Category Archives: Quick Thoughts

Quick thoughts on Netflix Everywhere

Note: you can see all my previous posts on Netflix here. The analysis here draws on financial and operating data I collect on Netflix, along with around a dozen other big tech companies. Subscribers to the Jackdaw Research Quarterly Decks Service get quarterly charts based on this data, and data sets are also available to purchase on a one-off or subscription basis. Please contact me if you would like more information about any of this.

Netflix just announced that it’s expanding to around 130 new countries including many of the largest countries it wasn’t in yet. This was a huge and unexpected move, at least so early in 2016, since Netflix had previously indicated that it would make this move more gradually during the year, with just a handful of markets pre-announced for early 2016. I want to focus on the possible financial impact of this expansion, because it seems to me that it could be significant.

First off, Netflix’s International business is significantly less profitable than its US business:Netflix contribution marginsI’ve written in the past (somewhat jokingly) that every time its international business threatens to turn a profit, Netflix expands into a few more countries. As you can see, though, its International segment continues to be unprofitable at this point, and has much lower margins than its increasingly profitable US streaming business. Why is this? There are several reasons, but they’re all applicable to this new expansion and the likely financial impact.

Free trials

Firstly, Netflix offers one-month free trials to customers. Naturally, the impact of these free trials is heaviest in new markets, and you can see this when you look at the percentage of Netflix customers in various markets who are included in its membership count but not in its “paid members” count:
Free trial subscribersIn the past, as Netflix has expanded into new markets, this percentage has been in the 30% range, though it’s recently dropped down to around 10% or just below. However, with 130 new countries going live simultaneously today, it’s likely that this number will skyrocket. With 25 million members in its existing markets, it’s easy to imagine that Netflix might garner comparable numbers of free trial subscribers in the coming months in 130 countries. As such, a huge percentage of its subscribers overseas will be incurring costs but not generating revenue.

Marketing costs and scale

Another major reason why new markets are less profitable is that Netflix has to do far more to promote itself in these markets where customers aren’t yet familiar with its brand. This chart shows marketing spend as a percentage of revenues for the US and International streaming businesses:Marketing costsAs you can see, marketing spend is a much higher percentage of revenue in the overseas business than domestically, where Netflix has actually been reducing its marketing spend other than in its big new content launch quarters.

As Netflix scales in a given market, this impact is reduced, as the base of revenues from existing customers allows the company to spread that high marketing cost over a larger base, and as awareness and therefore word of mouth marketing grows. However, with 130 new countries, Netflix will have to spend heavily on marketing in the coming months to promote its services. Another huge scale effect is the shared costs of doing business in a new country – converting content to new languages, hosting the content locally, and so on all adds up, and in these countries those costs will be shared by a very small number of subscribers in the short term.

Outgrowing the DVD business

One interesting aspect of Netflix’s business which I’ve covered in the past is the fact that the US DVD business continues to throw off very nice profits, which in turn has largely funded Netflix’s overseas expansion:International vs DVD contributionsThe problem is that this new expansion will be so significant that it seems very unlikely Netflix will be able to offset the losses with its cash cow anymore. As such, overall margins will likely suffer significantly.

The financial impact could be considerable in the short term

For all these reasons, I believe the financial impact of Netflix Everywhere could be very significant in the short term. I’ve been pretty bullish on Netflix overall, and I’ve felt that its slow and steady international expansion coupled with its gradual improvement in US streaming margins was a fantastic combination. This big-bang approach threatens to derail that strategy, albeit only temporarily, bringing what might have been a longer-term dampener on margins forward into a much more compressed space of time, but opening up a far bigger opportunity longer term. I’m very curious to see how Netflix talks about the financial impact here – its press release on the news was conspicuously devoid of any talk of the financial side. But the market certainly seems to like the news so far.

Quick Thoughts: Google’s OnHub router

It looks like a slew of reviews of Google’s OnHub router have come out in the last 24 hours or so, so I’m guessing some sort of embargo has lifted. It seems the reviews are decidedly mixed (which feels par for the course today for products both good and bad, as reviewers each seek to find something unique to say, often overreaching in search of something worth praising or criticizing). Glenn Fleishman has an interesting review of reviews of sorts at TidBits, which is  worth reading for its own sake and for the links to other reviews.

What I wanted to write up here quickly isn’t a review (I don’t have a review unit) but some other related thoughts I’ve had about the OnHub router since it was announced, and which have been reinforced by reading some of these reviews. We discussed some of this on the Beyond Devices Podcast a few weeks back, the week the router was announced, but I’ll probably go a bit deeper on some of it.  I’ll embed the podcast episode at the end of this post too.

A router for novices at a power user price

One of the things that struck me off the bat, but has really been brought home by these reviews, is that there’s a fundamental mismatch between the price positioning of this router and the target audience. The whole value proposition (at least for today – and we’ll come back to this) of this router is that it’s enormously simple to use. And yet it’s priced at $200, the same sort of range as the high-end routers in the market. The Wall Street Journal review notes some interesting trends in the router market, which I think serve as useful context for this device:

Cable companies and other Internet service providers now rent their customers basic Wi-Fi routers when they sign up for service. As a result, U.S. router sales have fallen from 6.1 million year-to-date in 2012 to 3.5 million in the same period this year, market research firm NPD reports… NPD determined that the average selling price is on the upswing. Cheap routers aren’t selling so well, but higher-end models… are. “People are willing to pay a lot of money—more than before—for an AC router with significantly better performance than they had in the past,” said Mr. Baker.

Here’s the thing: I’m betting that the kind of people who are willing to pay a lot of money for “an AC router with significantly better performance” are not the kind of people who feel intimidated by those routers. They’re the kind of people who know enough about routers to understand what distinguishes 802.11ac from some of the older technologies, and who are likely pretty comfortable customizing the various settings. And yet Google’s router is priced in this same range and yet removes many of the typical settings available to power users in other routers in that category. You can’t specify separate SSIDs for the 2.4GHz and 5GHz bands, you can’t get into a web interface, and many of the other deeper settings you can configure on almost any other router. This is a router for novices sold at a power user price point. As such, it’s likely to please neither group.

This is further reinforced by the fact that the router is clearly something of a trojan horse for Google in the home automation space, given the inclusion of Bluetooth and Zigbee, and yet again those users are likely to want far more control over their routers than the OnHub provides. Some of this may be solvable in software down the line, but I suspect Google’s whole mentality around this router is wrong, and that can’t be fixed by software updates.

A symbol of Google’s disjointed approach

To my mind, the OnHub router is also a symbol of Google’s disjointed approach to so many of its projects, and I worry that the Alphabet reorg will only make things worse. Google already has a home automation business, Nest, which not only makes its own products but has been the vehicle for both making further home automation acquisitions (Dropcam) and for acting as a hub for other home automation gear (the Works with Nest strategy). And yet, this product isn’t branded Nest, nor does it apparently sit under Tony Fadell’s hardware group, which also includes Google Glass. In fact, Mark Bergen of Recode and Amir Efrati of the Information have both suggested that this product actually came out of the Google Fiber team. I’ve written previously about how disconnected from the rest of Google the Fiber project has seemed, and it’s ironic to now see Google proper appropriate this technology just as Fiber is being hived off into a separate Alphabet company. The good thing about Google is that people throughout the organization feel free to experiment with various things, some of which eventually become products. The bad thing is that this means you could have several separate teams working on similar things in isolation, and in some cases you end up with several products apparently chasing the same use case (e.g. the Nexus Q, Chromecast, and Google TV/Android TV).

The naming of the OnHub router, the subdomain, the naming of the companion app (Google On), and so on all suggest that this is the beginning of a broader strategy (and we already know that there will be another router made in partnership with Asus). But this is yet another effort within Google to tie together the different devices in the home. Why isn’t it owned by Nest? How will it relate to Android TV and Chromecast, Google’s other living-room projects? So many questions, and so few answers…

The theory and the reality

One other thing Google has touted as part of its positioning for the OnHub is this idea that it is pretty enough to sit in the living room. This, too, feels like a very Googley statement – I’m not sure how many people with real design sense would actually want even a relatively good-looking router in their living rooms. But it’s also a bit of a non-starter as a practical matter – the router doesn’t stand alone – it needs both a power cable and an ethernet cable to function, and you’ll seldom find both of those in the middle of a living room. You’ll at least have power outlets in the wall, but your cable modem is likely to be in a closet somewhere rather than in your living room. And with only one jack in the back of the device, you’re going to need a switch somewhere else anyway for the rest of your hardwired devices, another example of the mismatch between functionality and pricing.

I could go on with all this, but you get the idea. Though an interesting product, the OnHub feels like it falls short on the theory alone, let alone the reality (where several of the reviews suggested it falls down too, even on the most touted features). But it also feels like it’s emblematic of several of the key challenges Google has – too many experiments and projects that are poorly coordinated, poorly thought-through, and ultimately poorly executed. I’m not convinced that the Alphabet structure will help with any of this, and in fact it’s quite likely that it will make the fragmentation problem worse rather than better.

Quick thoughts: Microsoft’s ad business

Given today’s news about Microsoft selling its display ad business to AOL and in turn replacing Google as the search advertising provider for AOL, I thought I’d quickly revisit some of my earlier analysis on Microsoft’s ad business.

By way of background, Microsoft has never directly reported the financials for its advertising business, but it has provided enough detail in its past financial reporting that I’ve been able to build a pretty good picture of this business over time. This past quarter, perhaps as a precursor to today’s announcement, Microsoft stopped providing any sort of information about its display ad business, but here’s a quick view of my estimates of Microsoft’s two major ad revenue streams over the past couple of years:

Screenshot 2015-06-29 16.02.44As you can see, Search advertising has been growing very well indeed, almost reaching the $1 billion per quarter mark last quarter, and likely to hit it very shortly, especially with the help of the AOL deal. By contrast, though, Display advertising has been heading south for some time now, and was under a quarter of a billion in revenue for each of the last two quarters of 2014. The split between the two, then, is roughly as shown in the chart below:Screenshot 2015-06-29 16.04.58In other words, search advertising was not only vastly outperforming display advertising in growth terms, but as a percentage of Microsoft’s overall online advertising business. As such, it’s made sense for some time for Microsoft to jettison this part of the business in favor of focusing on the part that’s working: search advertising. Part of the reason for the disparity between the two is general industry dynamics – display has been struggling for other companies too, while search continues to be one of the most effective forms of advertising and to command commensurate rates.  Microsoft’s display ad business, though, was also sub-scale, and hadn’t made the transition to mobile devices and native advertising effectively. Search, meanwhile, has benefited both from positive industry trends and the growth of Bing and Yahoo’s growth in search market share in the last couple of years.

The impact on Google

AOL’s decision to switch from Google to Microsoft is not enormously impactful on Google by itself, but in the context of Firefox’s earlier switch to Yahoo as its default search engine in the US, and the potential for a much more significant switch away from Google by Apple sometime this year, it’s part of a drumbeat of bad news for Google. One of Google’s challenges at this point is that it’s come to compete with many of its erstwhile partners, with Apple as perhaps the most striking example, and it’s arguably starting to pay the price for that strategy.

Yahoo’s tiny, slow-growing mobile business

We’ll have to wait for the transcript to be sure, but Marissa Mayer must have described Yahoo as a mobile-first business a dozen times or so on the earnings call this afternoon. However, as Yahoo has also been quantifying its mobile business lately, we have some hard numbers to evaluate this claim by, and they’re not all that good at backing up her repeated claim.

First, here’s mobile as a share of revenues for Yahoo, as compared with Facebook and Twitter. Yahoo’s number is expressed as a percentage of what it calls “traffic-driven revenue”, while Facebook and Twitter’s are as a percentage of ad revenue, which likely means much the same thing:

Yahoo mobile ad percentagesYahoo’s percentage is clearly growing somewhat, but not very fast, and it’s clearly light years behind Facebook, which began life as very much a desktop business, or Twitter, which has arguably always had a mobile-centric approach, especially to advertising. Those are mobile-first businesses: it’s hard to argue on this basis that Yahoo is.

Another way to look at things is revenue:

Yahoo mobile ad revenuesHere, fledgling Twitter, which had just $20 million in revenue in Q1 2011, has already passed Yahoo and seems to be lengthening its lead. Facebook, meanwhile, is in a whole different league, and also growing much more rapidly. Again, Yahoo is clearly no mobile powerhouse just yet, and at these growth rates it’s just going to get left further and further behind.

If Yahoo is taking a mobile-first approach to product development, that’s clearly a good thing, and it certainly seems to be the case. But a company that has over 80% of its business tied to the legacy desktop world has a long way to go before it can call itself a mobile-first company.

Quick Thoughts: Periscope and Meerkat

I’ve been meaning to write something about Meerkat for a while, but haven’t got around to it. Now that competing product Periscope has launched, I feel like it’s finally time. A brief primer for those to whom these names mean nothing: Meerkat launched several weeks ago, and is a barebones personal video broadcasting app for iOS, and Periscope is a very similar but more polished app which Twitter acquired a while back and launched yesterday.

Watch me talk about this stuff

I did a quick interview with Reuters TV yesterday on the topic of these two apps, and you can see that here.

Also yesterday, by way of testing Periscope, I did a quick Periscope session where I talked about the two, which I subsequently uploaded to YouTube:

Meerkat’s faster start helped rather than hurt Periscope

Given that people knew about Periscope even before it launched, there has always been discussion about how the two would compete and whether Meerkat would have such a lead by the time it launched that Periscope would be dead in the water. I was always skeptical about that – being first doesn’t always (not even mostly) lead to winning in these battles. and first-day results from Periscope seemed to bear that out. Dan Frommer of Quartz posted this chart this morning:

This shows the number of Meerkat and Periscope-related tweets over the last few weeks. As you can see, it took Meerkat most of that time to reach the 30k per day mark, but Periscope almost hit that number in its first day. What I’ve wondered about, and what now seems to be happening, is that Meerkat actually did Twitter and Periscope a huge favor by teaching people about the concept and letting them experiment with it, such that by the time Periscope came along people immediately understood the value proposition.

I prefer Periscope so far

On top of that, as I alluded to in my first paragraph and in the video above, I find Periscope to be much more polished and easier to use than Meerkat. Meerkat seems very much like a cheap Android app, somewhat ugly in its interface and hard to navigate around. Periscope seems much more better thought out, nicely designed and professional, much more at home on iOS (the only platform both apps are on, for now). And of course then there’s the fact that Meerkat got cut off from Twitter’s social graph shortly after launch, which will make it harder for Meerkat to sign up new users, because the process of finding followers will be much harder. For power Twitter users, custom curation of following and follower lists has always been important, but for mainstream users easy setup is key, and Meerkat has likely lost some of that.

Periscope isn’t perfect – the fact that I had to post the video of my session from my camera roll to YouTube rather than being able to share it directly from the app is something I hope they’ll fix in time. Both apps force you to use a portrait orientation, which is fine for selfie-style videos but poorly suited to either group shots or shooting a scene (and, as this BBC citizen journalist expert mentioned, poor for incorporating into professional news broadcasts too). But I like the fact that it keeps sessions in the app for viewing after they’ve ended, unlike Meerkat, although finding those needs to get easier.

It’s still early days here, but I think Periscope has a fantastic chance to dominate, given its polish, its ownership, and the fact that Meerkat laid the groundwork for it. Meerkat isn’t necessarily doomed, but there’s little in my mind to recommend it over Periscope, and I think the latter will pretty quickly become dominant.

Why did these products break through now?

The other question worth thinking about is why these two apps suddenly appeared at this time, and what’s made them so successful when other previous efforts have failed. I think it comes down to two main things, and a third factor which has helped too:

  • Mobile-first products – these products don’t really exist in any meaningful sense on the web or desktop – they’re mobile-first, and only enable streaming from iPhones. Meerkat allows sessions to be viewed from the web, while Periscope doesn’t, but they’re clearly designed for both broadcasting and viewing on the device you have with you all the time. This also allows them to tap into notifications, which are huge for real-time services.
  • Twitter as megaphone – speaking of real-time, Twitter has become the real-time platform, with many people keeping it open and watching the tweets roll in more or less as they happen. It’s already been the real-time platform for covering current events, but that has largely meant text and the occasional picture. Video adds a whole new dimension to this real-time element, and Twitter becomes the megaphone through which these broadcasts from apps get shared with the world. Tapping into existing Twitter audiences and leveraging the retweet as a way to seamlessly rebroadcast popular streams is a huge part of why these services are successful. It’s also the biggest reason Twitter’s acquisition of Periscope makes sense – it adds another vital dimension to Twitter’s “media” strategy.
  • Tech reporters getting on board – this was arguably less a catalyst than an accelerant, but quite a few tech reporters quickly jumped on board with Meerkat, tried it out, broadcast themselves and got the product through its awkward “here’s me eating my lunch” phase that all sharing platforms seem to have to go through. Though that phase was repeated as Periscope launched as regular people tried it out, that early reporter experimentation meant we quickly learned what worked and what didn’t, and this early experimentation by those with significant followings helped the product to get far more attention than it would have got through more of a grassroots adoption by regular people.

Breaking into the mainstream is the next challenge

The challenge will be transitioning from this early success among those who have significant followings on Twitter to regular people. Are ordinary people whose Twitter audiences consist mostly of people they know in real life going to use either of these two services? Or will their broadcasts only be meaningful when they happen to find themselves in the right place at the right time, as with yesterday’s building collapse and fire in Manhattan? I suspect that over time we’ll see some new Meerkat and Periscope celebrities emerge, just as we’ve seen Instagram and Vine give rise to new personalities with significant followings seemingly out of nowhere. But I think that, most of the users will be viewing, not streaming, and that actually fits great with Twitter’s new direction.

Quick thoughts: Defragmenting media on Facebook

Facebook seems to be working on two fronts to bring content from third party sites natively into the Facebook experience. This began with video, where Facebook has been quietly bringing both major traditional brands and smaller content creators into the core Facebook experience. But there are now reports that Facebook plans to do the same with news articles from major publications like the New York Times, Buzzfeed, and National Geographic.

I’ve talked previously about the video efforts, and in that piece I said this:

Facebook has become a massive destination for video, but almost all the video is actually hosted on other platforms. That obviously has cost advantages for Facebook, but it means that it doesn’t own the content, and therefore can’t monetize it effectively. It also means that engagement around videos on Facebook is fragmented, with popular YouTube videos attracting millions of comments scattered across hundreds of thousands of different user shares of the same video.

There are lots of aspects to all this, and Facebook is no doubt talking up the performance and monetization benefits of publications hosting their content directly on Facebook. But I think one of the other key benefits is the ability to overcome this fragmentation, and that applies just as much to news as it does to video. So let me expand on that a bit. The problem as things work today is that there’s no single version of either a video or an article on Facebook, just the original on the third party site. Meanwhile, there could be thousands of individual shares of that video or article on Facebook, with no connection between them. On the third party site there might be an indication of the number of third party shares, but the site has no easy way to digest what’s happening on each of those shares, and Facebook users have no visibility into how many shares, comments or other metrics the article or video is capturing in total across Facebook. Sometimes Facebook takes an article shared my multiple friends and bundles it into a single card with a single link, but there are still two entirely separate discussions happening among two separate groups of friends.

What Facebook could do if these things were being shared natively through Facebook is start to aggregate all this activity much more effectively, both for the content owners and for users, with commenting, stats tracking and so on happening much more effectively. And of course Facebook could share much more data about the users sharing the content with the content owners, so that they’d get a much better picture of who’s viewing the content. One of the key challenges with Facebook (in contrast to Twitter) is that sharing is inherently private, which provides almost zero visibility for content owners. With native sharing on Facebook, that could change, though of course it raises some interesting privacy implications. If you’re commenting on my video or article on Facebook, does that mean I now get to see your comments, even if they’re only shared with your friends and not public?  And in a world where many news sites have either switched off comments or left them on but failed to curate them effectively, could Facebook help to provide a better class of discussion?

There are so many aspects to all this, including some significant risks for the brands involved. But it seems to me that they would at least in part be making the following tradeoff: ceding control over hosting and branding the content itself in favor of better visibility and tracking of the engagement with that content. For some of them, at least, I’m guessing that’s a tradeoff worth making.

Quick Thoughts: Apple Watch and notifications

My thanks to Gregg Keizer of Computerworld, for sparking these ideas in a conversation we had earlier today ahead of Monday’s event.

I wrote a piece for Techpinions earlier in the week about the concept of intimacy as it relates to Apple Watch. In that piece, I wrote:

What I think intimacy in computing really means is going a level deeper on the personal side, and perhaps also stripping away some of the non-personal elements of the smartphone. The Apple Watch, then, becomes the truly personal device our smartphones have never quite been. Notifications come in noiselessly, communication with our Apple Watch-wearing significant other can be both more private and more individualized, the tasks we do on the Apple Watch can be limited just to those that are meaningful to us, leaving others for the smartphone, and so on.

I highlighted one phrase in bold there, because I think it’s critical both to the success of the Apple Watch specifically and to smartwatches in general. One of the criticisms of some of the other watches out there is that notifications are something of an all-or-nothing phenomenon: either you get utterly bombarded with notifications or you get none, but it’s been tough to get granular control. With the Apple Watch, this problem has the potential to get worse: with both traditional notifications and apps running on the Watch itself, there’s potential for even more of this. So it becomes all the more important that Apple get this right on the Watch. The ability to manage notifications and other activity on the Watch so that you only get notified for the things you really care about is a key value proposition. There have been a couple of pieces today about how the Apple Watch will handle notifications, but neither of them is clear enough to answer conclusively the question of how well the Watch deals with this issue.

I think there’s an analogy here to the early smartphones, which acted as task-specific extensions of our PCs. PCs were general purpose devices, on which we did essentially all our work. Early smartphones, and especially those made by BlackBerry, majored on email above all else, and allowed us to perform triage on this central form of communication on a dedicated device, leaving the rest for when we got back to our desks. I see the Watch performing a similar role in relation to smartphones, which have now become in some ways general purpose devices in a similar way to early PCs. If the Watch is to succeed, it will do so by allowing us to focus on those tasks that are most important to us, stripping away the rest of the stuff we’ll leave for our smartphones. And for many of us it will do this for our personal lives in much the way early smartphones did it for our work lives.

Quick thoughts: Another way to think about Nest

A couple of good blog posts from other analysts this morning triggered a thought in my mind about Nest.

The first of the two blog posts is from my fellow Techpinions contributor Ben Bajarin. I recommend reading the whole thing, but here’s the key thought:

Apple’s customers are higher value customers and their growing installed base means they are amassing one of the largest, if not the largest, installed base of premium customers on the planet. This observation has some striking implications…

He goes on to talk about two of the implications, and I think the insight there in  particular is great. One of them is the impact this has on the competition, among which of course we find Google, which owns the mobile operating system that’s mopping up the vast majority of the non-Apple customers.

Something else I read this morning (a post from Benedict Evans that’s really about something completely different) prompted me to think about this in the context of Google’s Nest acquisition. I’ve been thinking about Nest primarily as Google’s strategic play in the smart home space, and as the hub and vehicle for the rest of what Google will do in the home. And I think that may well be in large part what that Nest acquisition is about.

But thinking about Nest in the context of Ben Bajarin’s piece made me see Nest in a different way. Think about Nest for a minute: its characteristics as a product, the people it’s likely to attract, even the people who work there. This is a very Apple-like product, made by a former Apple executive, and I’ve always said that Nest was a much better fit at Apple than at Google. It’s even sold in Apple stores.

What if Nest is at least in part about capturing those very lucrative and attractive Apple customers without having to convert them to Android? What if what Google is building with Nest is at least in part a concession that Android and phones based on Android aren’t likely to attract these customers, but by playing in a completely different space – the smart home – Google can in fact attract those customers? And once it has them there, perhaps convert them to other aspects of the Google ecosystem, which after all is where Google really makes its money? There’s no particular reason why Google needs to have all its customers on Android anymore – it’s served its function of preventing Microsoft and Apple from dominating the mobile world. And of course, there’s been lots of talk about even Google’s services making more money on iPhones than on Android. What if Google could establish a different beachhead in devices that’s not dependent on first converting people to Android? What’s fascinating about Nest is that it’s about the only recent Google initiative that’s not about reinforcing Android as a platform – Android Wear, Android Auto and Android TV are all about extending Android specifically into new domains rather than simply spreading Google services into new domains.

I’ve no idea if this was actually part of the thinking behind the acquisition of Nest – most likely, like other acquisitions, it wasn’t about a single strategic objective but rather several. But it would certainly be an interesting response to the emerging reality that Apple has captured the vast majority of the most valuable customers within its ecosystem, and trying to win them back through phones seems to be a losing strategy.

The big downside to this, of course, is that Google has very deliberately kept Nest somewhat at arm’s length from the rest of the company (a point Benedict raised in his post). But this strategy only works if Google continues to build links between the two, as it’s already begun to do with Google Now integration. With Tony Fadell now overseeing Glass as well, there’s obviously even more linkage between Google proper and the Nest team, and it’s another sign that Fadell might be asked to oversee more Google devices and pursue the same strategy.

Quick thoughts on Microsoft and Amazon in the cloud

At the end of last year, I wrote a piece for Techpinions Insiders about what to expect for Amazon in 2015, as part of a series on major tech companies. As part of that piece, I wrote the following about Amazon’s AWS business:

AWS has been one of Amazon’s big success stories over the last several years, generating higher margins than e-commerce and growing extremely fast. But growth faltered a little in 2014, as competition, from Google and Microsoft in particular, intensified. The basic storage and infrastructure services are becoming rapidly commoditized, with plummeting prices and little real differentiation. As such, both margins and differentiation will move to what these companies build on the basic services; hence Amazon’s launch of Zocalo and other enterprise tools that sit on top of AWS. But here it is going up against Microsoft’s traditional stronghold and Google’s increasingly capable offerings. End user software hasn’t been Amazon’s strong suit, but it’s Microsoft and Google’s bread and butter. I remain skeptical of Amazon’s ability to successfully compete in this area. Meanwhile, others not in the cloud storage business are also building their own competing platforms at this higher layer, including Box, Salesforce and others. If AWS is to become the highly profitable core Amazon has always lacked, it needs to successfully compete at this layer as well as the basic services it has provided historically. It’s not clear to me Amazon will be any more successful at this in 2015 than it was in 2014.

This week, Microsoft reported its earnings for the December quarter, and among other things Satya Nadella said this about Microsoft’s own cloud business in response to a question about cloud margins:

Overall, the shift to the higher layer services is the real driver here, which is obviously Office 365 and its various levels is one factor. The other one is what I talked about in the Enterprise Mobility Suite, that’s really got fantastic momentum in the marketplace because the solution has really come together and is fairly unique, as well as Dynamic CRM.

So these are all got a different profile in terms of margin and they are all now pretty high growth businesses for us. So when you think about our cloud, you got to think about the low-level infrastructure. Even there we now have premium offerings and then we have higher level services. So that aggregate portfolio is what helps us move up the margin curve.

I juxtapose these two quotes – one about Amazon and one about Microsoft – because I think today’s announcement by Amazon of new enterprise email services plays directly into the challenge I described, and which Microsoft seems to be managing much better. It continues to be critical for both companies (and Google) to migrate their way up the cloud stack to the higher-layer services (as both I and Nadella called them), but Microsoft is already there, while Amazon continues to try to compete in a space I’m really not sure they can. We’ll see what AWS results look like tomorrow, but I expect this to be something that comes up on the call.

Quick thoughts: on Apple’s subtle machine learning improvements

John Gruber had a post on Tuesday about the improvements to Siri over time. He was mostly talking about the core functions of Siri itself, but I’ve noticed some other subtler improvements over the past couple of years that I’ve been meaning to write up for ages. Technically, these aren’t part of Siri per se, but they sit in the same category as Cortana or Google Now’s equivalent functionality, so I see them as almost an extension of Siri. What’s striking about these to me is that as far as I can tell Apple hasn’t explicitly talked about them and since they don’t trigger explicit notifications many users might not even notice them.

I grabbed a screenshot of an example of this back in February 2014, and I think it’s the best way to explain the sort of thing I’m talking about:

2014-02-08 07.08.45It’s the first paragraph there that I’m focusing on, and I need to provide some context to enable you to understand what happened here. Early last year I played for a few weeks in a church basketball tournament. I had games regularly each Saturday at a similar time, and at the same location. In my calendar I put simply “Basketball” and never entered the address, since I knew already how to get there. What Apple’s machine learning engine did here was (as far as I can guess 1):

  • Note that I had an item called “Basketball” in my calendar for that morning
  • Make a connection with past appointments on Saturday mornings also called “Basketball”
  • Look up past location behavior in its location database to connect a particular location with past instances of “Basketball” in my calendar
  • Look up this address and calculate driving time between my current location and this destination
  • Present it to me at a relevant time in the Today screen.

Again, Apple has talked up some functionality around using calendar locations explicitly entered in your calendar to provide these sorts of alerts, but I’m not sure it’s ever talked about the deeper machine learning stuff in evidence here. I’ve never seen exactly this sort of extrapolation from past behavior again since this occasion, but I have received other notifications on this screen that it’s time to leave for appointments where I’ve explicitly entered a location in my calendar, based on heavy traffic (it happened to me this past week at CES, for example).

I wonder how long it will be before Apple starts surfacing this information more proactively with notifications rather than merely displaying it reactively when users pop open this screen. And I wonder if Apple will begin talking about this functionality more in future – I note that Microsoft is currently running adds which state that Siri can’t tell you if you need to leave for a meeting based on traffic. That’s technically true – Siri is entirely reactive today – but iOS does tell you if you know where to look. That seems like something Apple could easily fix, but it would turn its machine learning capabilities from a background feature that’s so subtle many users likely miss it into a headline feature. And with that might come concerns about  the data Apple is gathering and how it’s using it. It may have decided that, for now at least, it would rather keep quiet about all this, but I wonder how long that will last.


  1. Another possibility is that Apple merely built a pattern of my past behavior on Saturday mornings without explicitly connecting them to the calendar item. I don’t see that it makes a big difference either way.