Why Every Estimate for Apple Watch Sales in 2015 is Completely Wrong

When I heard that RBC Capital Markets predicted that Apple would sell 20 million Apple Watches in 2015, I was highly skeptical. Now, a new report from Morgan Stanley puts the prediction at closer to 30 million. The gap between these two estimates already suggests that little can be truly known at this point.

Allow me to explain.

The problem with these predictions is that no methodology is ever presented. Nor is there an estimated error, a plus or minus number. One of the tip-offs that little mathematical methodology was used is the roundness of the sales estimate and the lack of an estimated uncertainty. Instead, a nice round number, like 30 million, is presented with the implied assertion that the number isn't far from reality.

Of course, the analyst can say that he or she can't be more precise than that, but that's because no estimated number fell out of the mathematical analysis which does tend to produce a number with a few decimal places. However, the way out of that is to also cite the error analysis. All these estimates make assumptions, and the combination of the uncertainty of each assumption can be mathematically combined to provide an overall uncertainty, and that puts the precision of the base number into perspective. For example, it's okay to estimate 18.571 million if the estimated error is +/- 7 million. See?

As for the elements of uncertainty, there are variabilities in the number that Apple will be able to build, sapphire ability, impact of price points, the state of the economy, the psychology of the customer, number of iPhones in the market place (see page 2 below), the effects of less expensive products from competitors, how many people at any given time need (and can afford) a new watch or think they really need an Apple Watch, the influential success of Apple Pay—just to name a few.

But, in fact, we already have an uncertainty range right now. The first estimate, with its own set of likely vague assumptions was 20 million. The next one was 30 million. So right away, from these two numbers, we could use simple mathematical tools that can suggest an average and standard deviation. No doubt, we'll get a bunch more swags later, all over the map.

Which is the right number? None of them. If each estimate were a careful math analysis with an uncertainty range, we might be able to combine each estimate with its numerical uncertainty and better home in on a good estimate. But we haven't been given that.

Using good math doesn't give away any proprietary secrets. It just reveals that one is a competent mathematical analyst. For now, I would be much more impressed if a detailed methodology were presented and included a result like 16.5 +/- 5.5 million. We might actually be able to do something by aggregating a bunch of well done estimates like that. When we did, we'd perhaps find (OMG) that the resulting uncertainty is so large that nothing meaningful can really be predicted at this early stage.

Next: The tech news debris for the week of November 17. Legends of Steve Jobs, Apple Pay, Cord-cutter bliss.

Next: The Tech News Debris for the Week of November 17


After Apple released WatchKit, The Verge has put together a nice article that combed through everything that can be known about the Apple Watch. "11 things we just learned about how the Apple Watch works." Notable is the fact that an iPhone is almost always required and that native apps won't be ready until later next year.

When discussing legends about Steve Jobs, it's often hard to distinguish a dramatic demonstration by the former CEO from a stinging harshness. In this story, Mr. Jobs made a strong point that perhaps an extended argument couldn't achieve. In the end, it's a testament to the man's imagination and intuition. "Steve Jobs Dropped The First iPod Prototype Into An Aquarium To Prove A Point."

Up until recently, cord-cutters were treated as a thorn in the side of the cable and satellite companies. In the tradition of American enterprise, Verizon has done the obvious: used its powers to create a package especially designed for cord-cutters. The plan consists of high-speed Internet, local FIOS TV with Network feeds, HBO, Showtime, and a year of free Netflix for $60. This is less than the average cable bill, and it caters to customers who are tired of paying for bundled TV channels they never watch. It's a brainy idea by Verizon.

We all know, or think we know, what kind of company Apple is, its vision, and where its future lies. On the other hand, it's no so easy with Google. Jan Dawson sizes up what kind of company Google is, tries to figure out where the company is going and how it will make money in the future. "What is Google?"

This article by Jon Brodkin at ars technica is almost 100 percent about Netflix. However, hidden in the discussion is some interesting data (in the charts) about the Internet usage by iCloud, FaceTime and iTunes. It's a sobering thought that iTunes uses almost 3 percent of all Internet traffic.

Over at six colors, Jason Snell has written up a very nice review of the iPad Air 2, complete with benchmarks and a discussion of the iPad's camera. As always, Mr. Snell brings on a boatload of insights in his "iPad Air 2 review."

Have I mentioned that I love Apple Pay? Here's a good article that notable for the lists of all the participating banks, store fronts and online places that will accept Apple Pay. "The ultimate guide on how and where to use Apple Pay."

One of the predictions I have about how CEOs will evaluate Apple Pay for their own storefronts is that they'll worry about losing their most affluent customers to the competition. Jonny Evans at Computerworld has data from retale.com that backs up that surmise.

[Apple Pay] users are already three times more likely to spend $250 or more than those on other services, the latest Retale survey claims.

I am sure Apple knew this would happen. There is more to designing a product than comparison charts and a price that races to the bottom.

You have probably heard about how Yosemite treats 3rd party SSDs and how it might affect users. Here's a tutorial by Chris Breen at Macworld ""Mac won't boot? About Yosemite and your third-party SSD. And in the spirt of the care and feeding of your SSD, Jonny Evans has followed up with some good advice. Please don't gloss over this one: "Dear Mac users, please, please backup today."

Finally, Apple customers are no stranger to minor difficulties when a new version of iOS is rolled out. But in the case of Android 5.0, "Lollipop," what happened here (and the finger pointing) is ridiculous. "Android Lollipop updated causes problems."


Particle Debris is a generally a mix of John Martellaro's observations and opinions about a standout event or article of the week (preamble on page one) followed by a discussion of articles that didn't make the TMO headlines, the technical news debris. The column is published most every Friday except for holidays.