I read all the analysis. It wasn’t particularly satisfying. Business Insider offered vague non-answers, reporting that, according to Didi president Jean Liu, “the two will be working closely together,” and going on to say it was very unclear exactly what that meant.
Brian Solomon at Forbes more or less gave up on finding a strategic tie-in, concluding, “the most likely reason is that such bets may no longer be out of character for Apple,” and that “[t]he main goal for Apple has to be a sizable return on its $1 billion investment.”
Buried in the list of reasons he dismissed, however, Solomon offered this line: “Apple and Didi may eventually find some operational synergies to exploit, but hardware and ride-sharing are quite different businesses.”
It niggled at me. Hardware and ride-sharing are quite different businesses. I slept on it. I chewed on it. I stewed on it. And I started to wonder: Are they, though?
Some time ago, there was this guy Steve Jobs who was obsessed with an integrated, end-to-end service product. Unlike other hardware manufacturers, his hardware only worked with his software. Unlike other software manufacturers, whose software had to work across multiple devices, his was monogamous, and therefore able to precisely maximize the features of the underlying hardware.
This integration meant that Apple controlled every detail of the end-user experience -- and it was a thing of beauty. Your computer worked perfectly with your phone, worked perfectly with your iPod, worked perfectly with iTunes, where you bought all the music — and you could just live happily in Appleville forever and ever.
Steve Jobs didn’t look at Napster and think, “Hardware and music-sharing sure are quite different businesses.” He thought, “Hardware and music-sharing should be the same business. It is all one continuous experience of music.”
So what’s the iPod/iTunes version of ride-sharing? It’s an Apple car: autonomous, obviously. Running Apple software. Summoned by your iPhone. Linked to your Apple Pay account.
And what does Apple need to get there?
Apple is long rumored to be building a self-driving car, and I have no doubt the company will create something beautiful. It might not even look like a car: a single-occupant pod is just as useful in an on-demand scenario.
Running Apple software, no problem. Summoned by your iPhone, no problem. Linked to your Apple Pay account, no problem.
So where’s the problem? “Autonomous, obviously.”
Building a high-quality autonomous car takes two things: engineering prowess (which Apple has lots of) and data.
But Apple is notoriously secretive with new products -- nobody can see them until they’re ready for retail. And it’s virtually impossible to accumulate the vast amount of data the company needs in secret.
How much data are we talking about? Well, you might remember when Google’s self-driving cars reached a million miles driven -- that was a big deal. Two weeks ago, the MIT Technology Review reported that Tesla had surpassed 100 million miles of vehicles steering themselves. 780 million miles of total data captured. Another million miles of data every 10 hours.
A week ago, Didi’s Liu said the company books 14 million rides a day. Reckon it would be useful for Apple to spend some of its more than $200 billion in cash to retrofit every one of those vehicles with data-capture sensors?
With its investment in Didi, Apple might only be looking for a return on some of its piles of cash. It might only be trying to give Uber a hard time, or gain more market intelligence in China. Or it may see autonomous, on-demand ride-sharing as the present-day equivalent of the Mac vs PC wars. If that’s the case, a billion dollars worth of data is pretty good ammunition.