This is in response to 16 mobile theses by Benedict Evans of Andreesen Horowitz, a firm that’s central to Bay Area VC culture. The theses are about half wrong, which isn’t too bad.

I’ll run through his theses one-by-one. But first, I think our differences center on two things; one that’s predictable given who I am, namely the cloud. The second is perhaps surprising: Whether keyboards matter.

Here we go; you might want to flip back and forth a bit because I reproduce Mr Evans’ subtitles but not his arguments.

“1: Mobile is the new central ecosystem of tech” · Tech is bicentric, these days: Cloud and Client. The cloud doesn’t care that much what the clients are, and the clients don’t care that much what’s inside the cloud. The technologies, implementers, and financial structures have become unsurprisingly disjoint over the years. Which I don’t think is a problem, and the fact that it’s even possible is a tribute to the power of Internet protocols.

So, half right.

“2: Mobile is the internet” · That’s just dumb. I work for a business with an annual run rate of ~$8B, growth of ~75%, and a healthy profit margin. If someone figured out how to take us and our competitors down, it’d blow a pretty big hole in the side of the Internet.

Also, large parts of the cloud spend their time talking to other parts of the cloud with little client involvement.

This kind of thinking is not only wrong but dangerous to investors’ money. Maybe not to a16z money, and Mr Evans is obviously talking his book, but still.

“3: Mobile isn’t about small screens and PCs aren’t about keyboards - mobile means an ecosystem and that ecosystem will swallow ‘PCs’” · Mobile is absolutely about what fits in a pocket, and increasingly I think that PCs are about keyboards.

I nearly always have instant access to a mobile, and most times to a PC too, but only if I’m willing to sit down, pop it open, and type a password. Since I’m impatient and short of time, the incentive to do everything on the mobile is overwhelming. And yet often I don’t, and the biggest reason is the keyboard. It’s just too much work and irritation to squeeze coherent rhetoric or pictures or conversations through the brain-dead fake on-screen keyboard.

Now, this is changing; the most recent tablets are starting to have keyboards that are pretty good. I personally have never yet run across one that would let me really work, although I haven’t tried the very latest from Apple and Microsoft.

But then, once it’s got a sizeable screen and a keyboard, it probably doesn’t go in a pocket any more, and… it’s a PC! Screw the OS, that’s a distraction. Thus, my conclusion is exactly the opposite: I think a “mobile” is more or less defined as “fits in a pocket, no keyboard” and “PC” as “has a keyboard you can use for work”.

So, wrong.

“4: The future of productivity” · The argument here is that big screens and spreadsheets and keyboards aren’t what matter for “real work” (quotes in the original). Instead, “What matters is the connective tissue of a company - the verbs that move things along. Those can be done in new ways.”

Well, I suppose that large parts of a VC’s job can be accomplished without recourse to typing or making presentations, and indeed we do find new ways of getting shit done. But for the moment, wrong, because spreadsheets matter and documents matter and keyboards matter.

“5: Microsoft's capitulation” · Fair enough, “Windows Everywhere” is history, and that’s a capitulation.

But… used a Surface recently? And while the success of cloudstuff like Azure and Office365 isn’t guaranteed, anyone who isn’t taking those efforts seriously is, once again, putting investors’ money at risk. Still, not calling this one wrong.

“6: Apple & Google both won, but it’s complicated” · Great stuff; really, if you’re not following along on Mr Evans’ side of the argument, flip over there now if only to read this little gem.

“7: Search and discovery” · The question is: What comes after search, and who gets the traffic?

Since it’s not conceivable that a serious Bay Area VC is ignorant of Twitter and Instagram and Facebook and so on, this little item totally baffles me. Dunno about you, but I discover the Internet through the people I follow, and I don’t think I’m weird. I suspect that there’s an interesting point here that Mr Evans just didn’t quite manage to connect with; but still, wrong.

“8: Apps and the web” · This starts out strong but then concludes that what matters is which icons are on your home screen. I’ll call that mostly-right.

But here are two things: First, a large and increasing proportion of my app interactions start with me pulling down my notifications, not hitting a home-screen icon. Maybe this is an Android-vs-Apple thing?

Second, the home screens on all my mobiles have Bookmark widgets for obvious online things and I hit them as often as any app.

“9: Post Netscape, post PageRank, looking for the next run-time” · This  —  it begins “For 15 years the internet was a monolith: web browser + mouse + keyboard” — is incoherent.

Sure enough, the platform contention on the client side, between Apple/Google/Microsoft/browser-tech, is fierce. And in the cloud, there is ferocious competition between vendors, to define the “X” in XaaS, and then over what X should be built with.

But still, he’s right that at the end of the day, it’s all about gathering users.

“10: Messaging as a platform, and a way to get customers” · Yep; good thinking here. Plus, I can tell you as a Cloud insider that messaging occupies a whole lot of our server-side thinking too; but you probably already knew that.

“11: The unclear future of Android and the OEM world” · I’m down with every word here. Android has served its purpose by keeping any player from getting a lock on the channels that advertising wants to flow through, and thus on Google’s oxygen; and then again by routing a bajillion people to Gmail and Maps and so on. But what next?

“12: Internet of Things” · I have to be careful talking about IoT, not so much because it might be career-limiting, but because smart people I respect seem, unlike me, to think there’s a there there.

But yeah, to quote Mr Evans: “talk of standards for IoT misses the point - ‘connected to a network’ is no more a category’ than ‘contains a motor’”.

“13: Cars” · Yep. Me, I worry a lot about whether the shiny New Economy has replacements for all those vehicle-driving jobs.

“14: TV and the living room” · Bah; the TV doesn’t care whether it’s being driven by a mobile device or a PC, and the distinction is whether or not I need a keyboard to pull up and manipulate what I want on the screen. Now obviously, if there’s anything you can do on either your mobile or your PC that can’t use the nearest TV as a display, that’s a bug.

So he’s half-right here.

“15: Watches” · Whatever. I just don’t see lives being changed or substantial businesses being disrupted, so this being here is wrong.

“16: Finally, we are not our users” · Inspiring, heart-warming, but wrong.

Most people are perfectly happy to use mobiles as designed, for chatting and YouTube and Candy Crush, and PCs as designed, for spreadsheets and planning and reporting. The last thing most people want is to “take ownership of the tech in their lives”; they just want it to work.

And the score is · By my count, a total of eight points I think wrong. If so, Mr Evans is still doing OK; I’ll take a .500 batting average on my tech prognostications home with a smile any day.

And also, he’s asking good questions, which I’ve long thought more important than offering the right answers. So, thank you Mr Evans. Seriously, go read the thing.


Comment feed for ongoing:Comments feed

From: David (Dec 29 2015, at 05:32)

I think your point on keyboards (point #4) is a bit shortsighted. I repeatedly see my college and early career aged kids prefer their phones and keyboard-less tablets to do things like photo editing, spreadsheets, and long form writing, even when a laptop is within reach. While you may be correct for people your (and my) age, I think each successive generation pushes keyboards further into irrelevancy.


From: Bruce (Dec 29 2015, at 10:40)

Good article by Evans and good commentary here.

As for IoT, there is a core of goodness there that is struggling to get out. I think it is going to happen, but in ways that are hard to predict right now. The hardware enablers are there, and some genuinely useful things are being done. But the ecosystem and business models are very unclear.

By the way, you confused me by referring to Mr Evans as Mr Bradley at one point.


From: Bill (Dec 29 2015, at 19:18)

Evans is blinded by the client-side and underestimates the rest of the infrastructure. The numbers on clients, and growth to the rest of the world, are naturally astounding. But they're visibly astounding. For those of us that see the cloud side, it's a little easier to see how BIG things get. But even then, it can be hard to comprehend :-) I think it's incomprehensible for those client-side biased. (Which opens up the possibility of interesting investments that those folks will miss...)

Another part of the syndrome is referring to cell phones as "supercomputers". While we've had impressive growth in performance, on the expected curve, that'll flatten out soon. And the perf levels are kinda funny relative to the things we're currently doing to accelerate ML things etc..


From: Mike (Dec 30 2015, at 11:02)

Regarding '8: Apps and the web' this is most likely an Apple vs. Android thing. I went from Android to Apple just this year when none of the Android manufacterers decided to release a compelling phone under 5 1/2". The notification pane on Apple is a pale immitation of Android, effective only at reminding me how much better it was on Android. It's bad enough that Apple could remove it from iOS and I think it would be an improvement (on the flip side, the quick setting pane on iOS is vastly superior to any I saw on Android).

So yeah, chalk this one up to Android vs. Apple.




From: Matěj Cepl (Jan 04 2016, at 13:29)

I am not sure what it means, but am I the only one who now makes majority of searches directly on Wikipedia without a detour through DuckDuckGo/Google?


From: Paul Guinnessy (Jan 07 2016, at 13:44)

I think the issue on number 7, search and discovery, is too limited in both arguments. If you're looking for the hottest research papers at the moment, you don't go to google, you either go to Nature, Science or arXiv. If you're on arXiv there are so many papers published each day its impossible to cover or read the abstract of them all.

Papers in scientific research are growing at about 6% per year so its a problem that's getting worse or worse (we'll skip over using publications like the one I work for as a way of keeping up to date). And the time spent reading papers keeps dropping.

Hence I would say there are some real opportunities in search and discovery that haven't been applied yet (although science publishers are investing a lot in trying to find the answer to that question) and its still open to all newcomers to find the answer.


author · Dad
colophon · rights
picture of the day
December 28, 2015
· Technology (89 more)

By .

The opinions expressed here
are my own, and no other party
necessarily agrees with them.

A full disclosure of my
professional interests is
on the author page.

I’m on Mastodon!