Review of KPCB’s 2017 Internet & Tech Trends

Learnings from a product person’s perspective of a 350-page deck

I. Setting the Stage

At the pace the tech news goes these days, you’ve probably already forgotten all about Mary Meeker’s latest gem of a trends report. This year was a doozy. Over 350 pages. Analysts, get you some sleep.

The slides are embedded below to make this easier and so you can follow along with our comments of how to interpret these findings. Use this to build your own startup, innovate at the company you work at, or just to learn something new.

Knowledge is a tool that must continue to be sharpened.

II. Advertising & Retailing

  • p5: global internet users and smart phone growth has essentially stopped. We’ve reached the fat middle of the market in terms of spreading and have plugged in half the human species. The only thing that’s spread further is life support (oxygen, water, food) and electricity. New enablement technology diffused? Check.
  • p9: adults now spend nearly a full working day interacting with tech and digital media in the US. So, you’ve got your work job and now you’ve got your fun job. The convergence has already begun. Social media marketers isn’t just some joke job. It’s a skill of the new economy.
  • p13: this “time spent vs ad dollars spent” chart has been out of whack for years. It’s finally getting in sync. Which means its time for a new medium to get released. Only this time, it’s going to be so personalized, it’s basically a reflection of you.
  • p14: the internet is our new television, is our new radio, is our new book.
  • p15: if you’re buying ads, trying to target a certain demo to create a conversion into a sale, you basically only have one choice: Google or Facebook.
  • P16–25: ads is where almost all of the money is flowing in tech. You may find it boring, but don’t forget, boring = cash flow.
  • p27: I’ve never been a fan of interactive or dynamic ads. It makes the user do two things they hate: 1) sit through another damn ad, and 2) you’re not going to make me interact with it?! Oy, the best ad doesn’t look, feel, taste, or sound like an ad. It’s compelling content.
  • p32: Google is basically the digital world’s most pervasive and compelling product. Because it makes people pay to be next to the stuff you’re looking for. In the real world, that would be like some dude following you around a museum asking if you want to go buy this paint brush or eat at Denny’s next to a picture of food. Not knocking it, it’s an incredible business, but there’s a lesson here in that the digital world has different rules than the real one. Expect that lesson to keep playing itself out over the coming decades.
  • p35: future ads are going to be pics instead of text. Humans process pics way faster. Pinterest and Snap might have an opportunity against the Googs here, but the search habit is so embedded in the human populace that now it’s part of the web browser URL bar. Snap nor Pinterest has that level of OS access.
  • p39: retail and brands. It’s not about just showing your product in context anymore. We believe it’s about creating an experience around your product that customers want to come for. An example is throwing a music festival in the summer if you’re a bottled water company. You create the context and leave your product as the only option. That’s how you reach Millennials and Gen Z. And Amazon.com is how you reach adults with kids and no free time. It ain’t rocket science, it’s just execution.
  • p40: Will Smith saved us from aggressive aliens. We appreciate him. And he was a former math guy so, street credibility.
  • p44: AR is going to be interesting. Yet to be seen how pervasive it will become from a visual perspective until we have the glasses that don’t look like cell phones strapped to our heads. Or as I call them, “Noggin Boxes”.
  • p45: AI Assistants, to me, are the promise of truly invisible software. Many have talked about it, but you can’t just say what you want and have it work right all the time, as well people are very uncomfortable talking to a machine in public. It just feels weird. No amount of NLP or AI is going to change human behavior. It’s going to have to jump the shark in favor of neural laces where I think it and it happens.
  • p53: customer support survey where people tell you what they want. I’ve got a better solution. Make your product and service so great that it makes support obsolete. I know, easier said than done, but it’s a goal that means your business is around forever. Look at Vaynerchuck. He didn’t sell wine. He sold “I’ll drive to you in a blizzard with a power generator to keep your business open and house warm”. Then you never buy from another wine seller again. That’s extreme support.
  • P55: that little intercom chat window is freaking everywhere.
  • more retail stuff with a bunch of metrics and pipelines and yaddy yaddy
  • P66: apartment lobbies become storage facilities for packages. Boy can I attest to this in Cali as well as NYC. It’s a really smart notice by the folks who put this together. What it means is again: Amazon is your retail strategy. Forget store fronts, websites, apps, etc. Just drive traffic to Amazon pages. Get Affiliate commissions on top of it. Reduce friction, just get the damn sale. Your brand is in front of people when they get the product home. You don’t need to own the web experience.
  • p68: food delivery is a huge deal. We’re tired, we’re sick of cooking, and entire startup businesses have been created on the back of a grubhub page. It’s like amazon for real-time, hot food. It’s a big deal and will only get bigger. Cheap + Healthy + Filling. That’s your magic formula.
  • p71: AI for clothing design of the future. Don’t count it out.
  • p73: people don’t want to just try stuff on. They want to get it now to wear now. Real-time is the key to in-person experience. That’s the excitement, not to wait for it to be shipped. Not a fan of the “waiting” model.
  • p74: Wal-Mart’s in real trouble.
  • p75: meanwhile, Amazon Basics is a gold mine.

III. Interactive Gaming

  • p81: if you’re on the internet, it’s highly likely that you game. Play is built into human nature because it’s escapism.
  • p82: I used to watch a room full of developers talk about and watch online gaming while they coded. This was back in 2013. Cooperative esports is not a fad, but something that’s going to be around for some time. The new sports agent will be after these huge sponsorship dollars.
  • p84: people in Asia like video games. No secret here.
  • p85: pretty even split amongst all age groups on who games.
  • p88–97: there’s a reason gamification works. The problem is always around defining and designing a concept called “fun”. Just because you slap points on something doesn’t automatically make taking out the trash fun.
  • p99: Twitch ain’t goin nowhere. Level that up with Minecraft and then add in some cryptocurrency where it becomes your job and watch a new economy blossom in a digital universe completely outside the confines of governmental taxation, which means you get to keep 33% more of that money you spent time making.
  • p101: I’ve never played fantasy sports but all my friends have been playing since we were kids. Weird that it took until only recently for the likes of Fan Duel to show up.
  • p103: more game machanics. Understanding that they exist is the first step, the next step is applying them correctly. Easier said than done. Trust me from experience running a gamification product for brands to increase engagement via social media.
  • p114: gaming beats social in engagement. The problem is that you eventually master the game but not your social network. So more value is continually added to Facebook, Insta and Snap while the creators of FarmVille, Clash of Clans, and Angry Birds have to keep creating hits. And content is hard.
  • p120: tech and sports has been talked about for decades but there hasn’t really been a breakthrough killer product that everyone swears by. Maybe the bracelet / watch that tracks steps but that’s pretty low-grade in terms of information density.
  • p123: state of the art for sports platforms is just analytics dashboards with some gamification added on top. But does that really increase your performance? Sure it helps to know you did a few more steps than yesterday and that you’re on a streak for working out 3 days in a row, but you kind of already knew that intuitively didn’t you? Where’s the leverage? Much like the iPhone made computing pervasive, did the Apple Watch really make sports pervasive and higher performance? It seems there’s a big hole that still exists in this area until we start tracking automatically what food goes into our body, the digestion that takes place, how exercise and sitting affects that, alcohol, work and environmental factors, and then gives you suggestions for improving your health.
  • p129: gamification in industrial settings is highly important. It gets talked about too much in consumer and not enough in aerospace, automotive, surgery, military, industrial manufacturing, robotics, etc.
  • p132: the volume of digital information created is on a power law curve, we’re still just at the very beginning.
  • p133: Moore’s Law still isn’t coming to a close, regardless of how many people talk about it. Because there is always room to pay more for faster compute.
  • p138–145: esports is a big deal. It’s growing 40% every year now reaching half the population of the United States.
  • p150: do trends in gaming predict future product around all of tech, much like porn used to in the video production space?

IV. Media

Editor’s note: after years of experience building online video/audio platforms, media businesses, and social apps, we are undoubtedly going to have a biased view about this section so you can either take it as we know some inside baseball or we’re bat shit crazy. So all I can promise is the unvarnished truth from my perspective.

  • p153: music streaming is pretty simple — make sure you have the song I want to listen to, give me some new songs I’d like but haven’t found on my own, and then charge me a flat monthly fee like $10 for it. Execute on those 3 things and you’ve got yourself some good bacon. Of course, Amazon is the real wild card here because as a Prime member, you get music plus a whole host of other things.
  • p154: online video vs cable tv is also very simple — give me the channels I want to watch, new movies the moment they hit the theaters, and charge me a low monthly fee. High quality is the name of the game here. Hollywood blockbusters and some personal favorites (HGTV, ESPN, HBO, Science Channel, etc). That’s it. Everything else is window dressing and noise. Apple Music is getting into Hollywood with original content. Netflix is killing the game but their library aside from originals is waning. Amazon Prime is coming in hot, and watch out for Facebook’s wildcard of a movie in a text message.
  • p158: streaming music subs are growing because nobody needs to buy albums anymore. Truthfully it’s Apple Music (install base of devices) vs Spotify (the O-G-riginal).
  • p161: Netflix subs are also growing rapidly as they expand their service globally. But the only thing that matters is getting the latest Hollywood blockbusters on their service. If they can’t continue to do that, it’s going to be a long painful death. I’m talking Avengers, Star Wars, etc. Basically, Disney. Which isn’t going to happen because they have their own streaming service.
  • p163: recommendation engines. Oy, have I heard the pitch for this a million and one times. Again, we’re back to “Ok this title I’ve never heard of is interesting, but what’s the new Hollywood flick I can watch?”. That has nothing to do with recommendation engines and everything to do with high quality content licensing deals with the studios.
  • p165: it’s YouTube vs Facebook. The former gets longer form, the latter shorter form. They’ll both battle for movie/TV-grade content but that’s Netflix, Amazon, iTunes. So there’s your split. Facebook is quick-hit social, YouTube is UGC around 15 mins, Netflix is originals and TV shows with some movies, iTunes is your Hollywood rentals and purchases, and Amazon is just the everything all the time so maybe one day.
  • p167: interesting that every one of the social networks is growing at the exact same MAU rate.
  • p171: the older you are, the more you watch. Less of a social / eventful life?
  • p173: pay TV will never go away because it’s more economical to pay the $100 to $200 per month for the cable, internet, TV bundle and get all the channels than trying to buy only internet service and then pay for one off channels and then movies on top of it. The reality is a mixed bag approach. Some internet, some cable, some OTT.
  • p176: the goal is a personalized linear channel. You tune into one TV station, let it play all day every day, and it’s filled with everything you want when you want it. Social videos, educational stuff, movies, TV, etc. Disclosure: I’ve got a patent for inventing this and for mobile live streaming.

V. The Cloud

  • p181: might surprise you to see the public cloud really hasn’t taken much of the spend away from a traditional data center / private data center.
  • p182: but AWS is kicking everyone else’s ass, which is no big surprise.
  • p182: security is a big deal, and as cyber becomes ever more important in the global economy (i.e., digital transformation), expect some big exits in this space due to the sheer need to protect not just information, but money.
  • p184: at this point, most people are familiar with APIs where as recently as 5 years ago I did some research and could find only about 10,000 public APIs in existence. Today, containerization is in a similar space. One way to think of it is like a computer dedicated to one API. Also, edge computing could be the biggest boon to digitizing our world, but the problem is fitting something powerful enough in a small space that doesn’t suck in a lot of electricity. You need efficient chips and efficient AI (i.e., connectomics).
  • p187: boy have I spent a lot of time on enterprise software. SAAS-based cloud products. It’s the playbook played out across Silicon Valley and internal digital transformation teams everywhere, now.
  • p188: at this point it seems laughable to say Design Is Important for digital products. Though the requirements for usability is much less, considering the Snapchat generation grew up with this stuff and can reverse engineer a UI just by tapping it.
  • p190: we don’t hear enough about cybersecurity, mostly because it’s a tight-knit group of folks who prefer to remain anonymous and protected. But it’s a big deal, especially in closed door conversations. Interpolate this to our infrastructure systems like power, water, etc and you start to realize how incredibly important this becomes.
  • p191: spam’s up? I would call many ecommerce shopping emails spam but that’s not what we’re talking about here. Where’s the startups that lets you unsubscribe from everything with a push of a single button?

VI. China

  • p195: China has nearly 4x the number of people of America with an economy and excess cash flow growing at staggering rates. Personally, I feel they’re iterating faster than the US and Rest of World in key innovation areas like AI, Space, Manufacturing and a single mobile app for the entire country. Watch out when they launch their own cryptocurrency that lets them optimize their monetary policy in real-time. I’ll invest in that ICO.
  • p196: service output is going up and to the right.
  • p197: it’s no secret that private enterprise fuels growth. China just proved it in a huge way over the last decade.
  • p198: focusing on health care (large population) and IT (the future). Smart.
  • p200–203: almost everyone in the country is connected to the internet via mobile and they spend almost all their time on it. Hello WeChat.
  • p204: No surprise on WeChat, but look at Tencent’s share of time. Crazy. More than half of China’s internet time is spent on a Tencent property, with a bit more on Alibaba to buy stuff. That’s like Facebook + Amazon in the US.
  • p209: live is a big deal over there, and has become an economy in and of itself. People are supplementing their work income with 24x7 livestreaming of their lives and getting paid by viewers with stickers and weird emojis. It’s also the place where many big companies are investing their R&D dollars because gaming and live are providing higher returns than investing in their own businesses. Isn’t that crazy? I take the money I make from my business and invest in this app over here because I’m going to get more money back. At what point does it suck up a lion’s share of the entire economy?
  • p210: Live is money.
  • p213: bikes are big in China because they’re cheap, don’t take up space, and they can fit in small spaces.
  • p216: bikes as the last mile of the commute. Interesting that people take a portfolio approach to traveling from A to B.
  • p219: expect this to explode when China releases their cryptocurrency.
  • p223: this is a great roundup chart to see which monetary services are getting the most users across China. Payments and Wealth Management, of course.
  • p229: China’s ad market doesn’t seem as fervorous as America’s. I have a feeling because the economy is more about direct payments.

VII. India

  • p233: it’s as big of a population as China, but a much smaller economy as their main export is software engineers instead of manufactured goods. Be careful though, it could be a sleeping giant as the GDP growth rate is the same as China. The US, while big, is only growing about 1/4th as quickly.
  • p234: far fewer of India’s population is connected to the internet. Only about 1/3rd compared to about 3/4th in China and almost everyone in the US. Internet penetration is highly correlated to the wealth of a nation.
  • p235: Android’s big there.
  • P236: smartphone shipments have basically flatlined.
  • p237: getting access to the internet is expensive for the majority of the population, hence the drag on the economy.
  • p242: China is selling India their phones.
  • p246: China is also India’s most-used web browser.
  • p247: US, China, India dominate the Android app store.
  • p251: expect internet usage to continue to increase based on bringing costs down which will have a positive impact on India’s future economic prospects. Internet > Education > Skills > Compensation > Productivity > GDP
  • p252: we’re all just humans regardless of the (country’s) name on the door. So yah, streaming music and video is big in India too.
  • p254: India’s leadership is focused on the right thing, see p251 above.
  • p260: a singular digital ID is smart. What is it in the US? Driver’s license and Facebook. In China? WeChat profile.
  • p262: payments and infrastructure are being built now. What happens when it’s fully built out?
  • p266–267: mobile is where all the time spent and no surprise that half is on entertainment with 1/3rd on social. Regardless of nationality, we’re all still humans.
  • p270: a massive portion of India’s compensation is spent on health care. Decreasing this cost + increasing access to the internet are the main two levers to increasing economic well being.
  • p274: VCs are investing in startups, helping to drive growth, though it has flatlined in recent quarters.
  • p278–279: I know that India’s young population presents a large opportunity and know some big transportation companies looking to capture that market.
  • p281: oh god, are you still with me? If you’ve happened across this chart, I think we’ve lost you. My fingers just got tired looking at this one.
  • p284: the US leads in number of years of schooling, and India trails far behind. Internet = education because it shows what’s possible and gives you the resources for free to get started.
  • p286: females in India aren’t working.

VIII. Healthcare

  • p289: have a digital helper analyze, diagnose, and prevent issues with your health is the holy grail. We’re close to that future, but knowing how lagging the healthcare industry is with tech overall, it’s going to be an uphill battle unless individuals take matters into their own hands. For examples, CIOs at hospital systems are focused on reducing sepsis which directly saves lives, but moving medical records and software to the cloud is a painful, decades-long process.
  • p292: health IoT including wearables are likely to expand in the coming years. Eventually the whole nanobots in your blood helping cells fix themselves in real-time is a nice thought but we’re a long, long way off from that.
  • p293: more blood and health tests are now available through labs, which means diagnosis is cheaper, faster, and more accurate.
  • p294: pay attention to sensors that are included in wearables. Without a sensor, there’s no data to analyze. So, the most pervasive (an accelerometer) will only give you number of steps. Not very helpful in helping you manage your health but I suppose it’s something. Once we get into heart rate, blood sugar, metabolism, etc we will start to unlock some pretty cool stuff. Then it requires a highly efficient, but low-powered AI like Biologic Intelligence to understand it all and detect anomalies in real-time.
  • p295: I’d say Google through Verily and Apple with their Watch wearable are best positioned from the big tech companies. Interesting that Amazon doesn’t really have a health-centric product yet. Now that they own Whole Foods and are pushing into groceries and food, expect that trend to continue into personalized health as part of a Prime subscription.
  • p299: patient data is necessary for understanding someone’s health but the flip side is pre-existing conditions and the impact it might have on you getting health insurance coverage, the price of your premium, and getting services paid for. Double-edged sword, this.
  • p302: look at this crazy exponential curve in medical knowledge. And there’s not much out there on the impact food we ingest has on our health. Expect this trend to continue into the future as our population gets bigger and older.
  • p304: personalization isn’t just for targeted advertising campaigns, biomarkers from DNA sequencing can be used to target clinical trials to get you the help you need from new procedures, drugs, etc.
  • p305: data, it works.
  • p306: the young are paying attention earlier in life. They’re listening to their parents and grandparents ailments and stories.
  • p309: a friend founded a telepharm startup called, well, Telepharm. He was early on the trend, but when you can’t get all the way to the doctor in person, live streaming is a great second best alternative.
  • p311: back to the tech adoption curve that I linked to at the very top of this analysis, of course we expect health to be adopted quickly. As we like to say in tech circles, find a problem and offer a solution. Health, when it’s a problem, is all you care about and are price inelastic.
  • p313: costs are decreasing.
  • p314–318: number of genomes sequenced are increasing. Just wait until we start capturing people’s connectomes.

IX. Innovation

  • p320: I take issue with this statement. It assumes that innovation is all it takes to raise capital, get big, and sell product. It doesn’t. Even scientific invention isn’t enough. You need to have relationships. Lots and lots of relationships to get capital from investors, corporate coffers, and customers. Then and only then does a product matter that someone can buy.
  • p322–325: US and China are dominating tech. The question is who’s iterating faster. Because that’s the slope of the line. And whichever line is steeper, predicts the winner. I have a gut feeling I know who it is.
  • p329–331: IPO & VC, a confusing chart with way too much going on but I think the story is it’s been more or less steady for about a decade. M&A is about the same, even though it cuts the timeline in the chart in half. Be careful, your graph axes.

X. Macro

  • p336: in the US we pay about 1/3rd of our income to the government. The question is what is it spent on. Unsurprisingly, health care for older people, defense so nobody hurts us, and interest on all that debt because taxes don’t cover the spending.
  • p337: the USA is not a healthy business if you look at it like a startup. Look at all that red.
  • p338: debt is highest during war. But that money gets spent somewhere.
  • p340: the US has the highest debt level in the world. China doesn’t even make the list because they’re printing money doing all the world’s manufacturing.
  • p341: giving people money and services for doing nothing is a great way to get an entitlement mentality. That’s why they’re called Entitlements and that’s where a lot of America’s income is going. It does help people who want to do something for that income, but are physically or mentally unable. That’s part of what a government is for. To help those citizens who can’t help themselves. But the fact that it’s increasing so rapidly is troubling.
  • p344: debt is debt. Most Americans have a bunch of debt in houses. Second is student loans. So living and educating. Not a great sign if you want to grow.
  • p347–348: immigrants account for a lot of value creation.
  • p350: as a world, we’re better off than we were. That’s progress as a species and something to be positive about.

Sean Everett

Read more than 750 current and emerging tech analyses on The Mission’s other publication, Humanizing Tech.


If you enjoyed this story, please recommend and share to help others find it! Feel free to leave a comment below.

The Mission publishes stories, videos, and podcasts that make smart people smarter. You can subscribe to get them here.