The Apple Watch. Photo: Apple.

Research firm IDC believes Apple's (AAPL 1.27%) watchOS platform will represent about half the smartwatches sold in 2019, while Alphabet's (GOOG 0.74%) (GOOGL 0.55%) Android Wear accounts for just under 40%. Other platforms, including Samsung's (NASDAQOTH: SSNLF) Tizen and Pebble OS will attract only modest interest from buyers, according to projections the firm released earlier this fall.

IDC remains one of the top third-party researchers when it comes to mobile computing devices -- its quarterly reports on the smartphone and tablet markets are widely cited. Yet, when it comes to projecting future market share, its track record is stunningly bad, and investors shouldn't put too much faith in its outlook.

Missing on tablets and smartphones
In 2010, IDC attempted to predict what the smartphone market would look like in 2014. Smartphones were considerably less popular back then (about one-third of U.S. adults owned one), but the market was hardly nascent: Apple's iPhone had gone through a number of iterations, and Samsung had just released the original Galaxy S. Here's the chart from IDC's September press release:


IDC believed BlackBerry and Nokia's Symbian would become less popular by 2014, but not to a major extent. Meanwhile, Microsoft would carve out a nice business with Windows Mobile, and, while Alphabet's Android would grow, it wouldn't dominate. "IDC believes the market will comfortably support up to five OS players over the next five years," said IDC's Restivo.

In reality, the opposite was true. By the second quarter of 2014, the smartphone market had coalesced around just two players -- Apple's iOS and Alphabet's Android. The iPhone accounted for around 11.6% of the smartphones shipped in the second quarter of 2014, while Android powered nearly 85%. The remaining share -- less than 4% -- was split mostly between BlackBerry and Windows Phone.

IDC's projections for the tablet market fared slightly better, but not to a great extent. In December, 2012, IDC offered up a prediction for the 2016 tablet market. Here's the chart from that release:


Given that it's still 2015, it's possible that these projections could prove accurate. However, it would require a rather remarkable shift in the tablet market. Apple's iPads are on pace to account for around 25% of the tablets shipped this year, while Android powers about two-thirds. Windows tablets have carved out a nice niche -- with around 8% market share -- but could still fall short of IDC's forecast.

In both instances, IDC greatly underestimated Alphabet's mobile operating system, and gave incumbents too much credit.

The smartwatch market is far more complicated
Now here's the chart of IDC's estimates for smartwatches in 2015, and what the market could look like in 2019:

IDC does not expect Samsung's Tizen-powered Gear watches to succeed. Apple's watchOS will see its popularity dip, but it will remain the most popular platform by a sizable margin. Alphabet's Android Wear will grow, but won't come close to replicating the success the company has enjoyed in smartphones and tablets. 

Smartwatches -- including the Apple Watch, Android Wear devices, and Samsung's Gears -- require a paired smartphone to function. The Apple Watch, in particular requires an iPhone, while Samsung's Gears require Android phones (and in some instances specific models from Samsung). Android Wear devices are compatible with the iPhone, but only in a superficial sense.

If this situation persists, the sheer quantity of Android handset owners suggests that Android Wear will become the most popular platform. In order for Apple's watchOS to remain in first place, either Apple's iPhone would need its market share rise significantly, or most Android handset owners would have to ignore the smartwatch category altogether. If Samsung's Tizen fails to find mainstream adoption, it's hard to believe the company would continue to support it over the next four years.

IDC's projections seem like safe assumptions given the current state of the smartwatch market, but its outlook will almost certainly be wrong.