Roger and Mike's Hypernet Blog

by Roger McNamee "Hypertext Mark-up Language 5 (HTML 5) is just a programming language." A famous investor said this to me 18 months ago. The statement is true, especially if you delete the word "just." But it misses the point. As programming languages go, HTML 5 may be unusually disruptive. Not only is it the platform for the next generation web, HTML 5 is likely to be the foundation for Hyperweb services on the Hypernet, which encompasses not the web, but also the app model and whatever new technologies emerge in the online world. ## The web needs HTML 5 ## Technology is one factor in favor of HTML 5, but timing may be as important. Apple's success with iOS has left HTML 4 market leaders Google, Microsoft, Facebook and many others with no profits in app-based device market, which now accounts for half of total connected devices. I believe it's too late for HTML 4 market leaders to build great businesses in the app world – Apple owns iOS and Google has open sourced Android to profitless prosperity – but HTML 5 is a new and open platform with the potential to compete successfully against the app model. HTML 5 represents new life for the world wide web, but it will almost certainly not be a smooth transition from HTML 4. HTML 5 is still in its infancy. Important functionality – such as that needed for commerce – has not yet been enabled. Even though it is not ready to replace HTML 4 on wired PCs, HTML 5 enables new and wonderful experiences on mobile devices. Ironically, the coolest HTML 5 apps run only on Apple’s iOS (there is no standard HTML 5 for Android). ## HTML 5 enables new content experiences ## My team has created two HTML 5 apps that clearly illustrate the potential of HTML 5: http://www.moonalice.com: - HD video streaming without a buffer over 3G wireless networks. [We broadcast 100 concerts a year and store them all in an archive on the site.] - Music player with 400 shows of audio These are just snowflakes on the tip of a very big iceberg. ## Pendulum of technology swings both ways ## In the decade before the iPhone, the technology of web and the behavior of consumers combined to commoditize most forms of online content. Fantastic for Google’s profits, commoditization was otherwise bad for business, especially the business of those who make content for a living. Story-level engagement dropped dramatically, as consumers navigated from page to page with increasing speed. In Google's world, publishers at all scales are captive to search engine optimization. Google transforms all content to a few lines of text in a common font, which approaches maximum commoditization. When Apple introduced iOS and the app model, content creators took advantage of new tools and a self-selected audience. Apple encouraged higher production values in apps, effectively favoring brands over commoditizers. This shift is consistent with history. Since the early 80s, digital content has swung like a pendulum between commoditized content and highly differentiated content. Before 1984, content appeared either as green ASCII text on a black background or amber ASCII text on a black background. In 1984, Apple introduced the Mac, with Adobe's PostScript enabling a What You See Is What You Get (WYSIWYG) display. This ushered in a 14-year wave of rising production values. Desktop publishing, Windows, PowerPoint, PhotoShop, Director, Flash, and Acrobat were killer apps in that era. When the web took off, finding content became the top concern, enabling Google's success with index search. Google quickly became the undisputed leader of the web, and the web economy optimized itself for Google's algorithms. The focus on SEO reduced the incentive to invest in tools, so HTML 4 (and Flash) remained predominant far longer than would otherwise have been the case. Static tools reduced the opportunity for innovation, especially in production values, which ultimately commoditized nearly all content on the web.

## iOS was a bet against the web ## When Apple introduced apps, it effectively bet against the web and Google. Apple's insight was that apps would enable consumers to find the content they want without the endless delays and unreadable type that characterizes the web experience on 3G smart phones. The bet paid off so well, that every smart phone vendor was forced to imitate the app model, including Google. For publishers, Apple’s app model enabled a modicum of differentiation. It also pushed the pendulum of technology away from commoditization. The impact is apparent in consumer engagement. In the search engine world of HTML 4, a huge amount of content gets discovered via search engines, and the engagement on these stories is typically 15 to 20 seconds. There is no way to monetize 15 to 20 seconds of engagement. In the mobile world of iOS, constant interactions with the web are undesirable, due to the poor data performance of 3G networks. As a result, consumers get most of their content in the form of apps, typically spending 2 to 5 minutes per text app, and far more for games. ## iOS + HTML 5: Publishers need both ## As the tool sets in mobile improve, app engagement could rise significantly, boding well for monetization. All of this makes iOS more attractive for publishers than the HTML 4 web, but there are costs: - no leverage from open source - smaller pool of developers than HTML - loss of control to Apple through the AppStore - 30% Apple tax on products sold in the AppStore - not compatible with desktop These factors are material, both in terms of cost and control. They create an incentive for content creators to move at least some of their business off iOS. HTML 5 starts with a cost/control advantage, but holes in functionality remain a barrier to adoption. Early adopters – including the New York Times – are investing in the belief that those holes will be plugged within a year or two. ## HTML 5: Appification ## HTML 5's advantage over HTML 4 is in its incorporation of the functionality of Adobe Flash into the language. This is a bigger deal than it may appear, as Flash is kludgy, buggy, and slow. By incorporating the Flash functionality into the HTML standard, HTML 5 gets additional benefits, including the ability to search any pixel on the page and the ability to "appify" any pixel. Not only will HTML 5 enable new page layouts, it will allow for better optimization of ads to content and users. It will also enable new “app” models, a few of which are as follows: - app delivery - e-book/Instapaper: just download content to your home screen - replace bookmarks: just download images, infographics, quotes to your home screen - layers: publishers can preload search results (bios, company information, etc.) into content so that consumers can get more info without leaving the page - longitudinal search: targeted search of a publisher’s past stories on a topic across time without leaving the content

The beauty of these new "app" models is that each can monetized, in most cases at rates better than the current web standard. Imagine you are reading David Pogue’s technology product review column in the New York Times. Today, the advertising on that page is pretty random. In HTML 5, it will be possible for ads to search the page they are on for relevant content. This would allow the Times to auction the ad space to companies that sell consumer electronics, whose ads could then look at the page, identify the products and then offer them in the ad.

One might even imagine Facebook creating a Mobile Connect product that stores private identity and transaction information, such as Amazon One-Click settings. If Amazon were the buyer of the ad, then the ad could feature One-Click buttons. This way the publisher could create demand and satisfy it without the consumer ever leaving the page. That isn't advertising in the traditional web sense, it is lead generation, which carries a much higher value in the market. This won’t happen overnight, but it will happen. ## New rules, new opportunities ## Even though HTML 5 is the next generation in web technology, I believe it will disrupt the current business model of the web. To compete with native apps, HTML 5 has to deliver engagement. To deliver engagement, HTML 5 must enable higher production values – and differentiation – than prevails in today's HTML 4 web. It should favor content publishers – and the tool vendors who support them – over the gatekeepers who currently dominate the web. There is no technical reason why Google, Microsoft, Yahoo, and Facebook cannot re-invent themselves, but the cultural barriers to change are huge, as has been made clear by all four companies’ inability to port their business model to the mobile sphere.

by Roger McNamee

In December 2010 I decided to open source my investment strategy, in the form of a slide show and presentation called Ten Hypotheses for Tech Investing.  When you open source ideas, you expose them to improvement.  I presented the Ten Hypotheses to many smart people, including executives at Google, Facebook, Twitter, Yelp, the New York Times, Wall Street Journal, NBC, and many others . . . and they shredded them.  It was fantastic!!!  

From that process emerged ideas like the Hypernet.  While I characterize the Hypernet as a hypothesis, it already exists.  We use it every day.  In reality, the first four of the Ten Hypotheses are new interpretations of the present, rather than predictions about the future.  This blog post will focus on those four hypotheses.

* * *

The world thinks about the online world in terms of a network layer (The Internet) and a software layer (The Web).  

The online world is no longer so simple.  Apple introduced the iPhone in January 2007, and with it a new model of online user experience: apps.  Today, app-centric smart phones and tablets represent half of all connected devices.  Half.  On app-based devices, the Web is just one application among many, rather than the center of online activity.  It is a huge change.

Since the online world now consists of two pieces of equal size, I believe the traditional vocabulary is obsolete.  I have proposed that we refer to the new network layer – the sum of the wired Internet and the mobile data infrastructure for cellular and wifi – as the Hypernet.  Its software counterpart is the Hyperweb, which today includes the traditional Web and app model, but which may evolve to include other technologies in the future.

Why do we need new vocabulary?  Everything about the app market works differently from the Web.  The failure to recognize this is one reason why Web leaders like Google have been unable to build profitable businesses around apps.  When companies incorrectly define their market – as the railroads did in the face of competition from trucks and jets – they leave themselves vulnerable.  Hence, the need for new vocabulary. 

When Google started in 1998, it transformed Web monetization with index search.  To make that happen, Google adhered to the cultural norms of the open source community by focusing on the long tail, making everything free, commoditizing content, and encouraging an “anything goes” atmosphere.  This was no problem when the Web was small, but once it hit critical mass the “techie” Web experience began to lose its allure.  One factor was the static nature of the Web itself.  For more than a decade, the web has been stuck with HTML 4 as its platform and Flash as its media format.  No wonder content was commoditized.  There was no way to differentiate without spending more than customers would pay.

Any time a product stagnates, as Windows did after 1995 and the Web has done in recent years, it is vulnerable to competitors.  In this case, the competitor was Apple and platform was iOS.

What Apple did with iOS was to create a user experience that was “one minus the Web.”  Instead of open source, Apple used a proprietary operating system (iOS).  Instead of long tail, Apple focused on branded products.  In addition to free, Apple encouraged the development of paid applications.  Instead of commoditizing content, Apple enabled limited differentiation among apps (but huge differentiation relative to Web).  Instead of “anything goes,” Apple offered a secure environment.  On top of that, Apple charged $400 to $800 for the hardware that delivers its user experience . . . using content and data that for the most part are already free on desktops.

What happened?  Consumers adopted Apple’s model faster than any technology in history.  When you include the dollars spent on hardware – how can you not do so? – it is possible that Apple’s ecosystem may be larger than the Web ecosystem.   Whoa!

For the leading Web companies – Google, Microsoft and Facebook - the economic consequences of Apple’s success have been masked by other factors, but they have been enormous.  We’ll start with Microsoft.  Before the iPhone, 95% of connected devices ran Windows.  I don’t know what Windows’ share is today, but it must be less than half that, as Microsoft’s share of app phones is roughly zero.  For all intents and purposes, Microsoft lost half its addressable market to app-based devices.  The same is true of Google.  Before the iPhone, Google accounted for roughly three-fourths of index search, which accounted for about 90% of Web search.  The other 10% of Web search was controlled by a rapidly growing group of specialized search engines: Wikipedia, Facebook, Twitter, Yelp, Realtor.com, Match.com, etc.  In the app model, customers use apps to search, not Google.  It is my understanding that for a given consumer, index search happens with the following frequency by platform: desktop web 100%, iPad 10%, iPhone 1%.  Given that high value transactions are moving rapidly to mobile, it seems unlikely that index search will provide Google with the control point it has grown accustomed to.  Margins must fall.  Notwithstanding the huge unit volume of Android, Google has yet to produce any profits from smart phones.  Even though Google’s Web business continues to grow, the company has lost half its addressable market.  The third Web giant is Facebook.  Like Microsoft and Google, Facebook has not been able to migrate its profit model to mobile.  All three derive their online profits from a business that will continue to lose share in coming years and may even begin to shrink.

It’s not too late for Microsoft, Google and Facebook.  All three can respond to this disruption as Apple did a decade ago.  Microsoft may already have made the critical first move.  The acquisition of Skype gives Microsoft the only product with a larger engaged online audience than Facebook.  Skype is perfectly positioned for mobile, especially in the developing world, and Microsoft got it for a small fraction of its strategic value.

To a much greater degree than Microsoft, Google has experimented with new business models.  Unfortunately, some of these experiments have come off the rails.  I believe Google has irretrievably messed up Android’s architecture and business model, and made a terrible mistake in buying Motorola, but the company still has fantastic human capital and more cash than Croesus.  But it has to recognize that it no longer controls the most important value streams in the online world.  If Google wants to be a market leader in the future, it will have to reinvent itself.  If it were to make a proprietary Motorola phone with Chrome OS, then Google would have a chance to leverage its strengths to compete with Apple.

Facebook was very slow to figure out mobile, but it has far fewer barriers than Microsoft and Google.  The company has just introduced mobile ads.  If Facebook were to introduce mobile use cases that are valuable (e.g., peer-to-peer wireless “friending”), it might find that the business model in mobile is better than in desktop.

The key point here is that Apple’s position of market dominance is unstable.  Giant tech companies with giant cash positions have been left out of the Apple ecosystem . . . and they will eventually seek a way to get back into the game.  Google, Microsoft, Facebook, Cisco, Intel, Oracle, and others have the scale, the cash, and the compelling incentive to develop an alternative to Apple’s app model.  One or more of them may succeed.

But success for these companies will not be significant if it is measured by market share of Apple’s ecosystem.  They need a new and really big ecosystem.  I believe the new ecosystem will be based on HTML 5.

I spent six months trying to convince entrepreneurs and investors that HTML 5 will change the world before I realized that my message was not getting through.  So I spent the past year developing and deploying HTML 5 functionality on behalf of my band, Moonalice.  One of our HTML 5 tools enables live HD video over 3G cellular with no buffer.  Another enables fans of Moonalice to listen to any of 400 shows.  Both applications are live on http://www.moonalice.com, but they only work on iOS.  Why? Android does not have a standard HTML 5 implementation.  Most people at Google don’t know that.

I have been told that HTML 5 is “just another programming language.” For content creators, however, HTML 5 offers the possibility of production values previously unheard of in the online world.

I am convinced that the pendulum of technology swings between commoditization and high production values.  Before 1984, there were no production values in tech content: green ASCII text against a black background was the standard.  Then Apple introduced the Mac, with Adobe’s Postscript.  Postscript enabled WYSIWYG, desktop publishing, Photoshop, PowerPoint, Acrobat, and increasingly high production values.  From 1984 to 1998, content looked better every year.  Then came Google and the commoditization of content.  That lasted for ten years before Apple’s app model enabled limited differentiation.  HTML 5 will remove the limits.

Online content has depended on Adobe Flash for video and animation for more than a decade, but the limits of Flash are significant, especially in mobile.  It was never designed for today’s mobile use cases, much less the ones that are coming.  Each instance of Flash on a webpage increases load times and instability.  In mobile, the overhead is so great that that live video experience on 3G networks is exceptionally lame. 

HTML 5 incorporates the functionality of Flash into the HTML standard.  In mobile, the overhead is tiny in comparison to Flash.  But that’s just the beginning.  As the tool set gets fleshed out, HTML 5 will transform the creation and presentation of content.  HTML 5 will be a creative canvas unlike any in the past history of technology.  Unlike the HTML 4 world we live in today, where every page looks similar and all can be compared in an index, HTML 5 will enable huge variation in production values, from the sepia of Kansas to the 3-D Technicolor of Oz.  Before Google, higher production values translated into higher economic value, and I believe that will be the case again, only more so.  In fact, the process has already begun with iPhone applications.  Major League Baseball has done a brilliant job of translating production values into revenue.  What they and others will be able to do with HTML 5 must be left to the imagination for the moment, because HTML 5 is still a very young language. 

by @m2jr

Even though our blog is less than a month old, we have gotten a lot of great feedback and suggestions from readers. Apart from questions like, “Do you really know how to surf?” many have asked for a more in-depth explanation of Roger’s 10 Hypotheses for Tech Investing.  We’ve also had requests to offer tangible examples of products and companies that that line up with these hypotheses.  To facilitate this, we are going to put together a set of “Exploration” posts over the coming months.  My hope is that they will bring the 10 Hypotheses to life and encourage like-minded people to engage in the discussion. 

Big Change #1:  The Hypernet Emerges from the Web + Cellular + WiFi

It’s hard to believe that only 5 years ago, Microsoft Windows ran on over 95% of Internet-connected devices.  For all practical purposes, smart phones were not even web capable.  There were no tablets as they are currently conceived.  It was a pretty simple Internet world.  PCs with browsers and windowed interfaces and applications accessed a single worldwide web with linked pages indexed primarily by Google.

But the advent of the iPhone changed things.  Apple transformed the smart phone from a carrier-centric device that could make calls and handle e-mails to an always-on computer ecosystem with a new way to access the web and a new application model.  Soon, Google and others created an alternative post-PC ecosystem and the race was on.

That Apple changed the smart phone market is well understood.  One only has to look at the collapsing fortunes of the leading smart phone vendors such as RIM, Palm, and Nokia to see this as well as Apple’s stock price since Steve Jobs’ iPhone keynote in 2006.  But the more important and subtle change has been that the new devices that we have put into our pockets and purses have changed the fundamental architecture of the Internet.  This is Big Change #1.

From this first big change emerges the Hypernet.  More precisely, the Hypernet is the physical infrastructure that results from combining the Internet with cellular and WiFi.  At present, half of the nodes are computers and half are smart phones, but the balance is shifting away from computers.

Big Change #2:  Leveraging the Hypernet architecture, the Hyperweb moves past browsers and pages

The potential of an architecture that merges the Internet plus cellular plus WiFi is bigger than it seems at first glance – it will change the way we interact with digital products and services. 

More precisely, the Hyperweb is the new user experiences enabled by a world of billions of nodes connected to millions of clouds.  Another key aspect to the Hyperweb is our belief that HTML5 will be a fundamentally important building block.

How will the Hypernet/Hyperweb Change Digital Experiences?  Examples

Now that we have laid the foundations for the hypernet and hyperweb, let’s explore some specific examples of how the technology landscape is likely to change:

Hyperweb Example 1:  New Ways of Hyperlinking

Like many people who follow blogs on the Internet, I found Dave Winer’s post on Why apps are not the future to be an interesting read.  In the article, he discusses how a key advantage of the web over apps is linking rather than the silo approach of apps in which only the app provider has control of what goes in and out.  Not long after Dave’s post, George Colony at Forrester gave a speech at LeWeb, which proposed an alternative viewpoint, declaring, “The web is on life support.”

Both points of view have make solid points.  Dave’s contention that an app world where everyone builds a separate skyscraper and allows only the people inside each building control what happens points out a lot of the limitations of the app model.  But George argues persuasively that much of the web is as messy as it is open and that users like the safety of the app world.  He also points out that the continued rise of processing power at endpoints as well as continued exponential increases in storage make the web model as we know it today not the best use of the computing resources at our disposal.

My view is that the Hyperweb will reconcile Dave’s affinity for links with George’s view of a future “App Internet.” I think that this is due to the fact that the hyperweb redefines linking in exciting new ways.

The diagram below shows how we think this will play out.  In the “simple” web experience, linked web pages are viewed via browsers primarily on Windows computers and are indexed by Google.  The hyperweb broadens our definition of linking.  A whole new set of instrumented smart nodes interact with a rich variety of connected clouds, apps, and services which can be hosted, local, or hybrid in terms of where they reside.

The examples of this today are early in their evolution, but as the industry begins to advance and build on the concepts of the hyperweb, things will get really interesting. In this and a few upcoming posts, we will start to offer ideas and examples, getting a little further out into the future as we go along.

Hyperweb Example 2:  ifttt

In the spirit of full disclosure, FLOODGATE is an investor in ifttt’s seed round. Ifttt stands for “if this then that.”  It is designed to put the early hyperweb to work for people by creating tasks that follow a simple “if this then that” structure.  “If this” is the trigger, such as “if there will be rain today,” or “if I get an attachment in Gmail,” or “if I take a photo on Instagram.”  “Then that” is the automated action, such as “then send me a text message”, or “then store it in Dropbox”, etc.

Ifttt “recipes” can link services together to enable early types of hyperweb experiences.  Below are some examples:

Right now, ifttt is in its infancy and in early Beta form.  All caveats apply.  But it pulls together a set of ideas that I believe have a chance to be very meaningful.

Hyperweb Example 3:  Content Unbound

It’s amazing that 1080p HDTV quality has entered the mainstream at such a low cost.  It’s equally amazing that we put up with the awful user experiences that are the front-end of this awesome hardware.  Every TV I own has multiple boxes fed by proprietary pipes owned by large companies with a distribution lock on content.  Each box feeds the TV with wires that are tangled behind a cabinet in an almost incomprehensible mess.  And having a simple way to access the content with a remote?  Forget about it.

 In our view, the hyperweb offers the potential to transform the use of video content and TVs the way the iPhone transformed the value and utility of mobile devices.  This hypothesis is a little bit further out in time, but already several pieces are coming together.

Imagine if, instead of having a jumble of boxes with pipes that are owned by big companies and wires that are a confusing tangle with an array of incompatible remotes, you could instead migrate all of the functions of these boxes to your smart device.  In this world, a player or a streaming device or a DVR is just an “app” or a programmable service.  Imagine further that you could take these devices anywhere (throughout your house, a hotel room in another country, your friend’s place – anywhere there is a screen) and zap the content to the screen wirelessly.  This is an example of a hyperweb-centric way of re-imagining connected TVs.  Not only this, a world of connected TV content “apps” on smart mobile devices would liberate the capabilities in new ways.  When a Blu-Ray player and a TiVo and a Roku box are physical devices, they are far less flexible than if they could behave like programmable objects that can link together.  Someday, we believe that content interactions between apps should be as easy as dropping a photo from an app like Instagram into a cloud service like Dropbox.

We also believe that the war to build the perfect general set top box or the ultimate smart TV user interface is an example of how companies tend to look in the rear-view mirror.  For example, if I were Google, rather than create a set top box to compete with traditional TV providers on their turf, I would use Android on smart phones and tablets to wildly disrupt them by migrating the functions to the devices in a way that would delight consumers and turn existing value delivery networks upside down.  Apple, Microsoft, and Amazon could do something like this as well.

Another belief of mine: If you had to bet on who is more likely to create a compelling interface for video and entertainment content, would you side with the set top box and TV hardware makers or the folks who have built the new user experiences on smart phones?  My money is on the latter group and when this is combined with the empowerment offered through mobility and programmable linking, I believe that it is only a matter of time before content becomes unbound as a rich new hyperweb-powered set of experiences.

Key Takeaways

The  iPhone did more than change the smartphone market: Its success serendipitously changed the architecture of the Internet.

The Hypernet is the physical infrastructure that results from combining the Internet with cellular and WiFi.  At present, half of the nodes are computers and half are smart phones, but the balance is shifting away from computers.

The Hyperweb is the new user experiences enabled by a world of billions of nodes connected to millions of clouds.  Another key aspect to the Hyperweb is our belief that HTML5 will be a fundamentally important building block.

Hyperweb experiences will redefine our expectations of how we interact with technology.  Today, we interact with web services via browsers and windows and PCs and graphical apps on tablets and smartphones.  But in the future, there will be entire new experiences that result from a new way of linking the billions of smart nodes in our lives to the millions of smart clouds that will deliver services by themselves as well as through their interactions with each other.

We offered ifttt and the idea of decoupling TV content from boxes under TVs as two examples of new experiences that the hyperweb will enable.  But there are many more examples, such as driverless cars, multi-tier cloud services, and others that we will discuss in the coming months.

We are extremely interested in your views as well!  Make sure to get in touch with us with your ideas and comments.

Next up:  HTML5, Driverless Cars, and Multi-tiered Cloud Services

Note: This topic is partly related to the Hypernet but also related to where we are in the Social Networking cycle.  It’s an atypical post for this blog, but a lot of readers asked me about this and how it might relate to my prior post on Technology Waves.

It was my last quarter as an Engineering student at Stanford.  Since I had finished my requirements for a degree, I had a different academic and social focus than my first 3.5 years. I planned to go to Europe and stay in hostels with a Eurail Pass in the summer, so it seemed only natural to take a class on Italian Renaissance Art.  Next up:  Windsurfing for credit.  Unfortunately, the course was fully booked, so I would not be able to benefit academically for my exploits on Lake Lagunita. 

In what felt like a less exciting turn of events, an aspiring lawyer friend of mine talked me into taking an Intro class on Logic.  I didn’t expect to like it very much, but since I was suspicious of lawyers for the most part, I thought it would be good to learn a few things about how they practiced their craft.  I was really surprised at how cool it was.  One of the topics that fascinated me was how people use logical and rhetorical fallacies to win an argument. My favorite example was the “false dilemma” which is also called the “either-or fallacy.”  After that class, I learned to always be on the lookout for the various rhetorical weapons used by people to advance their agendas.

Are We in a Social Networking Bubble? A Question Based on a False Premise

In my opinion, “Are we in a social networking bubble?” is an example of a rhetorical fallacy.  It assumes that social networking companies are either wildly overvalued like it’s 1999, or correctly/undervalued and there’s nothing to worry about.  (As an aside, another favorite of mine is the question, “What’s more important, the team or the market?” The question falsely assumes that great teams aren’t more likely to find the better markets, but I digress.)

Framing questions honestly and in a truth-seeking way is one of the most important skills of the innovator or investor.  My hope is to tackle the question of social networking valuations with this in mind.  By mapping the typical progression of valuations and investor sentiment to technology waves, I hope to show that investors and the market almost never know how to correctly value important disruptive technology companies.  I’ll also describe how I think investors make money during different phases in tech waves.  Finally, I will suggest that entrepreneurs starting companies today are better served by finding the next coming wave, which Roger and I believe is the Hypernet.

Technology Waves Create the Fundamental Value

My point of departure for social networking valuations is the notion of technology waves, which I outline here.  For those who want to skip that post, the core premise is as follows: The technology industry is unique because the exponential increase in performance of storage, processing and networking ensures that a new supply of game-changing companies will emerge on a regular basis.  These companies usually cluster in waves, such as the PC revolution, client/server, the Internet, and social networking. 

We have observed that these waves tend to follow a pattern.  They start with infrastructure.  Advances in infrastructure are the preliminary forces that enable a large wave to gather.  As the wave begins to gather, enabling technologies and platforms create the basis for new types of applications that cause a gathering wave to achieve massive penetration and customer adoption.  Eventually, these waves crest and subside, making way for the next gathering wave to take shape. (See the figure below) 

In my view, technology waves are where the core value creation occurs in the technology industry.  Value is created on the supply side because innovative startups create disruptive new products that exploit ongoing exponential price/performance improvements.  Value is created on the demand side because customers come to recognize the value of the new innovations and incorporate it into their lives – gradually at first, and massively later.

Sentiment Waves Drive Valuations

But value creation is not the same as valuation. Valuations move like a sawtooth around the value that gets created when a new technology wave takes hold. (See the figure below)

In the early phases of a technology wave, massively powerful gathering forces exist below the surface, but only a handful of visionary technologists, entrepreneurs, and investors really see what’s starting to happen.  I call this the Gathering Phase.  At this point in time, new technology startups are typically undervalued and under-recognized, relative to their current and future potential value creation. 

Eventually, people start to realize that something fundamental is happening.  Many of us in the industry remember when people said, “Facebook may have a lot of traffic, but people are throwing virtual sheep at each other.  How is that a business?”  But in the latter half of the 2000s, people began to say, “Oh my gosh – Facebook is 1/3 of the entire Internet, is a monopoly private network, and will be a huge juggernaut.”  This is when market sentiment and valuations transition to the Acceleration Phase.  If strong but unseen undercurrents characterize the gathering phase, the acceleration phase is where sentiment catches up to and surpasses the reality of the technology’s value, creating varying amounts of “froth” in valuations.

Eventually, the wave of sentiment crests and people start to say, “These companies are comically overvalued.  Time to sell.”  This is the Correction Phase, which starts out with a correction back to rational valuations, but usually ends up falling even below what the companies are intrinsically worth.  Eventually, the valuations bottom out and converge back upward to their rational values.  But by then, the gathering forces of the next big wave are usually in place and the interesting opportunities for massive growth lie elsewhere.

Thoughts on the Gathering Phase

First off, I should go on the record and say that I really liked Fred Wilson’s post about “Herky Jerky” investing.  In my opinion, great investors find a framework that plays to their strengths and stick to it in good times and bad, with occasional but infrequent opportunistic decisions to break the rules. Having said this, my favorite time to be a technology investor is at the Gathering Phase.  Without getting too academic, I think this is the best time for investors and entrepreneurs to be non-consensus and right.  An opportunity is created because there is a gap between the intrinsic value of new technology that’s being created and the World’s ability to recognize it.  Since the world has not yet seen the future, there is an undersupply of startups and an undersupply of investors.  In the social networking wave, the Gathering Phase started around 2002/2003 when the “PayPal Mafia” scattered and started exploring new concepts for the consumer internet.

The great opportunities during the ensuing couple of years were LinkedIn (well done, Reid Hoffman and David Sze), Facebook (kudos to Mark Zuckerberg, Peter Thiel and Jim Breyer), and Twitter (well-played Ev/Biz/Jack and Fred Wilson.)  These visionary founders and investors saw the trend of social networking before others and had the conviction to back up their vision by building or investing in some of the early, defining companies of the wave.

Thoughts on the Acceleration Phase

Some founders and investors are better at playing the momentum game and this is where the Acceleration Phase comes into play.  An opportunity is created because, even though valuations have become high, they are beginning to go even higher at a very rapid rate.  Success in this phase involves three key factors: First, you have to have the ability to invest in the companies early enough in the acceleration phase so there is room for the valuations to accelerate upward quickly (congratulations Yuri Milner).  Second, you have to have the “mojo” to get into the absolute best companies of the wave (way to go Andreessen Horowitz).  Overpaying for close-but-not-good-enough is really dangerous and potentially fatal.  Finally, you have to hope that you are able to exit in the acceleration window, before the correction occurs.  Time will tell who exits at a profit if the acceleration continues and who gets burned if valuations correct.

There is a lot of evidence that we are in the Acceleration Phase of the social networking wave.  Examples are an oversupply of startups, an oversupply of capital, faster investment cycles for venture funds, larger fund sizes for venture firms, the increased desire of public investors to buy into shares of private companies, and celebrities from Hollywood who have decided that they want to start tech companies and/or invest in them.  

In my opinion, the Acceleration Phase of social networking started when Goldman Sachs bought a big chunk of Facebook in late 2010.

Thoughts on the Correction Phase

If valuations play out as I think they will, we will see a massive Facebook IPO, followed by many awesome exits by other compelling social networking companies.  This will be followed by a set of exits and IPOs that are comically overvalued.  Eventually the market will see this and the valuations will drop precipitously.

Emerging Tech Companies are Almost Always Priced Inaccurately

I hope that the scenario I am laying out does not seem alarmist.  I am not worried about valuations per-se.  In fact, my core argument is different.  Namely, disruptive tech companies are almost always priced wrong.  They are either undervalued because the mainstream does not yet have conviction about the value of new innovation, or they are overvalued because the world has become overly enthusiastic after understanding the implications.  There are plenty of ways to make money during all of these times.  The key is to be objective about what’s happening and about one’s own capabilities.  In 1999, we had a bubble because the magnitude of the acceleration phase was extraordinary, but the cycle played out according to the same pattern.

Another related thought is that technology has been financed this way for the last 150 years.  Railroads were financed this way, as was radio, TV, cars, and other technologies that eventually reached the mainstream.  It seems that manias are the natural way for our economy to lower the cost of capital for innovative companies that emerge when a new disruptive technology takes hold.

Enter the Hypernet

What does this have to do with entrepreneurs and the Hypernet?

For entrepreneurs, the key message is to be really careful about doing a social networking startup in 2012.  The social networking wave is about to crest.  There are very few ideas and opportunities in this space that aren’t crowded.  There will be many opportunities for “quick flips” based on momentum, but the oversupply of startups makes it a very risky time to start a company in this area.

Using the surfing metaphor from my previous post, we believe that the crowded beaches of social networking startups and the frothiness of the water make it more compelling to seek the next gathering wave.  The Hypernet is still undersupplied and undercapitalized.  But more importantly, it is a wide open, blue ocean opportunity that has not yet been defined and can therefore be the basis of huge companies in the future, rather than speculative startups that are trying to time an exit window before it’s too late.

Key Takeways

Value in the technology business is created by technology waves.  The creation of value is driven on the supply side by the inexorable march of price/performance improvements in digital technologies.  It is driven on the demand side by customers, who see the value slowly at first, but rapidly later, which leads to mass adoption.

Valuations in the tech business start out by lagging the value that is being created in a new technology wave.  This creates an economic “surplus” for people who are smart enough to perceive the wave before others.  This occurs in the Gathering Phase.

Valuations in the tech business eventually exceed the value that’s being created and reach a peak.  This creates a different type of economic surplus for people who are able to invest in the very best emerging winners that are unavailable to typical investors.  This occurs in the Acceleration Phase.

Eventually, valuations come back to Earth in the Correction Phase.  They fall to their rational values and even below for a while, but eventually converge back to their intrinsic values.  However, by the time this happens a new wave is usually gathering.

Entrepreneurs should be cautious about starting new companies in the middle of an Acceleration Phase of a cresting wave.  Generally, they are better off finding the next gathering wave and a blue ocean of opportunity rather than building a speculative company in a sea of froth, based on a small idea that is designed to flip.

The question, “Are we in a social networking bubble?” misses the point.  The more important question is, "Where are we in the social networking wave, what is the valuation environment given where we are, what are the criteria for success, and what are the warning signs?"

After this post, I hope I never get asked about "bubbles" again :-)

For this post, I would like to offer a special thanks to the many legendary tech investors I interviewed in a 2011 “listening tour” about tech cycles.  Specifically, I would like to acknowledge Arthur Patterson of Accel, Paul Ferri and Andy Verhalen of Matrix, David Strohm and Reid Hoffman of Greylock, Bruce Dunlevie of Benchmark, Dick Kramlich of NEA, Kevin Compton of Radar Partners, Peter Thiel of Founders Fund, and Marc Andreessen of Andreessen Horowitz.

A few weeks ago, I had the chance to go surfing off the coast of Waikiki.  The waves there are perfect because they are just the right temperature and they are forgiving enough that I can get up on the board and have a great time, despite being pretty out-of-shape.  I have always been inspired by the metaphor of surfing.  On one hand, you are out in the ocean, paddling as hard as you can until you find the next wave.  It’s up to you to pick the right one and control yourself and the board the best you can as you ride the wave into the shore.  But there is also something vastly larger and more dynamic at work because every time you get up, the entire power of the ocean is behind you.  That awesome power makes the entire experience possible.

Waves in the Technology Business

The metaphor of surfing has guided my framework for thinking about the tech business ever since I was writing programs for the Apple II and the original IBM PC in Junior High and later as an entrepreneur and now as an investor.  In the digital world, rather than the forces of gravity and the tides and the water itself, the motive power comes from exponential technology factors such as Moore’s Law (which doubles the performance of computing every 18 months), as well as storage (which advances even faster), and networking connectivity (which is slightly slower).  Because of these powerful forces, new waves gather in the tech business all of the time.  Some waves are too small to be interesting.  Some (like the PC revolution, client/server, the Internet, and social networking) are so big that they spawn entire new industries and companies that change the world.  These waves guarantee that there will always be a new supply of innovative companies who push the technology industry forward.  (As an aside, it’s not accidental that Ann and I decided to name our investment firm FLOODGATE)

Technology Waves:  Overview

In my experience, every decade or so, we see a major new tech wave.  When I was in high school, it was the PC Revolution.  I made my career as an entrepreneur at the end of the client/server wave and in the early phases of the Internet wave.  Today we are at the mass adoption phase of the social networking wave.  I am obsessed with these technology waves and have spent a lot of time studying how they develop and what patterns can be observed.  I’ve also had the opportunity to talk to some of the great founders and investors in the technology business to refine my thinking further. I am going to spend the rest of this post describing some of what I’ve learned and connecting these thoughts to why Roger and I believe the Hypernet/Hyperweb is the next great technology wave.

We have observed that these waves tend to follow a pattern.  They start with infrastructure.  Advances in infrastructure are the preliminary forces that enable a large wave to gather.  As the wave begins to gather, enabling technologies and platforms create the basis for new types of applications that cause a gathering wave to achieve massive penetration and customer adoption.   Eventually, these waves crest and subside, making way for the next gathering wave to take shape. (see the figure below)

In the recent social networking wave, broadband penetration created the infrastructure for billions of people to be always connected.  Next came the enabling platforms from companies like Facebook, Twitter, and LinkedIn, who created the various types of social graphs and connection frameworks for people to socialize.  This was followed by applications such as Zynga that took advantage of these underlying platforms and connections, along with business software companies like Bazaarvoice and Jive.  A similar story could be told for every wave back to the PC revolution (which incidentally had infrastructure such as semiconductors and disk drives, enablers such as DOS, PostScript, and NetWare, and apps like Microsoft Office).

The Importance of Enabling Technologies

Another interesting feature of most tech waves has been that the enablers (Microsoft with DOS in PCs, Oracle with Relational Databases in client/server, Facebook with the social graph) have turned out to be natural monopolies.  They appear just as the early infrastructure has been built and create a way to translate the new capabilities into a platform for applications that eventually reach very large audiences.  Enablers are the special companies that convert the energy of a gathering wave into the opportunity for a new set of technologies to affect millions or even billions of people.  There are very few companies that end up playing this critical enabling role, but the companies that win end up being massively valuable.  There are many debates as to why this is true.  My belief is that there are so few because the technology community needs to agree on a few standard platforms to build an industry that can achieve mass adoption, penetration and scale.

What It Means for Entrepreneurs and Investors

One of the other features of this metaphor that I like is that it provides a good framework for investors and entrepreneurs.  For example, when I first arrived in Silicon Valley in 2005, many VCs were saying that consumer internet companies made no sense – that the dot-com bust proved that “you can’t make money with just a lot of eyeballs.”  But what this analysis missed was that we were in the beginning phases of a new wave called Social Networking.  The Internet wave had crested and a new wave was gathering.  Since I was Angel investing with this framework in mind, my focus was to find the enablers of social networking (or what many called “web 2.0” at the time).  Unfortunately I moved to California too late to invest in the Angel rounds of Facebook and LinkedIn.  I wrongly assumed podcasting would be an enabler, but fortunately I invested in a podcasting company called Odeo, which became an incubator for Twitter.  Twitter turned out to be one of the key winners in the enabling technology/platform phase.

In late 2006, we decided that the enabling forces of social networking had already gathered and consolidated and began shifting our focus to applications that leveraged the connections that were being created by Facebook, LinkedIn, and Twitter.  Examples of such companies we funded include Bazaarvoice, BranchOut, Chegg, MassRelevance, Modcloth, Socialware, Spiceworks, and many others.  During this period of time, we saw many “Facebook-killers” or “Twitter-clones” or “LinkedIn for another segment” types of companies, but we passed on them because we thought that the enabling phase was over.

Why It’s Too Late to Start a Meaningful Social Networking Company

Today, we believe that it is too late to start a meaningful company in the social networking wave.  When the cycle started to move from enablers to applications, it was a very good time for two reasons.  First, it was late enough that companies could make a bet on a few platforms that had achieved scale and were likely to achieve much greater scale.   But, it was also early enough that the map of the world had yet to be drawn for apps.  This meant that companies could acquire users for much less money and monetize them better because there was less competition.  These companies also had the opportunity to build network effects among their users and create sizable entry barriers, particularly in terms of scale and distribution strength.

The social networking application space is now very crowded.  The wave is about to crest.  There is more competition for users which means that the costs of acquiring users is increasing while the lifetime value of users for new products is under more pressure.  This is a natural part of how a tech wave evolves.  In the early days, there are very few “surfers.”  There is plenty of room for everyone.  But pretty soon, lots of people crowd to the same beaches and it’s just not the same.  Anyone who surfs during a busy time in Santa Cruz has probably experienced a crowd of angry surfers acting territorial and saying “dude – get off my wave.”  That’s what life is going to be like for new social networking startups going forward.

Some people will misinterpret this point.  I am not saying that social networking is “dead” or that we are in a “bubble” or that the current crop of major social networking companies aren’t valuable.  My business is startups and where the opportunities are to start meaningful game-changing companies in the future.  Social networking will matter for sure.  But entrepreneurs from this point on should look at it as a “feature” rather than the next great startup opportunity.

Enter the Hypernet

One of the reasons I am so excited about this point in time is that, in the last few years, the new infrastructure has been put into place for a major new wave. Roger and I think the new infrastructure combines the web plus cellular plus WiFi (which we call the Hypernet), combined with potential new user experiences that involve billions of nodes and millions of clouds (which we call the Hyperweb).  We believe that these are the prime forces of a massive new gathering wave.  In the next few years, a handful of entrepreneurs will begin to build the enabling platforms that define some of the great technology companies of tomorrow. We also believe that this wave will be much larger than the social networking wave because it has a larger hardware component to it.  Generally speaking, when a wave has hardware components to it as well, there is an even larger ripple effect and the value multipliers are even stronger.

Part of my purpose in working on this blog is to find the startup founders who have unlocked a fundamental insight for how these enabling technologies and platforms might take shape.

Key Takeways

Technology waves are a powerful framework for entrepreneurs and seed investors who wish to start or invest in companies that have a chance to be extraordinarily valuable. In the future.

Technology waves evolve along a pattern, starting with infrastructure, moving to enabling technologies, and then kicking off a phase of applications that take the wave to mass adoption. The key is to recognize what phase of the wave you are in and time your startup or investment strategy to align with its realities.

The enabling technologies and platforms that win are natural monopolies  There are a small number of them but they create massive value.

Applications build on the enabling technologies and platforms that consolidate and take the technology of a new tech wave to he masses.  The applications companies that win are those which wait long enough to make the right platform bets, but enter early enough that the map of the future is yet to be drawn.

Social networking is the current Tech Wave, but it is shifting from a new cycle where there are opportunities to build big startups, to a mass-adoption phase where social networking is now a “feature” rather than uncharted territory.

To borrow a quote from Wayne Gretszky, the best entrepreneurs “skate to where the puck is going.”  Roger and I believe the puck is heading to the enabling technologies and platforms phase of the Hypernet and Hyperweb.  We believe that some startup team right this minute is working on a project that will likely create a massive enabling technology or platform that powers the mainstream adoption of this new cycle.

If social networking can be thought of as a part of the ocean that is crowded with surfers, the Hypernet and Hyperweb is the opposite.  Right now there are very few of us paddling together in the water.  We are having a great time because we have seen conditions like these before.  There is still room for a lot more people and we hope that those of you who are excited about these ideas will come join us!

The driverless car is a great example of a new, unexpected experience made possible by the Hyperweb (see Roger's hypothesis #1).

The car will be a node, and it will receive data from multiple clouds.  For example, it might take orders from its owner to go pick up the kids, and it will monitor traffic servers to choose the optimal route.  

In his post Koushik Dutta points out that the driverless car will fundamentally change how we think about car ownership, sharing, parking and insurance.  This wouldn't have been possible before the Hypernet era.

Read more on TechCrunch and Hacker News.

Roger’s 10 hypotheses:

“Next” Web Architecture = Hypernet + Hyperweb

Enterprises Adopting Consumer Technology 

Index Search is Peaking

Apple’s App Model Has Undermined Economics of HTML4 Web

HTML5 is Game Changer for Publishers

Tablets Are Hugely Disruptive

First Wave of “Social Web” Is Over

Smartphones in US: Apple + 7 Dwarfs

Wireless Infrastructure Is a Competitive Threat to US

The big opportunities for entrepreneurs and investors are created by new technology cycles.  Examples from the past are the PC Revolution, Client/Server, the Internet, and Social Networking.

Roger and I believe that social networking has moved from an early stage investment opportunity to a feature – it’s time to find the next gathering wave.  We think it will come from a new infrastructure that combines the web plus cellular plus WiFi (which we call the Hypernet), combined with new user experiences enabled by billions of nodes and millions of clouds (which we call the Hyperweb.)

So far, Roger has done most of the work and most of the postings on this topic, but we will both be contributing our ideas on the Hypernet/Hyperweb in this blog.  We may not always agree with each other on everything (and may even change our minds about some things), but we broadly agree on the important thing – that we are in a new technology cycle that will be extremely exciting.  We want to engage in a conversation with the early pioneers who wish to build the next trillion dollars of value in the technology business.