the span between prediction and fruition, the more likely it will be accurate, but even then analysts are just as susceptible to hype as the rest of us. There’s a few numbers that are generally trustworthy. Sales numbers given by the companies themselves. These numbers can be (and frequently are) framed in the most flattering, if not the most honest way, but they’re at least generally accurate. Also, marketshare numbers from firms like comScore, Nielsen, and Canalys. These tend to be fairly stable and accurate. The numbers differ slightly from one analyst firm to another, but generally speaking, trends are at least consistent (i.e., comScore and Nielsen both reported when Android overtook the iPhone, even if they disagreed somewhat over by just how much).

There’s a certain fine art to analyzing smartphone market share (or any business analysis) and it can get pretty confusing at times, but remember to take every number you hear with a heaping helping of salt, trust reported numbers more than projected numbers and never, never, ever make your stock investment choices based on what an analyst predicts will happen three or four years from now. Ever.

* — In November of 2010, comScore reports 61.5m smartphone users in the U.S., of which Apple had 25% and Google had 26%. This is equivalent to 15.375m iPhone users and 15.99m Android users. December 2010, the smartphone market grew to 63.2m users, of which Apple had 25% and Google had 28.7%. This is equivalent to 15.8m iPhone users and 18.134m Android users. According to these numbers, in the time between November and December’s records, Apple added 0.425m new users in the U.S. while Android added 2.144m new users, for a ratio of roughly 5:1 new Android users to iPhone users. This trend was common (allowing for minor variance) for much of the latter part of 2010 as well as the beginning of 2011 by comScore’s numbers, up until the advent of the Verizon iPhone. comScore has not yet released any U.S. sales numbers beyond the month of February which saw the Verizon iPhone’s release.

Source: BGR

2 thoughts on “PSA: Don’t Trust Analyst Estimates More Than A Year Out”

  1. Too many typos and I’ve been too lazy to go back and edit it, but I said as much here:

    Also, you see why I make fun of analysts and refer to them as sorcerers and charlatans.

  2. Pingback: comScore Reports March Marketshare Numbers, Android, iPhone Up, RIM Down, Water Wet

Comments are closed.

Scroll to Top