Company / division: Alphabet
Corporate venture capital in Silicon Valley is always fascinating: these are the things big companies choose to invest in indirectly rather than directly, which often leads to fairly complex relationships between some of these companies (notably Google’s relationship with Uber). My Beyond Devices Podcast co-host Aaron Miller did a great breakdown of this feature of the tech industry a while back – have a listen to this episode if you’re interested. It appears Google and Intel were the biggest investors in 2016, with some really sizable bets made by Google in particular during the year. I wish I could see more details from the report, but CB Insights pulls a bait and switch in which it first asks for just an email address and then follows up with a demand for a phone number and job function, so I’m not downloading it.
It’s interesting to see Google working with the MRC around auditing now too – Facebook just announced MRC auditing a couple of weeks ago, but it had of course had an embarrassing series of screwups relating to metrics for advertisers and content providers, whereas YouTube didn’t. However, this is reflective of a broader mistrust of online advertising by big brands and marketers, and an inconsistency in the use of major metrics like viewability. From what I’ve read, the MRC standards are pretty minimal as far as what counts as a view, but at least there’s consistency there, which is a start.
As usual, it would be great to understand in more detail the methodology behind this survey, but it’s not available. The Verge seems to have got the rankings wrong – from what I can tell, Samsung was 7th and not 3rd last year – but it’s also worth noting that Samsung’s score dropped from 80.44 to 75.17, which sounds a lot less dramatic than dropping from 3rd (or even 7th) to 49th. The fact is that there are a lot of companies clustered together between 75 and 87 points and so a small drop in the score produces a big drop in rankings. Since the survey was also conducted in November and December last year, when the Note7 debacle was still very fresh in people’s minds, I’m guessing it would score a lot better just a few months from now. Though the Verge picked up on Samsung’s drop as their headline, it’s worth noting where other tech companies sit too: Amazon is #1 (score 86.27), Apple #5 (82.07), Google #8 (82.00), Tesla #9 (81.70), Netflix #18 (79.86), and Microsoft #20 (79.29), all of which classify as either very good or excellent. It’s also worth noting that big cable companies like Comcast and Charter score in the low 60s, which qualifies as “poor”, while the major wireless carriers score 66-72 (“fair” to “good”), with T-Mobile top and Sprint bottom.
This is a really important aspect of autonomous driving that’s not talked about nearly enough. In the SAE levels system for describing autonomy in vehicles, all the layers between 0 and 5 require the driver and vehicle to work together at least to some extent, which means that even when the car has taken over a task, the driver is supposed to remain ready to take over when the car requests him or her to re-engage. The problem here is that we tend to switch off, whether deliberately or merely passively, when our focus isn’t actively required, and that means that machines have to give us an awful lot of notice when we need to take over. In practical terms, that’s often impossible, and that can actually make cars operating at levels 3-4 in particular less safe rather than safer than human drivers. That has important implications for those manufacturers which seem to be trying to work incrementally up from Level 2 to Level 4 or 5 over time, like Tesla, because there seems to be an increasing consensus that we may need to skip those middle levels entirely. And it also means, as I’ve pointed out a couple of times before, that lots of experience operating test or production vehicles at Level 2 or 3 is not nearly the same as being ready to produce a Level 4 or 5 vehicle.
via Bloomberg (we discussed this topic in depth during this episode of the Beyond Devices Podcast and this talk by Gill Pratt, head Toyota’s Research Institute, is also very illuminating on the same topic)
I think the framing here is exactly right – this is part of the broader crackdown at Alphabet on some of its longer-term and less financially viable projects. The new approach – targeting balloons at specific regions rather than trying to blanket the globe – always seemed like the more obvious way to go, but of course balloons are inherently hard to navigate, so I’m intrigued to know how they will manage that. Two big questions remain: firstly, whether Internet access delivered from the sky can ever be really good (see existing satellite-based Internet access, which tends to be slow and bandwidth limited), and whether Alphabet should be in the access business at all (see also yesterday’s Google Fiber item). At least it sounds like this particular project might generate revenues sooner rather than later (and eventually even profits!) but it’s still not clear that it’s going to benefit the core Google business much.
“Just like Alexa” is a bit of a stretch here, because the whole point of Alexa’s ordering is that you know the products will come from Amazon. Google Home, by contrast, will order from a range of different Google Express merchants, only some of which are available nationwide. And because most people don’t have a Google Express account set up yet, they’ll have to do that first before they order anything. Lastly, unlike items bought using a Prime subscription, shipping will be charged extra after a short promotional period. Despite all that, this is obviously an area where Alexa has had unique capabilities and where Google Home has now closed the gap a little. By far Home’s biggest disadvantage is still its lack of awareness and distribution.
Google has officially released its Android-apps-on-Chrome beta and it’s getting fairly terrible reviews. While Google is legendary for attaching the beta label to things that feel like they’ve long outgrown it (see especially Gmail), in this case the label actually feels premature, since the experience on these new Chromebooks sounds much more like an alpha than a beta. The software is very buggy, and the implementation of Android apps on ChromeOS is if anything even worse than it long has been on Android tablets. Getting Android apps onto a laptop sounds really compelling – all the apps you’re used to on your phone, now on your PC! – but the reality for now sounds extremely disappointing. It’s hard to avoid the sense that this is a misstep by Google, which promised these launches a few months back but clearly isn’t ready to deliver on them yet. In fact, these devices appear to suffer from some of the same lack of readiness as the recent Android Wear 2.0 release, although that was supposedly a finished product. This is uncharacteristic for Google, and it’s a worrying sign that things are being rushed out of the door to meet deadlines rather than because they’re ready.
Google Fiber and Alphabet’s broader Access unit within which it sits has been somewhat in limbo since late last year, when it lost its leader and canceled all its expansion plans. The story the company told then – and still seems to be telling today – is that it intends to pursue the same goals in new ways, principally through wireless. The fact that a new head has actually been appointed means that it’s at least somewhat serious about that goal, and isn’t just going to sell off or shut down the whole business, but it’s still possible that it might sell its fiber assets even if it pursues wireless technologies in future. Meanwhile, I still don’t think there’s a good reason for Google to be in the access business at all at this point.
99% of Mobile Malware Targets Androids Because of Open Store and Infrequent OS Updates – F-Secure (Feb 15, 2017)
This data comes from the blog of F-Secure, a European cyber-security company which tracks malware. The key finding here shouldn’t be a surprise – Android sees 99% of malware activity on mobile, for three simple reasons: it has by far the largest share, its app stores are open and often weakly policed, and Android devices are often very slow to get OS updates and software patches, although it has been doing better on that last point recently. Interesting, there’s still far more malware being created for Windows PCs than Android, even though there are fewer of them, but the range of malware being created for Android is approaching that which targets PCs, even though the main focus is still trojans. All of this, of course, only serves to reinforce the narrative about Android being insecure.
If only the device you use as a voice assistant had phone functionality built in! I’m being facetious, but it’s interesting to watch Amazon and Google potentially working backwards from a non-phone device to something capable of making calls. This is a logical extension of a voice search for a local business – I already regularly do this using Siri, especially while driving, and it’s very useful. As with yesterday’s Nest story, this is a great illustration of the benefits of software-based products – you can provide meaningful additional functionality through an update and suddenly the device you already have becomes more functional. I would guess that Amazon would need a partnership for local business search, whereas Google of course has that functionality in house – it’s in domains like this that Google has an advantage over Amazon despite the latter’s early big lead. I’m very curious how far out these efforts are – unusually, the WSJ is reporting on both companies’ efforts at once here, but they may well be at quite different stages of development. And of course Google famously stayed out of the phone business when it launched Google Fiber because of all the regulatory headaches and fees that go along with being a fully-fledged phone provider – it might try to stop short of going that far this time around too.
Nest adds automatic door detection and animated push notifications for subscribers – VentureBeat (Feb 14, 2017)
This is fairly minor news from Nest, but that seems to be the only kind of news it’s capable of making these days. Other than a new outdoor camera in the middle of last year, it’s mostly just refreshed existing hardware over the last couple of years, and there hasn’t been a completely new hardware category for several years. However, these software and machine learning-based enhancements do show the value of a smart device – hardware already in market just got more functional thanks to a software update. It’s not clear from the coverage here whether Nest is leveraging any Google expertise or whether it’s building the necessary technology in house, but one hopes it’s the former.
YouTube Orders First Original Kids’ Programming for Red Subscription Service – Variety (Feb 13, 2017)
YouTube has some original content for adults already, almost all of it tied to YouTube creators in one way or another, but it’s now extending that investment into kids’ programming to go with its YouTube Kids app. Again, several of the shows feature YouTube personalities, so it’s leveraging its access to content as well as giving creators yet another reason to stick with it rather than switching focus to, say, Facebook. The YouTube Kids app has been a bit of a mixed bag so far – at a time when several big traditional kids’ programmers eschew advertising, it’s shown ads (unless the viewer has a YouTube Red subscription), for example. But this is an interesting next step.
I’ve changed the headline here to make it a bit more specific, but there’s actually quite a lot more to this speech, and although the article is a little hyperbolic, I do think this is important. Procter & Gamble is the world’s biggest advertiser, so its views and policies with regard to digital advertising are worth paying attention to. Its chief brand officer just gave a speech in which he railed against programmatic advertising and the broader opaque digital advertising supply chain, the power of Facebook and Google, inconsistent standards for measuring ad viewability, and more. Some of the very same things big ad-centric companies are constantly touting as key to their businesses are the same things that are causing consternation among major advertisers, and that’s a tension that isn’t going away anytime soon. Facebook is making strides on its metrics screwups from late last year, but programmatic – which Google talks up every quarter – is getting terrible press at the moment in relation to ads showing up on unappealing sites, and it feels like there are changes coming here. Worth reading the whole article just to see some of the big frustrations advertisers are working through and the possible impacts.
via Marketing Week
One Reason Staffers Quit Google’s Car Project? The Company Paid Them So Much – Bloomberg (Feb 13, 2017)
That’s an interesting hook for the article, but far more interesting than the ease of hiring away Waymo employees is the fact that Alphabet had such a kooky compensation scheme in place at all – more evidence that the Porat era has introduced much-needed financial discipline in the Other Bets. Long story short: Alphabet paid massive bonuses to employees in its autonomous driving division based on “project milestones” which had nothing to do with financial performance (since the division won’t generate revenue for years). This, in turn, has apparently loosened up some employees who have enough financial security to take risks on leaving for competitors. It really is remarkable how the Google/Other Bets split has shone a spotlight on some of the crazy largesse in the latter businesses.
Billboard does an annual Power 100 ranking of the most important/influential execs in the music industry. Coming at this from a tech angle, there are several notable companies on the list: Spotify’s Daniel Ek takes the top spot, several Apple folks are at #4, Amazon at #12, iHeartMedia at #19, YouTube at #30, Pandora at #34, Facebook is at #54, and various others are scattered through the second 50. Amazon’s ranking is surprisingly high, but is entirely due to Billboard’s perception of Echo and Alexa’s role in transforming music, as illustrated by Billboard’s interview with Jeff Bezos and Amazon Music head Steve Boom. I think the take here is a little overblown, but there’s no doubt Echo and Alexa are changing the experience of music for the small minority of people who use them. YouTube’s relatively low ranking is surprising given how important a channel the site is for the music industry, but of course its relationship with the labels and artists is complicated. This kind of ranking exercise is always somewhat arbitrary, but it’s interesting to get a music industry take on the tech companies and their relative importance here.
This is a big hire for Lyft, which so often plays second fiddle to Uber in so many ways. Being able to recruit a top notch mapping engineer like this away from Google is a great validation of Lyft as a company and as a recruiter specifically, and should make it easier to hire in other talent for mapping and autonomous vehicle technology at Lyft. It’s also notable that Vincent would be willing to leave Google, which obviously has far bigger and deeper efforts underway around mapping and autonomous driving than Lyft does.
It’s always been odd that Google has two separate music streaming apps rather than simply two on-ramps to a single streaming app, so it’s good to see it combining the teams behind them as at least a first step towards eventually having a single music experience too. Neither of Google’s apps ever shows up on industry charts of subscribers, so the numbers on both are likely small still, so this is a great time to make a change before foisting a lot of upheaval on a large base of customers. Note also the bizarre final paragraph in this piece, which somehow tries to tie Google’s move here into a narrative about apps in general falling out of favor with users.
via The Verge
Android Wear and LG Watch Reviews Are Mixed at Best (Feb 8, 2017)
It looks like Google and LG lifted an embargo this morning on Android 2.0 and LG’s two new smartwatches. My first reaction to the reviews here is that the new watches sound pretty terrible, and that we’re back to grading these smartwatches on a curve, something I first noted back in 2014 before the Apple Watch was announced. The Verge review is illustrative – it notes that the Sport version is uncomfortable and enormous (it doesn’t fit under shirt cuffs), doesn’t have interchangeable bands, the Android Pay app takes too long to load, and can’t be used while swimming; the Style version lacks most of the more interesting features on the Sport, looks cheap, and the batteries on both versions struggle to make it through the day, while Android Wear 2.0 is pretty buggy. The Verge’s rating? 7.1 for both. Their rating for the Apple Watch Series 2? 7.5. Android Wear has struggled to take off ever since it launched – it’s just never felt like Google or its OEMs understand that watches are fashion accessories, and need to be designed for that job. Packing a billion features into these watches isn’t going to cut it, especially if they don’t work well, or they end up looking ridiculous on your wrist. I’ve seen nothing here that makes me think Android Wear 2.0 is going to do any better than the previous versions.
The article doesn’t mention Microsoft once, but talks about Google’s consumer products several times, which makes it feel like this is rather missing the point. This is an enterprise offering and therefore goes head on against various Microsoft products and services intended to achieve similar aims (as well as those from Box and other smaller, more specialized enterprise software and service providers). Both Google and Microsoft are focusing on their AI skills as a source of differentiation in enterprise file management, by promising to help employees find the files they need. Search is, of course, a core Google skill, but it operates very differently in an enterprise file system from on the open web, and Microsoft may actually be better placed here given its long history and the massive investment many companies have made in Microsoft tools in this setting.
Although most of the attention Alphabet gets around cars is around its Waymo division, which is focused on autonomous driving, Google’s Waze group is also doing interesting things around another of the three major shifts happening in transportation: ride sharing and urban mobility. In this case, it’s not so much an Uber- or Lyft-like driver-provided service as a simple carpooling arrangement with a little compensation for the driver, but in that sense it has the potential to be more cost effective, using existing infrastructure to provide lower cost transportation and reducing congestion in the process. Google is building a ride sharing capability much more subtly and quietly than it is autonomous cars, but the former has far grater potential to make a big difference in disrupting traditional transportation in the near term.