by Elad Gil
At Livongo, we took a consumer-centric view and created a service that allows them to “disengage” from their disease as much as possible. They want to be healthy. Livongo gives them a connected glucometer, monitors their data with machine learning, and intervenes if the consumer needs to make any adjustments in diet or exercise to remain healthy. This approach, in my estimation, will save us over $100 billion a year at scale in managing diabetes. Also, as we collect more data about consumers’ metabolic patterns, we are recognizing that there are many types of diabetes that we previously thought of in the “type 2” bucket. Highly impactful.
Taking a persona, like the type 2 consumer diabetic, and thinking about product mechanisms that are going to deliver experiences that consumers need, across industries, is an enormous opportunity.
Elad: Is that the common thread—that customer exploration and focus—that is causing these companies to break out? Are there specific tactics that these companies have in common where, you know, it doesn’t matter if you’re working in health care or financial services, you’re basically doing these three things?
Hemant: One thing these companies have in common is that these markets are so large. Arguably these are all markets that are $100 billion or more. And companies tackling them, the ones that get it right, have the potential to be $100 billion market-cap companies, something that used to be crazy to think about in venture capital ten years ago.
I think the commonality between all of them has been that they identified their target customer persona precisely at the beginning. So, for Stripe, it’s the developer trying to do online commerce and who needs an API for accepting payments online. For Gusto, it’s the modern HR person in a small business that is on social media and can be reached in new ways. For Livongo, it was a diabetic who always checks their blood sugar and is employed by a self-insured employer.
I think taking these massive markets and narrowing product focus down to a specific persona to start has been a core reason why these companies that have gotten it right.
Elad: How does that change as the company scales? You start dealing with product complexity. You start adding additional product lines. Maybe you want to broaden your scope because your first market was very high potential but a niche, which is often how great companies start. How do you think about that transition? Are there specific examples that come to mind of people who’ve navigated that well?
Hemant: That’s a great question.
If you look at how Stripe has scaled, providing the online payments API globally is a massive opportunity. While they have built an “AWS for commerce” stack for their customers to deepen their product stack, the traction of their core offering just dwarfs everything else.
And you want to take advantage of that. I think for them it was, “We want to make sure the majority of the new companies that start today start with us,” because that represents the future. By focusing on that, they have secured their future. Only when the organization became great at doing so did they start to focus on going upmarket to also serve companies that have been around for a while.
For Livongo, the playbook for scaling is different. Their objective is to keep consumers with diabetes healthy. Those consumers have many co-morbidities, like hypertension and obesity, and Livongo needed to deepen its offering much sooner in its lifecycle despite having a similarly large market to serve because it was the right thing to do for its customers to be truly impactful.
In general, it’s different for every company. But if a company has a large market to go after with your initial target persona, it should maximize scaling there before diversifying and diluting the organizational focus.
“We’re in the hot seat. We’re building modern education, modern health care, modern financial services.”
—Hemant Taneja
Elad: What are the common failure modes or patterns there? One could imagine either being too focused on the core markets and never broadening into something else. Or, to your point, getting distracted by new add-ons when you should really be doubling down because your market is so large and singular. How do you know which of those situations you’re in, or how do you navigate that situation?
Hemant: These are not winner-take-all markets. So first of all, the emphasis on “market leadership” becomes hard. The companies, instead, need to focus on charting a course with aggressive but manageable growth rates over a long period of time.
There is an optimal growth rate that is unique to each company. It depends on the physics of the operations: What is the pace of hiring that’s required to service their initial customer versus expanding to the next market segment? How complex are the logistics in delivering the services? What is the capital intensity required to diversify? Those questions start to become paramount in figuring out a growth plan that makes sense.
Good founders have done it. They’ve remained fiscally prudent. But a lot of companies have tried to do too much and diversify too much because lot of capital was available for cheap, but then they often had had difficulty raising money or growing into their valuations because they just couldn’t maintain focus on the unit economics.
Elad: That’s a really interesting point about winner-take-all markets. Because in the early 2000s, mid 2000s, a lot of the emphasis was on the fact that markets are winner-take-all and that investors should only invest in the winner-take-all companies. I think that’s because you were dealing with a lot of truly network effect–driven businesses. Do you think most venture-scaled businesses are winner-take-all? Or do you think you have more complex oligopoly or other market structures? In other words, how important is winner-take-all?
Hemant: First of all, I think it depends on the market. But for a lot of these large markets that have been regulated and are just opening up, I think it’s hard to think about being winner take all. Gusto is an example of that. There are millions of small businesses that they can serve with a modern people platform. Should Gusto focus its product and sales on companies that are 1 to 10 employees, or 10 to 100, or 100 to 1,000?
You have to carve out your niche and become really great at it, and not worry about what other companies are doing. Figure out how to build a business that is foundationally designed to grow at what you think the optimal growth rate is, and don’t worry about how you take the disproportionate share.
Elad: That’s super interesting. Because the canonical view would be: Google was winner-take-all, Facebook was largely winner-take-all. But to your point, if you’re dealing in health care or fintech, that same characteristic doesn’t seem to apply as much due to market size and structure, and fragmentation.
You’ve also been a big advocate for building responsibly, and I’d love to hear some of your thoughts on that notion.
Hemant: Tech companies spent decades focusing on creating software to deliver efficiencies across every sector, including health care, education, and finance. Now we’re rethinking how those services are delivered from first principles. New companies often begin around profound ideas: What does it mean to recommend to a diabetic how they should spend their day and what they should eat? Or what does it mean for teaching behavior in a classroom full of high school kids?
Those are big responsibilities. The traditional mindset has been to grow at any cost, right? Silicon Valley has been obsessed with this idea of backing the hacker-entrepreneur who moves fast, breaks things, and iterates. But that mindset and growth at all costs is not well suited to building responsibly in areas that greatly impact people’s daily lives.
So for me, there’s this notion of the empathetic entrepreneur who understands their customers’ needs deeply, figures out what’s the responsible way to serve them, and makes sure that is represented in their MVP, even if it comes at the expense of some growth.
Traditional large corporations that serve these markets have this idea of corporate social responsibility. I think startups building in Silicon Valley today need to have their own startup social responsibility. In my opinion, that is met b
y being transparent about how algorithms and machine learning are being used to deliver services, and by building measurement systems to make sure that the company’s success is not based on taking advantage of things like bias and discrimination that we want to avoid in society. I’ve generally looked for founders who truly understand this, respect this, and are building with those principles from the beginning.
Elad: As a company scales, how do you think it can reinforce those principles through the culture or the day-to-day work? What have you seen work well tactically?
Hemant: It really is about starting to build measurement systems around the impact that your products are having in society, in whichever dimension they’re meant for—be it social emotional learning, chronic care management, fiscal literacy, media, or what have you. And being transparent.
Founders need to be prepared for the yet-unknown. It’s not always going to be immediately obvious when you’re deviating from your company’s core values. Teams need to be thinking about building what I call “algorithmic canaries” into their products and metrics dashboards in order to catch unintended, negative outcomes before they become broadly adopted.
If algorithmically-driven sales optimizations lead you to excluding groups of people from even learning about your product, or managers in your company tend to review older workers using different language and criteria than with younger employees, algorithmic canaries should throw a red flag. The same is true for understanding how your product is being used: canaries should alert you to your product being used in ways you never intended it to be.
You can look at how some companies just abused practices to grow. Think about what happened with Zenefits. Think about what happened with Theranos. Think about what happened with Uber. They all took shortcuts to grow fast and put their employees, customers, and other stakeholders at risk.
I think that you can have a team that values “responsible innovation” from the beginning. You can state that weaponization of technology is just not something you’re going to use in your product development. Or else you can end up being one of these companies that eventually become kind of fatally flawed.
I am convinced that we are a decade into a thirty-year cycle. You know, we talk a lot about high valuations, or when’s this bubble going to burst. I would pay less attention to that and think much more about the broader digitization that we’re in the midst of, one that Silicon Valley is leading. I care deeply about embedding this thinking in our culture.
We need to think hard about how our collective work is coming together. For example, right here in the Bay Area, we have companies building self-driving trucks. We have companies working on longevity, including Spring Discovery. And we’ve got experiments going on in basic income. What if it all comes true? Some day we are going to have to talk to three million truck drivers in the country: “We’ve got good news and bad news. Good news is you’re going to live thirty, forty, fifty years longer. Bad news is that you won’t have a job. And oh, by the way, we’re going to put you on a stipend and eliminate all your pride and self-esteem.” That sounds terrifying.
Just thinking systemically about what we’re creating as the technorati is also important. That needs to become more of the conversation in our tech ethos.
“We need to think hard about how our collective work is coming together.”
—Hemant Taneja
Elad: To build on that, when do you think founders of breakout companies should get involved with either philanthropy or politics or thinking about broader societal issues? One argument would be that you should be solely focused on your startup. You should be focused on scaling it until you have the free time and money to help society more broadly—the Bill Gates model of philanthropy. The other argument is that as soon as you are seeing success you should really be engaging with society more broadly—immediately. I’m curious how you think people should handle that transition and when they should do it.
Hemant: I’m a believer that getting involved in philanthropy from early on is a good thing. I actually cofounded this organization in Boston called TUGG (Technology Underwriting Greater Good). The whole idea was to have the tech community help mentor social entrepreneurs.
In Silicon Valley culture, there’s a lot of mentorship that goes around in helping each other build startups. I think that even if you can’t afford to help others financially, whatever time you’re spending mentoring other entrepreneurs contributes to the pay-it-forward culture that we have. Apply a reasonable portion of that time toward helping social entrepreneurs. Then over time, as you’re successful because you have been deeply obsessed with your startup and things worked out the right way, start to get into philanthropy financially as well.
That’s the continuum that I think about. So start with the “mentoring” time that you have today. And as you achieve financial success over time, contribute financially as well.
“I think startups building in Silicon Valley today need to have their own startup social responsibility.”
—Hemant Taneja
Elad: And how do you think about political activism and things like that in the context of a startup? Because on the one side, there’s all sorts of examples where you want to create a balanced view within your company, because your company or employee base really represents all different types of people. But at the same time, Silicon Valley has been notoriously terrible at engaging in the political process in a way that is positive for society. Philanthropy is one part of impacting society. Politics and regulation is another. How should companies or founders think about that?
Hemant: I like the role of business to be in helping design the right policies, and stay away from politics. I’ve done a lot of work in the energy sector for the past decade, and it’s always been about how to help create policies that lead to the adoption of clean, affordable, and secure energy.
Our bias should be to have more transparency in how we use software and data in our products and services. Engaging in a dialogue with policy makers, and demonstrating self-regulation via algorithmic accountability, is a way to productively engage with policy makers. Missteps by the companies in the social media sector around this make us take a step back in this regard. We need to learn from those experiences, because otherwise we run the risk of more regulation.
I think that it is a very worthwhile investment in time to bring regulators along with our use of data and AI. Because less regulation implies more innovation. But I think that only comes from investing in a trusting relationship with the regulators.
This interview has been edited and condensed for clarity.
APPENDIX
Things to just say no to
If you are in Silicon Valley long enough, you feel like you have seen it all. Here are some things to “just say no” to.
Envelopes full of cash. Google used to give out everyone’s Christmas bonus as a thousand dollars in cash in a plain envelope on the same day in December. Urban legend has it as Googlers would disembark off the shuttle to San Francisco, they would get collectively mugged of tens of thousands of dollars by thieves who heard about the company’s generosity.
China. Uber had come the closest to succeeding in China with its 20% ownership of Didi, but basically every other tech giant has been blocked and then cloned in China, with early operations shutting down. For most companies, a China strategy ends up being a painful, money-burning fail.
Giant chrome pandas. Dropbox bought a giant chrome panda at the height of its funding success. As time went on and the company focused on frugality, the giant chrome panda became a sign of its earlier wanton spending, and became a constant reminder that you should save money. You can instill that lesson without wasting a ton of money on a chrome sculpture—for example, one company I heard of recently put a Juicero machine in its office as its own version of a “chrome panda.”
Pool tables. When I first moved out to Silicon Valley, I joined a Sequoia-backed startup that had 120 people. Within three months it grew to 150, and t
hen shrank over the next 9 months down to 12 people over four to five rounds of layoffs. After the first round of layoffs, the company bought a pool table to “help employee morale.” Unfortunately, anyone spotted playing pool tended to be laid off in the following round. Eventually shooting pool was a sign that the people playing would soon be gone. The pool players probably had the free time to play pool, so probably that was correlation rather than causation. Still: Not a good sign.
