s3e17: Funny, I’ve always believed that the world is what we make of it 

by danhon

0.0 Station Ident

Tuesday 26th April, 2016 at the XOXO Outpost after having taken an hour out of the morning to go to depressingly middle-class-white-stereotypical Music Together (Hello, everybody!) class with wife and son where today’s dance party was Prince’s 1999. I’ve also managed – somehow – to continue the trend of naming newsletter episodes after quotes from Contact[0].

[0] Contact (Amazon USAmazon UK)

1.0 Google’s Innovation Labs

So Google’s Bradley Horowitz (née Yahoo Brickhouse[0, 1]) is setting up an in-house startup incubator called Area 120[2, 3] alongside Don Harrison that’s partly claimed by to be an effort to stop talent from jumping ship. I don’t know anything about the talent jumping ship issue (other than what’s been reported in the press like a few high-profile staff moves like the former head of Google’s ATA, Regina Dugan, moving over to Facebook[4]), but boy do I have wildly under-informed opinions about startups and “innovation labs” that you might be interested in!

These opinions are spurred in part by having seen a lot of innovation labs happen around the place – from my personal experience in ad agencies, and seeing friends and colleagues be located under the (protective?) aegis of innovation labs at companies like Yahoo! and Facebook, at media companies like the BBC, Channel 4 and Penguin, in government (local, metro, state and, uh, country-level) and, uh, other places I guess. One pattern that I want to highlight is clarity around what a startup incubator or innovation lab is intended to do. What’s its purpose?

If you can’t come up with a good reason as to why a startup incubator or innovation lab exists and what it’s supposed to do then I suppose my gut instinct is to respond with a sort of hollow laugh because ultimately it might feel like you’re going to be wasting your time. I’m assuming, of course, that you want to actually effect some sort of change. Some [citation needed] of the time, these labs and incubators are just an executive hedge: something about our business is under (perceived or real) threat, and we must be seen to be doing something about it to appease a third party. Whether or not the executive leadership will then actually *recognize* a real threat and then *do something about it* based on whatever comes out of the (directed or undirected) innovation lab or incubator is a completely different issue. Which is why it feels like when it comes down to organizational strategy, an innovation lab or incubator (I’m just going to call them innovation labs now) feel pretty useless unless your organization itself has a pretty good sense of self-esteem and purpose. In other words: if it’s true that Google’s lab is called Area 120 in an homage to the 20% projects that resulted in Gmail and AdSense/AdWords, then that’s perfectly fine but let’s be clear: those projects were green-field areas for Google. They didn’t cannibalize any existing business. They were areas where there wasn’t any organizational drag or any political fallout (well, not much) if they worked. They solved clearly and cleanly defined user need problems without stepping on any divisional toes. I always keep coming to my advertising example because a) it’s one of the closest that I have to hand and b) it’s also one that’s still (admittedly) a little personal. But when I left Wieden Portland, one of the issues was that if the agency *really* wanted to figure out what its potential role in “digital” could be, it had to let digital compete with its bread and butter. It had to genuinely wonder and follow-through on what a digital alternative to solving a problem for a client might be instead of producing a :90 Superbowl manifesto (which are good at solving a certain class of problem!) and lets those new solutions compete on their own terms. But, it’d be hard, and it’d be, in a way, potential cannibalization.

I draw an analogy here with where Ben Thompson has recently been talking about Apple’s organizational crossroads[5]. The argument that Thompson makes is that Apple is a “unitary organization” – it’s organized around expertise, not products. There’s a marketing function, a product function, an operations function and so-on. There’s no “Mac” division fighting against the “iPad” division fighting against the “iPhone” division in the way that Microsoft used to (and maybe still does?) have warring factions between Office and Windows and Server and so-on. Apple has one Software Engineering function that does everything software for everything, one Hardware Engineering function that works across all hardware usw.

The problem with saying that you’ve got “innovation” or “incubation” over in a separate part of the business is that, well, they’re in a separate part of the business. That’s where innovation happens. That’s where the new stuff happens that those people over there do, which, at some level, is a potential if not direct threat to the livelihood of other people in the organization. How should I act, as VP for Gmail, if Area 120 comes out with something Slackish that promises to Kill All Email? Do I support it? Does the Slackish thing – say, wave – become part of my purview? Do I give a shit about it, or do I protect my empire? Does it mean my budget will decrease if they’re successful? As Thompson says, not having titles like “VP of Mac” means that it’s entirely possible that someone else can come along and do something else that “does computery things for people who use computers to do things” better than the Mac and there’s less chance of organizational inertia or resistance to the new thing. Because there’s a higher purpose.

That’s my big worry or reticence about the ability of innovation labs to do the job-that-needs-to-be-done in organizations that say they want innovation. A) Do they actually know and understand and are prepared to accept the problem that they’ve tasked themselves to solve and b) are they prepared to accept the consequences?

[0] Bradley Horowitz – Wikipedia, the free encyclopedia
[1] Yahoo To Close Brickhouse By End Of Year | TechCrunch
[2] Google Preps New Corporate Incubator — The Information (value-wall)
[3] Google reportedly starting an in-house startup incubator | The Verge
[4] Facebook swipes the head of Google’s ATAP lab to lead a new hardware division | The Verge
[5] Apple’s Organizational Crossroads – Stratechery by Ben Thompson

2.0 Nokia and Withings

So, Nokia, a company that used to make boots and tires and then was famous for making phones that could only really do things like play snake, make phone calls and send texts and maybe take some but not that many photographs, but is now famous for completely missing the point on smartphones bought Withings[0, 1, 2], a company that is a) French and b) makes, well, French-ish internet of things smart object type things ranging from internet-connected scales to internet connected blood pressure monitors, internet connected baby camera monitors and smart watches that are smart but don’t have a conventional computery display.

Some things that were of interest to me:

– the purchase price was only $191m which holy crap seems really low for a company that – for me at least – had a profile like Withings and did make really nice stuff that more or less worked (I guess I’m damning with faint praise) and has 200 employees
– what is it with European internet companies?! Are they just incapable of making things at scale these days? I mean the market cap for Fitbit is 3.74 billion! Do European companies just not know how to, uh, market? I mean, it’s not like the Withings stuff is shit! It actually works, unlike stuff from Jawbone[citation needed].
– a long time ago, I had a surprise interview with Nokia Sunnyvale (I know! They had an office in the valley!) that turned out to be for an interaction design director position and I had lots of Opinions About Smart Devices and how most of them weren’t smart they were really fucking dumb and they should hire me to have a point of view about this kind of thing. One thing that I gushed about at the time was the Withings Activité[4], which long-time readers will recognise from me having gone on about for ever and ever and ever because it was a) a watch, b) was smart, c) didn’t look like a dorky computer on a wrist. It just looked like a watch and it was smart and it talked to other things and it was just comfortable being a watch or a piece of jewellery and offloading other user interface tasks to things that were, well, super user interfacey. At the very least I think it had an interesting and distinctive point of view and a certain je ne sais quois that provided a valid differentiator to the abundance of GRAPHS AND TRACKING AND NUMBERS AND HOLY SHIT HAVE YOU SEEN WHAT WE’VE QUANTIFIED NOW coming out of Silicon Valley quant. self. companies.

But now we have Nokia buying Withings and my thoughts go to:

– well I guess Nokia might still remember how to have pretty awesome industrial design and supply chain management, so they’ve got that going for them
– the Nokia brand still means something to people
– but I think we can all agree that we never really thought Nokia was that awesome at software and services, and the Withings stuff is what I’d mark as “competent on the verge of passable, but certainly not worthy of merit in any way whatsoever” or even just plain “workmanlike” notwithstanding the really stupid fucking issue of sending a “hey, we can help you lose weight!” marketing email to a user account where THEY KNOW THAT THE USER IS LESS THAN THREE YEARS OLD.

I didn’t end up taking the job with Nokia for a variety of reasons, but I’m super interested in, for lack of a better term, a European sensibility to quantification and introspection. What would that look like?

[0] Nokia plans to acquire Withings to accelerate entry into Digital Health | Nokia
[1] Withings joins Nokia
[2] Nokia is buying digital health firm Withings for $191 million | The Verge
[3] Fitbit Inc: NYSE:FIT quotes & news – Google Finance
[4] Withings Activité

3.0 Grab Bag

* I love this Verge/Circuitboard writeup of the Starbucks emoji language[0]
* GERTY is coming soon to a hospital near you (if you live in Japan, most likely)[1]
* the best one is that Google *says* that they’re deploying a new recommendation “based on deep learning technology” for a new video recommendation system for YouTube, which I instantly (if uncharitably) rephrase as a submission to arxiv titled “Novel Deep Neural Network Technique For Promoting Consumer Purchase Activity Based On Web Browsing Activity”. Part of the issue with all of this deep learning stuff is that it’s going to be ever more trivial to set up a giant Deep Learning stack in the cloud, feed it a bunch of data and then come up with a (not really supported) conclusion that anyone else could’ve given you in about five minutes glance because hey, that’s just something you should try, but now THE ALGORITHM says that you’re allowed to do it and if THE ALGORITHM told you to jump off a cliff would you and I guess that these days we have definitive evidence that the answer is yes, some people would, in which case  ¯\_(ツ)_/¯. I mean AT THIS POINT I really just want to see what sort of data set I can stick in front of an off-the-shelf LSTM RNN to see if it can predict with a high degree of probability exactly how minutes I am away from my next Diet Coke.

[0] You language 4 brand | The Verge
[1] Panasonic’s pink Dalek delivery bot has been cleared for hospital work | The Verge
[2] Official YouTube Blog: A new mobile design for your Home

IN CONCLUSION: innovation labs – probably not worth it (says pessimistic me) unless you have strong signals and priors that executive leadership will actually do something with the innovation outside of pat you on the head and say “good innovation, now we need to protect our existing businesses), and Nokia x Withings is super interesting if only because who knows what those crazy Finns and French will come up with next remember after all where would be without wifi-connected rabbits[0].

[0] Karotz, Nabaztag & Cie…

Send notes, my best regards and all of that jazz,

Best,

Dan