On the eve of New Year 2010, I published a set of predictions about what the technology industry would look like a decade later, in 2020.
Since we are now halfway through that decade, I thought it would be a good time to check in to learn some lessons from how they are shaping up. I had some big misses. I thought that data and platforms would be much more open by 2020 than I now believe they will be. I missed some big trends, like the importance of mobile apps. In other areas, like startups and content, my predictions appear to be holding.
Overall, I think I underestimated the power of incumbent financial and distribution structures and missed how fundamentally different the mobile ecosystem would be from the desktop world.
Here’s a breakdown by area, with the relevant parts of the 2010 predictions and my evaluation. You can read my full 2010 predictions in much more detail here.
STARTUPS
New Year 2010: I asserted that “The Internet will still be very interesting, and a small handful of long-term meaningful Internet companies will be born—but the rate of turnover of power will change significantly. Web applications will be mom-and-pop businesses built on scalable clouds and highly abstracted languages. Biotech/Nano/Robotics/Health will actually be meaningful in the real world, rather than just promise.”
New Year 2015: It is interesting that in 2010 I didn’t mention mobile apps at all, which have clearly been the dominant story of the last several years. It is possible that by 2020 the story will have ended and we will be back to a world where we don’t talk about them anymore; however, I think that is unlikely. Eighteen months after Apple launched the App Store, I just don’t think I had internalized how incredibly powerful the combination of distribution and monetization would be beyond the games and other trivial apps that dominated the charts at the time. It wasn’t clear to me that the smartphones would lead the world in a different direction from the desktop Web.
Otherwise, I still believe that by 2020 the Internet will be dominated by a handful of long-term players and that the rate of turnover in power will have declined (though in the last five years, the shift to mobile has strained this prediction).
ADVERTISING
New Year 2010: I wrote that “Traditional advertising [will start] to decrease because it [will have] fully shifted online, and it turns out it doesn’t really work at scale very well anymore.” I also predicted most brands would have moved their marketing from cost-per-click models to cost-per-acquisition and beyond to customized real-time generated offers.
New Year 2015: Advertising is shifting online, and despite the fact it has been a long time coming, I still believe that by 2020 the role of agencies will have dramatically diminished. It will have become clear that the return on investment in advertising exists in the targeting and customization provided at scale by search and social but not in things like custom campaigns and native advertising.
I was overly optimistic about the speed at which real-time offers would become relevant online. I’ve learned how slowly most of the world moves on measurement and how hard it is to leverage scaled data for targeting. I don’t think the majority of big brands will have shifted to real-time customized offers by 2020 because I think it will be simply too difficult for them to do it well. If big brands are doing highly customized on-the-fly offers by 2020 it will be because they are letting companies like Google and Facebook do it for them.
HARDWARE
New Year 2010: I said the “trend will be towards interfaces generally becoming commoditized and getting really cheap and open. I don’t think we will care very much about the specific boxes that host our connectivity in 10 years, just as now we don’t care about a lot of the components that were once central. Apple is the only company that has a even shot of maintaining a premium position in the human interface market.”
I said that the laptop will still be a primary interface in 2020 and that digital cameras will become irrelevant and integrated with phones. I also made the now seemingly crazy assertion that only 25 percent of my input usage would be on a touchscreen.
New Year 2015: I’m amused I was so bullish on my laptop. While I am sure I will still use a laptop daily in 2020, clearly the phone has become far more important than I gave it credit for, as is further illustrated by how much I still believed in physical keyboards.
As for the theme of openness, I have a hard time evaluating it in 2015.
I still believe that by 2020 most of our hardware will be cheap and open; however, that has not been the story of the last five years; there has been clear consolidation in the handset market around leaders like Apple and Samsung. When it comes to the associated platforms, iOS has maintained far more differentiation and market share than I expected, and Android has gone from nowhere to dominance and Google continues to exert major control over it.
Still, I believe that at least as far as devices, and perhaps operating systems, the next five years will see a huge move towards the market fracturing —though perhaps not true openness and commodification. This is a dangerous bet. I could easily be dead wrong here; however, when I look at trends in Asia, with companies like Xiaomi and OnePlus, and platforms like Tizen and Cyanogen, I think we could easily see a world where there is far more diversity, competition, and even commoditization by 2020. It will have not been a straight line, but I could see my openness bet directionally playing out by 2020 even though it looks totally wrong now.
INTERNET SERVICES
New Year 2010: I wrote “Google will still lead search, but search for static answers will be a relative commodity, and AdWords will not be as high margin a business as it currently is (though it will be far larger in terms of absolute revenue and dollar profitability). Facebook will dominate identity by dominating the socially validated “graph” and feed will provide leveraged scarce information–the social “graph” will not be open, but other communications forms and transactions will provide different topologies of people, places, and things. Overall there will be massive consolidation of time spent on the Internet, and most services will go hyper vertical or completely commodity.”
New Year 2015: Google is still clearly going to be the dominant search engine in 2020, but search is proving a much tougher business for the company on mobile, and margins will compress. It feels clear that Facebook is going to be in a very good position. The company does have a hugely valuable proprietary data set. A few key platforms have already massively consolidated time spent online.
That said, when I wrote that applications besides Facebook would have different social “topologies,” namely social graphs, I didn’t anticipate that Facebook itself would become a federation of different apps, like WhatsApp and Instagram, with different social data. That will clearly be the case by 2020. (I joined Facebook in 2010 after making my predictions and left this past fall.)
MEDIA
New Year 2010: I wrote, “content is everywhere and always accessible…. people stop caring about owning it locally entirely, and everything is subscription and/or pay as you go. The conduits of content are devalued. The best content creators do relatively well–though the average content producer does much much worse.”
On television, I predicted that Hulu and Apple would give way to Google and on music I predicted my favorite services of the time—Lala, TheSixtyOne—would give way to “ubiquitous distribution and consumption.” And I wrote that creative talent would hold the most leverage over other media stakeholders.
New Year 2015: The general theme that media is widely available and goes subscription and pay-as-you-go feels correct. But I was way off on the services that are going to dominate. I think I was just looking at the wrong racehorses at the time.
On TV I completely missed the importance of Netflix. At the time, they seemed to me like a physical media rental service my parents subscribed to, struggling to make the difficult move to streaming. At this point, people seem to barely even remember that they once worked with the the U.S. Postal Service. It might be that YouTube turns around and by 2020 validates my bet on Google, but it isn’t looking likely that they will be the undisputed TV leader in five short years.
On music, I clearly missed Spotify, and dramatically underestimated Pandora, both of which I was aware of and using at the time. Thinking back to the end of 2009, I was primarily listening to music via a series of small streaming services, and I think that led me to believe that music would be everywhere and what service one used wouldn’t matter. I underestimated rights holders’ power and distributors’ willingness to endure the pain of working with them.
I also overestimated the power that creative talent will have over other media stakeholders in 2020. Services like Twitter, Facebook, and Instagram are connecting creators more directly to their audiences. But those distribution platforms have become so powerful that they, to an extent, feel like the heirs to the power left by the publishers and rights holders rather than the artists themselves. So, I think the real media story of this decade will be about how Internet distribution platforms subsumed traditional publishers.
COMMUNICATION
New Year 2010: I wrote: “Email will technically still be dominant, but it will have evolved so far (albeit on a relatively kluged core system) that the fact that it is technically email and called email will not really be relevant to the end user experience—Gmail squared is NOT Wave. Voice phone use will be down on a per user/per month basis… IM and SMS will be massively up, but will be so deeply integrated that we won’t think of them as separate services, rather than just irrelevant communication routing options.”
New Year 2015: I still believe that email will be my primary means of messaging in 2020 (which I know will make me old, but it will be true nevertheless). But I clearly missed the rise of mobile messaging apps and push notifications, which have been the real story of the last five years and likely will be the major trend for the next five. I didn’t see messaging as part of the web-services layer.
While there have been stabs at one, there has yet to be a successful meta-messenger service to commoditize WhatsApp, Facebook Messenger, Snapchat, Line, WeChat and others. And frankly, I think there are very strong competitive pressures that will keep a meta-messaging service from being successful. But, at the same time, I do believe that the jury is still out on the stickiness of any of the messenger service as an individual platform. The fact that messaging follows a strong power law, where the vast majority of one’s usage is with a handful of users, should mean that the networks are far more replicable than social networking.
REGULATION
New Year 2010: I predicted that by 2020 “Amazon and Facebook will be targeted from antitrust perspective.”
New Year 2015: I don’t think Amazon or Facebook are going to face antitrust issues in 2020 because competition is getting more intense. If anything I think Amazon will feel incredible pressure from Google and possibly Uber in the U.S., as well as, perhaps, entrants like Alibaba.
Facebook I think, will be extremely powerful through a federation of applications; however, the social space is expanding and changing so quickly that it is hard to see them drawing serious antitrust scrutiny. The rise of companies like Snapchat has punctuated this. Further, the Facebook “graph” itself doesn’t create the sort of long-term structural advantages I once thought it would. Mobile address books have been a key, and for me, unanticipated, enabling platform that allow anyone to wire up their own graphs relatively easily. What is harder to replicate is the technological footprint and ubiquitous install base that Facebook has achieved with its apps.
That said, my predictions missed local regulation. I suspect the real story of regulation by 2020 isn’t going to be about blockbuster national antitrust. Instead, it will be about how cities and states deal with technology companies whose products are threatening their industries.
CONCLUSION
I don’t think that 2015 is going to be a very eventful year in the technology world. I don’t expect that products like the Apple Watch will be relevant on a larger timeframe, and while there is a lot of very exciting work going on in everything from virtual reality to drones to self-driving cars, it is all early. Perhaps these things will have a meaningful impact by 2020, but it’s not clear.
Looking a bit further ahead to 2020, however, I suspect that in another five-year window there will be another major shift or two that I am currently massively underestimating just as I did the impact of mobile during the last period.