My views on the market, tech, and everything else

Business Lessons from Jim Clark (Silicon Graphics, Netscape, etc.)


Jim Clark is an entrepreneur and computer scientist. He founded several important businesses including the two named in the title and Healtheon (which eventually merged with WebMD). He is currently the founder of the building management systems provider CommandScape. Clark is a friend of my friend Craig McCaw and I have heard some colorful stories about him as a result. He is not boring. As an example of the Clark’s non-typical approach to life, Mark Andreessen said during an interview once: “I just ran into an entrepreneur who said, ‘I just ran into Jim Clark at a resort town in Italy. Jim was in a hot tub carved into the side of a mountain. I said, ‘Yes! That was Jim Clark.’”

  1. “Don’t be afraid to cannibalize your product. You must be willing to challenge your own product lines. For example, Barnes and Noble could have addressed the Internet, but didn’t until Amazon forced them to. That is the worst way to do it.”

Clayton Christensen argues that only the upper left quadrant in this matrix is genuinely “disruptive” in the way uses the term in The Innovator’s Dilemma. The best way to see which view is right is to put some real businesses in the quadrants:

Overlooked Low-end Customer Segment Most Demanding Customer Segment
Inferior to existing solution Netflix in 1997

Amazon in 1994

Superior to existing Solutions Uber/Lyft






To be provocative in order to to make this blog post more interesting I will argue here that there are other ways to be disruptive. You might want to think about where some other businesses fit best in this matrix. For example:

toy 3

In thinking about this issue it is useful to review what Christensen wrote in The Innovator’s Dilemma:

“An innovation that is disruptive allows a whole new population of consumers access to a product or service that was historically only accessible to consumers with a lot of money or a lot of skill.” “The very decision-making and resource allocation processes that are key to the success of established companies are the very processes that reject disruptive technologies. These are the reasons why great firms stumbled or failed when confronted with disruptive technology change.”

This classic example is often used to explain The Innovator’s Dilemma:

“Minicomputers were much smaller than mainframes, which had appeared in the 1950s, yet much larger than the personal desktop computers that followed them, beginning in the early 1980s. In the 1970s, minis ruled much of computation. But by the late 1980s, the business desktop microcomputer was eating DEC alive. “People attributed DEC’s demise to [CEO Ken] Olsen,” Christensen says. (Olsen had regarded desktop computers as toys for playing video games and publicly predicted they would fall flat in the business market.)”

What is Innovator’s Dilemma? Is it just an academic way of describing how businesses sometimes purposefully or accidentally end up in a competitive battle possessing different strengths and weaknesses? In my view a disruption based strategy is fundamentally about using asymmetry to create competitive advantage.  The key to a successful asymmetric attack in business is to pick a strategy where the competitor has a hard time responding to the attack directly from a position of strength. Guerilla warfare is a classic example of an asymmetric strategy.

  1. “Microsoft was founded the same year as SGI, and they both went public in 1986. I had the experience of my own foolhardy opinion of the PC in those days — that it was a ‘toy’ unworthy of the attention of real computer scientists.” 

Steven Sinofsky describes the crux of the issue that Clark is talking about in this way:

“As many have recognized, when inventions and innovations first appear they are often (always) labeled as ‘toys’ or ‘incapable’ of doing ‘real work’ or providing ‘real entertainment’. Of course, many new inventions don’t work out the way inventors had hoped, though quite frequently it is just a matter of timing and the coming together of a variety of circumstances. It can be said that being labeled a toy is necessary, but not sufficient, to become the next big thing.”

Sinofsky’s list of one time “toys” is instructional:


What “toys” would you insert in these years?

Toy 2

Chris Dixon elaborates on the “at first it is viewed as a toy” investing thesis used by many venture capitalists and founders:

“Disruptive technologies are dismissed as toys because when they are first launched they “undershoot” user needs. The first telephone could only carry voices a mile or two. The leading telco of the time, Western Union, passed on acquiring the phone because they didn’t see how it could possibly be useful to businesses and railroads – their primary customers. What they failed to anticipate was how rapidly telephone technology and infrastructure would improve (technology adoption is usually nonlinear due to so-called complementary network effects). The same was true of how mainframe companies viewed the PC (microcomputer), and how modern telecom companies viewed Skype. (Christensen has many more examples in his books).

This does not mean every product that looks like a toy will turn out to be the next big thing. To distinguish toys that are disruptive from toys that will remain just toys, you need to look at products as processes. Obviously, products get better inasmuch as the designer adds features, but this is a relatively weak force. Much more powerful are external forces: microchips getting cheaper, bandwidth becoming ubiquitous, mobile devices getting smarter, etc. For a product to be disruptive it needs to be designed to ride these changes up the utility curve.

…A product doesn’t have to be disruptive to be valuable. There are plenty of products that are useful from day one and continue being useful long term. These are what Christensen calls sustaining technologies. When startups build useful sustaining technologies, they are often quickly acquired or copied by incumbents. If your timing and execution is right, you can create a very successful business on the back of a sustaining technology.”

  1. “The decision to put money into the Internet in 1994 was considered by many of my colleagues to be borderline insane. Most people said things like, The Internet is free; you can’t make money on that! I literally had people telling me I was going to screw up the Internet by bringing more traffic to it.” 

Chris Dixon believes that the best ideas to invest in are good ideas that look like bad ideas (AKA “borderline insane’). If what investors like Clark and Dixon say is not true, the optionality associated with the investment is unlikely to be mispriced. In other words, if everyone thinks the idea for the business is a good idea from the beginning the entry price of the investment will be too high to be profitable in a way that is likely to generate a venture capital style financial return.

Which businesses today are good ideas that seem like bad ideas? Buying Dentacoin is a bad idea that seems like a bad idea. What other investments fall into the “what seems like bad idea is a bad idea” category? In what areas are people investing which are so flooded with money that an attractive financial return is unlikely to be realized?


The Y Combinator slide replicated above is: (1) not drawn to scale and (2) varies over time, domain of expertise and investor. Most bad ideas are bad ideas, but a few are not.

What is an example of a business that seemed like a bad idea to some investors that turned out to be a good idea? In November of 2013, Jamie Siminoff was a guest on the television show Shark Tank, He asked for an investment that valued his WiFi enabled video doorbell business at $7 million. Four sharks passed. Kevin O’Leary offered a loan and royalty deal that Siminoff declined to accept.  Five years later Amazon agreed to acquire the business for a reported $1 billion.

The founder tells the story of his Shark Tank rejection in this way:

“I will never forget leaving the set without a deal. It was horrible. I could not believe that we had done all of that work and were walking away with nothing. Sure I thought if we aired (the episode has a lower chance of airing without a deal) that we would get a little bit of traction, but I did not think it would be enough to make a real difference for us. I was gutted telling the team. But the show must go on and we went back to work, maintained focus and did what we could. And then we aired… and our lives changed forever. The bump we got from Shark Tank was not decent, it was extraordinary. And it wasn’t just something that lasted for the weekend. It’s still happening today, two years later. In terms of dollars, it was worth millions, but it also brought and provided an incredible amount of credibility and awareness for us with industry partners. Thanks to Shark Tank, we were able to hire more engineers and take the company to another level. It was like replacing the gas in our car with jet fuel…”

  1. “What I recognized after talking to Marc was that the Web was to networks in 1994 what the PC was to computing in 1982.” “Of course, I knew what the Internet was. But I hadn’t thought about what the implications were in terms of its growth rate.” 

Bob Metcalfe who co-invented Ethernet and founded 3Com said once in a meeting I was in: “No one ever made a nickel directly from Ethernet technology itself. Businesses have instead made a profit using Ethernet as a foundation for something else.” What Metcalfe was saying is that sometimes the profit is made indirectly from a technology or phenomenon. A classic example of a technology or approach producing profit indirectly is open source software. Many entrepreneurs have realized that a viable path to creating a new business is to create an open source software project and to then build a community around it. The business then tries to hire as many of the best people who worked on it as they can. The advantage that a business company like this brings to customers is how it adds value over and above the open source software — usually offering a combination of complementary services, tools and functions not necessarily available in the free version. Red Hat, Canonical, MySQL, WordPress and Mozilla have adopted this approach. What other businesses would you put in this category?

  1. “In the first year of business [at Netscape], we had almost no sales force. We were just taking orders.” “We recognized in the beginning that the Netscape required a different marketing strategy. The only way we could get large market penetration, was to allow to be freely downloaded, and besides the internet already has that culture in place – a lot of software was free. But we felt that by letting people download the software, we would be able to create a very large market share, and it worked. In a year and a half, we created 40 million users.” 

Bill Gurley describes the approach Clark is taking about: “If a disruptive competitor can offer a product or service similar to yours for ‘free’ and if they can make enough money to keep the lights on, then you likely have a problem.” The business model Gurley describes is commonly known as “freemium.”  A loss-leader product is not a new concept. Free Tapas have been served in bars in Spain since at least the middle ages. Gillette selling razors at a loss to sell profitable blades is a more modern example. In this model users typically get some value for free and are charged a fee for other complementary services. In some cases the service which has a monetary cost is more advanced in other cases it is less advanced. All businesses today must be prepared for competitors to give away what they sell as an incentive for customers to buy something else. What is different about a freemium strategy today is that many of the free services are digital and have close to zero marginal costs to create and distribute.

Since I am asking many questions in this post, which businesses in the news in 2018 best exemplify this freemium business model described by Clark. Atlassian and Dropbox are two examples. Spotify has a freemium business model. What other companies have adopted a freemium business model?

6. Silicon Graphics was] an excellent group of people but they – and people at every other company – begin to define themselves by what they have been doing, not by what they can do.” 

What businesses fit within Clark’s description in this quote? IBM? One way to avoid the trap Clark is describing is to have great “product people” who can create new products that drive the business forward. Fred Wilson their attributes here:

“The product person sets the overall requirements, specs them, focuses on the UI and UX and manages the process. The engineering person builds the product or manages the team that builds the product, or both.” Great products need both types of people.

Steve Jobs was obviously a product person:

“My passion has been to build an enduring company where people were motivated to make great products. The products, not the profits, were the motivation. Sculley flipped these priorities to where the goal was to make money. It’s a subtle difference, but it ends up meaning everything.”

A related problem is what Yoky Matsuoka has called “the Valley of Death” between academia and business. Business Insider describes how Matsuoka views the challenge:

“Research is all about proving an idea that’s never been done before. Have an idea, write a grant, hire research students, get proof-of-concepts and have everyone publish papers. Those papers bring in more grant money and lead to tenure. The gap comes at that point. Researchers assume that some great product person will take the research and turn it into a product to be used by millions of people. But it’s not easy to take a product “that works for 10 people and getting it working for a million or a billion people,” Matsuoka says. And the work required to bridge that gap “is boring for everyone,” she says. Researchers want to focus on new stuff that’s never been done before. They don’t want take something proven and published, and make it stable for a billion users. And product people don’t want to spin their wheels experimenting with early technologies that have only worked for 10 people. Their attitude is “We’re working on real products,” she describes.”

Truly great “product people” are a rare commodity. Who are the best product people you know who are actively involved in a business today? Do you have a product person on your team?

7. At Silicon Graphics I had advocated using cable-TV systems for all kinds of media distribution, for movies on demand and things like that. We did a contract for Time Warner in Orlando that used a computer that was equivalent to the set-top box. All that stuff was expensive—$5,000 per set.” “I was kind of a lone voice [at Silicon Graphics] I was babbling about cable television – and into the wind for a lot of the time. The reaction I got was, ‘Well we’re not a consumer electronics company. Why do we care about cable-TV boxes? Who cares?’” “I believe the Internet is the Information Highway. I’m religious about this. I don’t think it is cable television.” 

The Information Highway period in history teaches important lessons about how conventional wisdom can lead businesses, investors and government leaders astray. The metaphor described a plan to create a controlled environment with defined on ramps and off ramps. It was the wrong idea at the wrong time and was buried by the rise of the Internet.

The Internet’s distributed nature is the inverse of what was intended by the promoters of the Information Highway. The founders of the Internet adopted these four core principles:

  • Each distinct network would have to stand on its own and no internal changes could be required to any such network to connect it to the Internet.
  • Communications would be on a best effort basis. If a packet didn’t make it to the final destination, it would shortly be retransmitted from the source.
  • Black boxes would be used to connect the networks; these would later be called gateways and routers. There would be no information retained by the gateways about the individual flows of packets passing through them, thereby keeping them simple and avoiding complicated adaptation and recovery from various failure modes.
  • There would be no global control at the operations level.

The Internet’s explosive popularity took many people by surprise, including successful software businesses like Microsoft. Some people saw the warning signs earlier than others. Business Week reported at the time:

“The Web-izing of Microsoft begins in February, 1994, when Steven Sinofsky, Gates’s technical assistant, returned to his alma mater, Cornell University, on a recruiting trip. Snowed in at the Ithaca (N.Y.) airport, he headed back to the Cornell campus. That’s when he saw it: students dashing between classes, tapping into terminals, and getting their E-mail and course lists off the Net.

The Internet had spread like wildfire. It was no longer the network for the technically savvy — as it had been seven years earlier when Sinofsky was studying there — but a tool used by students and faculty to communicate with colleagues on campus and around the world. He dashed off a breathless E-mail message called “Cornell is WIRED!” to Gates and his technical staff.”

Bill Gates responded to the change by writing his famous “Internet Tidal Wave” memo, which today would be a rights protected document in the cloud rather than a memo.

There are many other examples of conventional wisdom that turned out to be wrong or a blind alley. Often the mistake is a result of a group of businesses trying to hype their future prospects. For example, telecommunication equipment suppliers have on several occasions been a source of problematic hype. I already mentioned the Information Highway failure but one should also attribute much of the Internet and telecom bubbles to the same source. Equipment suppliers and some operators like Worldcom/UUNet told tall tales about traffic growth which in no small part caused the telecom and Internet bubbles. These bubbles eventually popped as you know.

Is anything like the Information Highway or the telecom/Internet bubbles happening today? The press is repeating massive 5G forecasts for spending on infrastructure right now. For example:

Reuters (MARCH 2, 2018): “GSMA, which represents nearly 800 operators and some 300 suppliers, forecasts capital expenditure (capex) on mobile networks worldwide will be $500 billion over the three years between 2018 to 2020. Expanding 5G could mean capital expenditure rising to 16 to 17 percent of revenues generated by the mobile industry from 2020, up from 15 percent now.”

Is that $500 billion estimate real? What could be motivating that estimate? Could it be, that: “For network equipment makers, such as Ericsson and Nokia, which are struggling with declining sales for 4G gear, the rollout cannot come soon enough.” What could possibly go wrong? Have we seen anything like this before?

What is “5G” anyway? A recent report on the big event at the Mobile Word Congress repeated what was said about 5G at a panel of executives from big equipment suppliers:

“’It’s a new radio, meaning, a new format in which antennae will control electro-magnetic waves,’ said one panelist. Another person said it is a new ‘network architecture.’ Another finally concluded, ‘So we don’t have a definition of 5G.’ The problem is not just coining a succinct description: The technology is “the most hyped thing,” in one panelist’s words, ‘it is all things to all people.” And in that, it is something of a mess at the moment.'”

The Reuters article went on to say:

“’5G is, so far, too much hype, in the sense of its position as a new revolutionary technology,’ Telenor Chief Executive Sigve Brekke told Reuters at Mobile World Congress in Barcelona… CCS Insight analyst Ben Wood said one mobile handset company exhibited a showcase of 5G phones in Barcelona, only to have one display model drop on the floor and break open. ‘It turned out it was completely empty inside,’ he said.”

A lot of what 5G is today is a marketing slogan for many things. The discussion is about to get a bit technical for the last few paragraphs but that is unavoidable. One way to look at 5G is in terms of buckets of things. Just three of these buckets are:

  • One bucket of 5G is about better software and protocols. Internet of Things (IoT) applications and services may work better as a result of new and better 5G software and protocols for example. Lower latency may enable some new IoT applications. 5G standards enable things like reduce the number of functional components that data must traverse between the device and the servers, enable the deployment of a collection of services on virtualized hosts and implement better spectrum aggregation and sharing. If a wireless system does just the things in this bucket, is it 5G?
  • One bucket of 5G is about more fixed wireless to homes that may capture at most 5-10% of homes vs fiber/coax cable. ALso in the bucket is more backhaul to base stations using 5G radio frequencies like 28 GHz creating better 4G densification, but small cells are only economic to deploy in some areas.
  • One bucket of 5G is hype about handsets that receive signals at frequencies above 6 GHz. This claim about the use of so-called millimeter wave frequencies to serve handsets almost all slideware and press releases so far. There may be close to zero applications customers will pay more money for that would justify the higher handset and systems costs that would be required to receive millimeter wave frequencies in a hand held phone. A long time industry expert said this to me in an e-mail recently about millimeter wave frequency use at 5G in handsets:

“They’re putting in lots of antennas so by using beam forming it’s probable that you’ll have some usable 28 GHz signal in some situations. But the primary goal is to use the 28 GHz and above spectrum to serve devices when they’re not in your hand and not moving. Range will still be limited and propagation is challenging. This is the vision, but it does depend on base sites 300-500 feet away. The highest and best use  of 28 GHz is for fixed with high gain antennas, etc.”

Below 6 GHz frequencies in handsets work just fine, can deliver higher data rates every year anyway and at far lower cost. The real magic that delivers higher data rates over that last wireless link from a wireless base station is a technology called MIMO.

That’s enough about technology since this post is getting too long. To be consistent with the theme of this post, it seems appropriate to ask readers one final question. Can you think of a better anagram for Information Highway?

  • Hey, ignoramus — win profit? Ha!
  • A rough whimper of insanity
  • Oh, wormy infuriating phase
  • Hi-ho! Yow! I’m surfing Arpanet!






















Categories: Uncategorized