Howard Morgan got node 50 on ARPAnet in 1972. Forty-six years later, he watched MemSQL, one of the companies he invested in, demo a database that processed a trillion records per second. Few technologists can claim either of those experiences. None can claim both. Where it gets astonishing is the roll call of tech and founders he’s guided over 50 years. The more one gets to know Morgan, the more he morphs into tech’s Most Interesting Man in the World.
The professor, former CEO, longtime investor, and co-founder of First Round Capital has been at the forefront of startups since before they got that name. After years teaching at Cornell, CalTech, and the University of Pennsylvania, he became a founder himself in the ‘80s with Interactive Systems, the first commercial Unix licensee. Around that same time he began investing, and he’s been a fixture (and holiday video star) in the venture world since. He’s a founding investor at Idealab and launched First Round with Josh Kopelman in 2004.
Morgan’s had a hand in many noteworthy companies along the way, and learned lessons from each of them. There was GoTo, which became Overture and was the first paid search company at Idealab; CarsDirect, which became Internet Brands; and Savage Beast, which became Pandora. He saw the dangers of the dot-com bubble’s inflated valuations with eToys, which eventually went under. And more recently, he’s seen how the rapidly changing availability of high-speed storage has made companies like MemSQL possible.
Simply put, he’s an encyclopedia of tech history and startup lessons. And all that he hasn’t experienced firsthand, he soaks up, too. (He keeps a running list of all the books he’s read over the last 30 years; it’s over 5,700 entries long.) In this exclusive interview, we dig into the practice Morgan has seen differentiate good founders from great founders: the ability to relax constraints. Here, he explains how leaders should look into the future—and how casting off limitations can be your smartest strategy.
If you can’t think out of the box, make the box bigger. Start by relaxing constraints.
Relaxing constraints isn’t a one-and-done proposition, or a task you can check off your to-do list. It’s a habit, a regular practice of analyzing what’s limiting your vision. Startups evolve quickly. “To the extent that you can think about what might constrain you as you grow and understand where you might want to get to, you can do things that will make that growth easier,” says Morgan.
Take a simple one. “The one constraint that none of us human beings can fix is time. We only have so much of it. The question then is, ‘How do you use it effectively?’" Because while you can’t change time, you can relax how much that constraint impacts you by consciously making space for your most important work. As an investor, Morgan urges his partners and founders to put an hour of “me time” on their calendars each day for focused thought work. “You can very easily let the outside world constrain you by filling your entire day with meetings,” he says.
Technology itself can become a constraint, too, if you let it. It’s a rapidly changing landscape, and if you get stuck in today’s trends and capabilities, you’ll feel ahead but can quickly be left behind. “As a professor, I would advise my graduate students to think about what they felt might be exciting five to 10 years from now. That was the subject they wanted to work on. In today's world it might be where CRISPR is going to go, where gene editing is going to go, where the blockchain is going to go.” If it took three or four years to complete their PhD, those graduates would then be well positioned to tackle the next big thing.
If you relax the constraint in your mind, you can build for it — and better circumvent it later.
Wherever you are in your startup’s life cycle, there are four commons types of constraints that Morgan urges founders to revisit. Relax these constraints, and you’ll be best able to expand your thinking, maximize your opportunity and grow your product.
“When we were building internet companies at Idealab in '96 through 2000, the bandwidth that was available to most people was 56k dial-up. But to build interesting user interfaces, you had to assume broadband,” says Morgan. “When Tim Berners-Lee was building the Internet at CERN in Geneva, they were building it for a one and a half megabit connections — the T1 connections that the top universities and research institutions had. They were not building it for the dial-up line. When Marc Andreessen and the people at NCSA in Illinois started building what became the Netscape browser, they were very aware of the constraints. They said, ‘Look, we know if we relax the constraint we can get to T1 and much faster broadband speeds. We want to be able to build something that will really work well at those much higher speeds, but can give us reasonable performance at the slower speeds."
To build something that will change how people live their lives, you need to think beyond what’s available today. A similar pattern played out in the mobile space. “Mobile bandwidth on 2G was not enough to do interesting data,” says Morgan. So successful developers imagined away that constraint. “They said, ‘What if that goes away? We can build GPS. We can build rich applications that need hundreds of megabits per second. When that bandwidth is finally here, we'll be ready if we relax the constraints in our minds.’"
Questions to ask to relax technical constraints. Achieving those big technical leaps forward may be arduous, but identifying the constraints you need to think past is more straightforward. Morgan recommends that founders ask a couple of simple questions:
What technology is missing that is stopping you from achieving your end goal?
What can you do now that lets you fold your more ambitious aims in later?
A caveat to consider. One challenge with relaxing this constraint is spanning the gap between building for the long-term but being able to sell something for the short-term. “Google realized this and started with a very simple interface which could get great speed. As bandwidth grew, they could put more on the screen faster,” says Morgan. “Building something that scales from the very constrained to the relaxed constraint is one way to stay in business while waiting for that constraint to go away. When those limitations disappear, you’re positioned to swiftly and dramatically improve your product or service.”
Looking past cost constraints is crucial for two reasons: first, it will ensure that you’re developing the best product, not the most feasible. And, second, it will prevent you from investing time in a product that plummeting costs will make obsolete as soon as it launches.
“Back in the early ‘80s, we had a company called Proximity Technology, which did fuzzy search—it allowed you to search names even if you misspelled them, looked for sound-alike terms, things like that,” says Morgan. “We built a custom chip to do that fuzzy search, PF 474. I was talking about it one day with Bill Gates, who said, ‘You know, Howard, that's great, except that two years from now Intel's next chipset’—which he knew about—‘is going to be fast enough to do all that in software. The constraint that you see, that people have to buy this expensive chip to do those algorithms, is going to go away."
It’s a story that’s played out again and again. “The cost constraint for Infonautics was literally just buying servers and licensing software. We paid hundreds of thousands of dollars—millions, in fact—for Oracle licenses. All of a sudden three years later when Half.com was starting, there was MySQL, which was free,” says Morgan. “Developing AI technology was cost prohibitive for all but the biggest companies, until TensorFlow and parts of Watson were open-sourced. You don't have to license or buy expensive software or try to build that software.”
Storage, bandwidth, processing power, speed. In recent years, the costs for all these essential building blocks for software have plummeted. Now, founders need to think about the next barrier that’s going to fall. “For example, 5G technology, which should be ubiquitous worldwide by 2024, means 24/7 connectivity anywhere can be assumed,”says Morgan. “Take a second to picture that new reality. What can you build knowing that?”
Questions to ask to relax cost constraints
As your business grows by 2X or 10X, what component is the biggest cost growth?
What grows, non-linearly, as you're scaling your business? What if you could get rid of that constraint?
A caveat to consider. Morgan and the Proximity Technology team could have saved a ton of pain if they had relaxed the cost constraint before building. Luckily, Bill Gates shortened that wrong turn, but not every founder will be so fortunate. Morgan has a few pieces of advice for startups to hedge as they relax cost constraints. “You can put up wallpaper to cover it up, but the writing will still be on the wall. You may not be able to pivot immediately when a cost constraint disappears in your industry, but start clearing a new path forward,” says Morgan. “At Proximity, we sold boards for a few years with the chip, but then worked build a software version which led to the Franklin Spelling Ace handheld dictionary. That product generated $100 million in the late 80s and was cheap because it didn’t need the expensive custom chip.”
When it comes to developing solutions to previously unsolvable problems, founders will also come up against the limitation of what is currently known, or knowable.
One startup that Morgan invested in recently is Openwater, which aims to change how we read and write to our bodies — in the form of a machine that can scan the body at a billion times the resolution of an MRI. But that’s not all. “The question they’re looking at is how do you do computer-to-brain communication, not just brain-to-computer communication?” says Morgan. “That’s science fiction.”
When you reach for abstract, futuristic and grand ideas, you’ll find that there are no pre-trained professionals you can simply hire for your outside-of-the-box ideas. And that’s another type of knowledge constraint. “It’s difficult to hire experts in a technology that doesn’t exist yet,” says Morgan. “If you want engineers who can do computer-to-brain technology, there aren't people trained in that.”
When there is no perfect hire, just start with smart people who have expertise in one piece of the puzzle. Then get them up to speed on the rest. “At Idealab, we have something called Energy Vault where we're doing energy storage at grid scale. We had to find civil engineers and other people who could build the concept we had even though they had no knowledge of the energy space,” says Morgan. “We started hiring and training as opposed to waiting to find the people who knew it, because there wasn't anyone who knew it. You had to grow them.”
Questions to ask to relax knowledge constraints
What’s the science fiction version of your product?
Have you pondered any particularly intriguing “what ifs?” as you developed your product?
When hiring teams to build groundbreaking technology, ask: What's the closest analog to what I want to do? Where can I find those people to relax that constraint?
A caveat to consider. “When relaxing knowledge constraints, it’s still important to observe the laws of physics. We’re not likely to change those, so antigravity or perpetual motion machines should not be part of your plan,” says Morgan. “You should also understand that what may be known in 50 years is not that relevant to today’s startup. Thinking three to 10 years out is about as far as one should go for almost any founder building a company.”
In the old days, the time it took to get a startup off the ground came down to how long it took to actually build the product. Now, unless you’re in the hardware space, that time line has been slashed. Cheap servers and high bandwidth have made it easier than ever to launch a company. “What we see at First Round is in that first 12 months after we've made the investment, if the company hasn't gotten their product out it's usually a problem of some sort,” says Morgan.
The goal post for what a meaningful customer base looks like has moved, too. “It used to be the measure was how quickly could you get a million users. Then it became 10 million users. Now it's a hundred million users,” says Morgan. “When something really blows up on the Internet, we’ve seen companies get a hundred million users in three months.”
In Morgan’s experience, what holds people back from amassing users comes back to hiring. “That’s the thing that blocks most companies from getting the customer base quickly. They need to hire more people, whether it's in sales or in engineering or whatever,” says Morgan. “But as companies grow and hire, they encounter a new host of challenges: more processes need to be put in place, and structuring and managing teams begins to consume a larger portion of leaders’ time and headspace. That slows you down a little bit, but appropriately so. But you have to keep pushing, particularly on customer growth. The whole growth hacking world, of course, is built around that.”
Questions to ask to relax time constraints
What’s holding your startup back from quickly finding and growing your customer base?
What’s in the way of scaling your hiring so you can get more minds working on the challenge your company is addressing?
Today, the time constraint is much more the time to get a customer base than it is the time to build a product.
Founders need to develop this muscle—this habit of assessing and relaxing constraints—at every stage of their companies’ growth. But Morgan shares a few pivotal company-building points, in particular, where he’s seen it make all the difference.
At ideation, don’t limit yourself to the current reality. Morgan has seen both seasoned and green founders constrain their big idea from the get-go. So, when you’re getting started, don’t be shy or afraid to engage in truly futuristic thinking.
Take perhaps the greatest example of audacious thinking, Google. Web search was already being done, but it was a very different prospect when the web was 10 million pages than it was going to be as that number exploded. “Google figured out a way to both crawl and then organize and get rid of the constraint that you couldn't search an entire web as it got to be a billion pages, or 10 billion pages,” says Morgan. “Both the page rank algorithm and the parallel distributed systems architecture that drove Google came out of thinking about that constraint at the very beginning.”
Early on, the constraint for MemSQL was high speed. “Its answer was, ‘Let’s assume everything fits in high-speed memory.’ Even though you couldn't buy that much high-speed memory at that time, they said, ‘If we assume you can, or close to, then what do we do? How can we solve a problem and make something that’s 10 to 30 times faster than what’s out there?’” They followed that line of thinking, and their database now processes a trillion records per second.
As your company scales, don’t become the hiring bottleneck. Most founders will come up against a constraint in the personnel space, and that’s the desire to interview every prospective hire themselves. “When you realize that your time is too constrained for you to actually interview every single candidate—which may come at 20 people, it may come at 100 people, but not much more than that—you have to think about how you overcome that.”
Typically, the way to do that is by implementing processes and systems to ensure that you can step back from those personal interviews without sacrificing candidate quality. “You have to create methodologies that let you keep the kind of people you want to bring into your organization, but do it without having to physically spend an hour with each one,” Morgan says.
With customer acquisition, don’t rely on a single channel (no matter how effective). Time and time again, Morgan talks to startups, particularly in the consumer space, that have one straightforward acquisition strategy: buying Facebook cohorts. “So I say to them, ‘Okay, well that's great and you're getting a great pack of users from that particular cohort,” he says. “But how many of those people are there on Facebook? There's only three million. If you get all of them, you've still only got three million users, and you want to get to 20 million or 50 million.’"
Founders need to think about where their users will come from early, to avoid constraining themselves by putting all of their eggs in one acquisition basket. Consider as many channels as you can. “The earlier you think about how you’re going to grow the acquisition process, the faster you'll be able to grow.”
To achieve lasting success, reimagine what your company might looks like in the long-term. As they strive for the next level of growth, founders may need to relax the constraint of pursuing only one possible outcome for their startups. Most founders dream of taking a startup from idea to IPO — and that’s a good thing. But relaxing constraints about how a company exists or exits is a healthy thought exercise to do annually — if not more frequent.
In many cases, this examination means relinquishing the notion of being an independent company at all, at least as they’ve currently conceived of it. “If you're Instagram and you have 20 million users and you all of a sudden want to be available to a billion users on Facebook, acquisition gets you there.” As part of a larger company, startups suddenly get resources and distribution that would otherwise take years to acquire. “It jumps you forward in time, and that may in fact be preferable to slogging it out alone.”
There’s one last constraint that comes up again and again, and it’s perhaps the hardest to see: the limitations of founders themselves. Starting a company is a Herculean feat, and often a very personal one. Founders take on the role of not just the face of the company, but its heart, soul, and muscle, too. But if a startup is going to endure, that might not always remain the case.
Founders have to relax the constraint that they're controlling and doing everything.
Think of the tech giants of today. At a certain point, the founders of those great success stories realized they couldn’t do it all and brought in expert help. Mark Zuckerberg knew there were things a CEO needs to manage that he wasn’t good at, so he hired Sheryl Sandberg as Facebook’s COO. Sergey Brin and Larry Page wanted to focus on developing game-changing technology, not company building, and brought in Eric Schmidt to serve as Google’s CEO.
But among the hard-driving, high-achieving founders of the world, this kind of self-awareness is too rarely the case. “One of the problems when you have very smart founders is that they believe that they can do everything well if they can just learn it. Except they don't have the time to learn it and you're training them to sort of learn by doing,” says Morgan. “Whereas experience can really shorten the time to get a job done or to get to scale.”
Successful business leaders take their ego out of the equation to best serve the work they want to get done. In his own career, Morgan lived that lesson when he retired from First Round last year and took on the role of special advisor. It was all part of the generational succession that he knows is so important in venture capital.
“Doing the job right demands a heavy focus on deal sourcing and getting out to meet founders. Alan Patricof, who's 10 years older than I am, and I would see each other at 8:30 p.m. at a meet-up at a bar to talk to founders and say, ‘We're both too old for this.’ Eventually, I determined that was too much of a constraint on First Round. I don't mean to be ageist, but somebody 25 to 35 is much more likely to deliver the necessary high-intensity interactions at those meetings. I like to live in multiple generations at once, but I understand that my interaction with a 25-year-old is going to be different than that of a similarly-aged peer of theirs. I had to relax a constraint, which was figuring out how to do generational transition.”
Whether it’s a business model or a business leader, the exercise is the same: what happens when you see each opportunity and constraint as a variable — and remove it? If it’s a company, that might be technical, cost, knowledge or time constraints. Founders must also must not hold themselves as sacred, if they truly want to build something enduring. Because many times, delegating or stepping aside from something is actually the best way to ensure that your vision goes as far as possible.That’s what developing the practice of relaxing constraints is all about.
“Part of this skill is to explicitly think about constraints once every so often, whether that's once a week or once a month. Really explicitly take some time, 10 minutes to an hour, every month and say, ‘What am I doing that I feel constrained by? What could I relax that could make things much faster, better and more scalable?" says Morgan. “Look at it another way: we all write words — whether it’s a book, an email or a text. But don’t just flex your writing muscle. Tap into your ability to edit. How does a sentence — and a person, a company, an industry — change when you take away words? If no word is sacred, you start to serve the message rather than the author. Released from those constraints, the enduring idea emerges.”
Photography by Michael George.