Back to Blog

Big Companies Act More Like Startups So People Don't Leave: Kleiner Perkins' Mike Abbott

By Christopher Steiner  •  Apr 13, 2017

The preferred landing spots amongst strivers and high achievers in college were once mostly old blue chip companies in predictable spaces: investment banking, consulting, law. But startups have become part of the zeitgeist with young people. Many fresh graduates would pick a job at a quickly-rising 10-person tech company before they went to Goldman Sachs.

This is a reality that has leaked into the software industry, notes Mike Abbott, a general partner at Kleiner Perkins Caulfield & Byers and a 2014 investor in Snapchat. The effect has become sharp enough that big company software teams have changed how they release code and products to more closely mimic the environment of a startup.

Companies now focus on microservices and containers not only because they make sense, but also because they enable tighter iterations and releases—two things that are important to young software engineers, and therefore important to recruiting.

Young engineers who are used to hacking on their own projects or working at startups expect to ship code and to see their work out in the real world at a fairly regular cadence. It provides satisfaction and an immediate sense of accomplishment for engineers to be able ship that way—and if they're not getting that, their eyes could quickly wander toward other opportunities.

"If you're a new grad and you join a company and you're not shipping anything, it's unlikely you're going to stay very long," Abbott says.

Engineers are the most valued resource in Silicon Valley and other tech hubs, and getting them on board and keeping them on board remains top of mind within the industry—even if it means changing how products behave or how they're released.

This is one insight that Mike Abbott shared with FundersClub CEO Alex Mittal during a FundersClub Live event. Abbott has a resume that includes starting companies, selling companies and now investing in budding startups aimed at the enterprise and a few at consumers (Snap, of course).

Before getting into the startup world, Abbott was actually pursuing a PhD in Molecular Biology at the University of Washington. A couple of years later, he raised $70 million at the height of the dot-com boom.

Here are some of the other key takeaways from Abbott’s talk:

  • Entrepreneurs can be shaped by failure as much as they’re shaped by success. That's not necessarily a bad thing. Learning from mistakes that are actually made is one of the most forceful ways in which to learn anything. Any entrepreneur who is made a particular mistake in the past is very unlikely to make it again. Abbott experienced incredible highs and lows as being part of a startup during the dot-com boom, and the lessons he took away from that time have shaped him to this day.
  • Very few people, certainly not Abbott, aspire to be in venture capital when they are young. One of the biggest things that made him look at the industry as a possible career at some point is the knowledge that there aren't enough VCs who understand exactly what it is that founders go through in building a company. The best VCs understand the struggle, the work that it takes to get a company off the ground. It's not glamorous to be a founder—and that's an understanding that Abbott brings to the table.
  • As a VC examining companies building for the enterprise, Abbott looks for startups who are providing a painkiller for big companies, not just a vitamin. These are products that solve sharp needs inside the enterprise. They're not just supplemental, they're integral. One way to see if a startup is indeed providing a painkiller type of product is if big companies are willing to pay for a pilot. Paying anything above zero is relevant. If a significant number of enterprises are willing to write a check for the pilot, then it's very likely the startup is onto something.
  • More than anything else, venture investing is about people. There are trends in technology and commerce that Abbott believes in, to be sure, but he ultimately sees investments as bets on people. As for areas he likes, Abbott thinks that machine learning and AI will play large roles in customer service within the enterprise at some point, and he's keenly interested in people who can execute on that vision.
  • One of the biggest challenges facing founders who are piloting growing startups is maintaining communications throughout the company. When employees complain that a company no longer feels like a,startup, that it feels like a big company—that gripe can often be traced to a lack of communication throughout the startup. When a startup is just five people, communication occurs naturally. Just about the entire company interacts with each other every day. When a company grows, however, if certain instruments aren’t put in place to ensure communication occurs on a regular basis, the company will naturally feel bigger than it is. By keeping information and communication flowing throughout a growing startup, it will help it feel smaller longer.

Below, readers will find a version of the conversation between Abbott and Mittal. While this conversation has been edited for clarity and, in some cases, brevity, it should be noted that this was a live exchange with questions from an international audience arriving in real-time. Many of the questions come directly from viewers. Mittal’s words in bold:

I’d like to note something interesting off the bat: you received your undergrad degree in biochemistry. How did you go from that to starting Composite software?

Yeah, totally the normal path of things, right?

So, framing my background a little bit here - my family goes back four generations in the Bay Area, my grandfather taught engineering at Cal, my dad's an engineer, so I grew up around computers. My introduction to writing software was on the Apple II Plus and even in my elementary school, there were a lot of computers and so I was used to all of that, it was normal to me. Clearly, that's not totally normal. In high school, though, I became very interested in microbiology and some of the advances around just genetics. That's what led me to consider the opportunity to see something novel in science, and when I was doing my undergrad I kept coming back to computing, which is why I went up to the University of Washington to look at grad school opportunities around the intersection of computing and biochemistry.

After about two years I wasn't sure I wanted to be a professor and a friend of mine got a grant to start a company, and he talked me into taking a leave of absence. That was my foray into entrepreneurship, though we didn't raise any venture capital, but just writing grants, which for us, frankly, didn’t turn out great.

Got it. My understanding is your following company, Composite Software, was actually quite successful and in the end was acquired by Cisco for $200 million. How did you accumulate the wherewithal to do that?

With the company before Composite, I experienced an incredible disaster, which I think informed both how Composite was built and, frankly, is one of the things that contributes to how I work with founders now. I was involved with a company started off with $16 million, because one of the co-founders had taken a company public, a company called US Web. Then in March of 2000 we raised another $70 million—this is at the peak-peak.

Then a year later I found myself having to layoff 180 people out of 220. I was the last executive and I was able to sell the company in the summer of 2001 because we still had $15 million in the bank.

As a 20-something, it was an incredibly humbling and hard experience.

So all of this was a guide to me as we did the Series A for Composite in the summer of 2002. My introduction to this world was actually sort of a disaster. But I think it's important to realize from the beginning that startups are hard. As many of you know, you learn a lot when something doesn't work.

What caused you then to transition back into larger technology companies? I feel like more often than not, people who actually leave a Google, Facebook, Microsoft or Apple, tend to then go into entrepreneurship. I think your career trajectory kind of went back into the larger companies, though.

I tend to do things backwards. Everyone goes this direction, I go the other direction. It seemed like things were migrating toward the cloud and I felt like there were probably two places I could do that myself, Microsoft or Google. I actually went up to Microsoft initially as an individual contributor. Eventually I was asked to become the first general manager of the effort that led to the building of Azure, but I very quickly discovered that the company, at that time, didn’t seem to have real conviction behind the product. I was then recruited to go lead the team to design and build WebOS on the mobile side. I had never done anything in mobile or consumer, that was totally brand new for me.

To continue that arc from there to venture, what was the ah-ha moment for you to say, "hey, actually this is where I want to be"?

Yes, when I was 18, I didn't even know what Venture Capital was. One of the contributors for me to go into venture, candidly, was seeing all of the bad behavior by investors when I was an entrepreneur. I did have some good investors and advisors, however, and at some point it occurred to me that needs to be more people on that side of the table that actually can empathize with what founders go through. As many of you probably are dealing with now, being a founder is a very lonely and hard thing to do. It's not glamorous to go be a founder and CEO. That's the spirit with which I try to engage with companies I work.

Just flash forwarding to where you're at today, Kleiner, are there any particular areas that you've chosen to focus on in terms of companies that you're looking, seeking out or already working with?

I spend most of my time on the enterprise by nature of my background, especially on the engineering side, so I spend probably three-quarters of my time meeting great people in that space and investing in them. I am not the type of person that goes to sit in some dark room to come up with six theses that I'm then going to try and find and invest in. There are trends I believe in, for sure, and places I think the world is going to, but ultimately, so much of venture investing, to me, is about bets on people. Will this person or team be able to filter the right feedback from customers?  Will they put the right things into the product or ignore their customers’ needs? There's areas that I'm interested in, for sure, like machine learning and AI, and some of the advances in neural nets, but it all starts with a person. I can't project a vision onto someone. I would not want to ever do that.

At what stage are you looking to get involved with an entrepreneur? You develop early relationships when they're still at the idea stage, when they're already out there with a working product or service, or some other point and time?

Probably the spectrum. I mean there's certainly a number of folks that I’ve had the chance to work with who may be too early for us because we don't tend to really do a lot of seed investing. I might introduce them to other seed investors. There's a number of companies right now that fall into that category.

As for what stage I try to catch companies at, if they’re on the enterprise, the B2B side: Maybe they have a 1.0 product, maybe there's some pilots going, and it's still very early. Maybe they even have some revenue, but more importantly I feel like they're providing a painkiller versus a vitamin, something that real businesses need.

I think some in our audience might not immediately know what you mean by painkiller versus vitamin, but I think in venture and in startups, it's a very salient point. Could you explain for us?

Sure. In most companies of any size there's a procurement team that for different amounts of money, have different approvals. The goal is to reach those teams and try to line up pilots for your concepts to be tried out, where people might not pay because they're going to try your product to see if it works or not. If the company perceives that you're solving potentially a real pain, a must-have versus a nice-to-have, I've found that people are willing to pay something for that pilot. As I tell entrepreneurs, it's actually not that important about how much they're paying you, but that they pay you something greater than zero. Not only will that validate your product and idea, but you'll also get, frankly, better product feedback because you’ve actually gone through the pain of dealing with their procurement team. They'll more likely have a vested interest in helping you succeed.

Maybe we could use that as a segue into almost a completely different spectrum of companies. Going to the consumer side, which is very different, but I understand that you led Kleiner's Investment in Snapchat, for example, and they of course had a phenomenal story.

Unfortunately the investment was at a later stage than I would have liked, but Evan is phenomenal as a product person.

So you're not actually looking at how much are individual consumers paying Snapchat, obviously, it’s about something else. What was the thesis or idea around backing Evan and the team at Snapchat?

I tend to, on the consumer investments that I've done, which is a handful, like Mobcrush and Plex, they tend to have some set of metrics or quantitative data where you can see some sort of very rapid organic growth, depending on the product. There is data there that substantiates the idea. But it always comes back to the founder, and it's been amazing to me to watch Evan pilot Snap and decide how his product will evolve. He’s made great, non-intuitive decisions to introduce new features and products that will eat away and eventually kill some older features and products. It's a real testament to him.

Cannibalize yourself.

Exactly, and it's easy for us to talk it, but to do it is, it takes a lot of intestinal fortitude. I really admire that.

This is an audience question from an angel investor: As an angel who writes a smaller check but still really wants to support the company in some way, what is the most important thing that an Angel investor can do to improve their return? So thinking about it as a financial undertaking for somebody who wants to support early entrepreneurship, what's the one thing that they might be able to do?

Yeah, I will try my best to answer. I'm just reflecting on when I was an angel investor and invested in companies like Hearsay Labs or Kildare and others, it still comes back to the people. I happen to know one of the founders, like in the case of Hearsay Labs I worked with Steve at Microsoft, I got to know Jeff really well at Kildare when he was at Facebook. But in terms of them improving returns, I think that's actually no different than maybe for us, which is, I think there's two big things that venture firms at the top level can help with. One is recruiting and then the other is partnerships or potential business development opportunities. I think beyond that, from a firm perspective, it comes down to the individual partner and what he or she brings to the table to help that entrepreneur finally have an unfair advantage.

I think that actually is no different in the angel realm. Obviously, bigger venter teams have more resources than one individual, but if you have a network of people, of engineers and designers that you can help that company get access to, that's going to increase your probability of actually getting a better return, right? Because you're de-risking the business.

What are your thoughts on blockchain technology, like Bitcoin?

I spend a lot of time working on a number of blockchain and Bitcoin companies during the last year or two and, as I've said before, we don't have a like a concentrated focus on seed. A gentleman and I started something called KPCB Edge about a year and a half ago, in which we basically write checks no bigger than 100K for startups in five different technology areas. We feel like these tech concentrations may require longer lead times and blockchain is one of them. Bitcoin, this is my own view there: if we're really going to invest in a Bitcoin company, it may make sense for us to simply buy Bitcoin, because most of those companies will likely be indexed to the value of the currency. What I do think, actually, is as a distributed ledger technology, blockchain is really quite interesting. We have actually done investments in companies like Align Commerce that happen to be in implementation detail for delivering a blockchain solution to the end customer.

Are there particular companies that you're most excited about right now, specifically in computing and infrastructure?

That's a broad question. Earlier this morning, I spoke to a group of CIO's in the South Bay from a bunch of Fortune 100 defense companies. What's interesting is even there you're seeing this kind of fascination around the rise of micro services and containers. There's this demand or desire to move faster. Part of the driver behind that is not just actually having those teams move faster, it’s changing how well these companies build software to be able to attract and retain talent. If you're a new grad and you join a company and you're not shipping anything, it's unlikely you're going to stay very long.

Even the Fortune 100 companies are realizing this, so I think the implications for infrastructure containers, the next thing beyond virtualization, are that things will continue to go in that direction.

Is KPCB Edge building tools and products? Why would a VC firm build such tools and what areas are you looking at building product for next?

Sure. Guys still write software here and we worked on building some software a couple of years ago, to help us to try identify these fast rising consumer companies. Then as an experiment, I'm a big fan of experimenting, we saw that there are some technologies that may have longer lead times than the kinds of companies we normally invest in. We wanted to augment our network in those rising areas. So Edge has three or four people who spend half their time writing software and the other half actually finding these entrepreneurs in these five technology areas.

The software that we built focuses on three areas. One is software for the firm to give us either a better way to engage with our customers, entrepreneurs, identify these companies. The second is tools that our entrepreneurs need to get access to, right? You want to get an extra introduction to Pepsi, like how do you find that within a firm? Historically, that’s been a process that plays out through email. That’s a very inefficient way to leverage a network that, in our case, has been built over 44 years.

The last area is software that we can offer future entrepreneurs who we should be backing. I tend to believe that over time there's going to be more and more engineers and designers who work at venture firms tp actually try to address this. Venture doesn't scale, but there's aspects of the business, let's say especially on introductions, that can scale through software.

Out of curiosity, have you guys started funding companies out of Edge?


What is the right way to go about coming up with a valuation for an early stage start up?

Well there's this really fancy equation... no, no - it's purely art, right? I do think that one of the things I've seen is that there is a recognition in the industry that if the founders give up too much ownership early, like in seed this movie will not play out well.

Can you give some color to that? Why?

Why doesn't it play out well? I won't name the company, but there's one company that I had to help resolve this issue, where a great entrepreneur, who maybe was out-negotiated, gave up too much of his company. This is something that has to be fixed, in my opinion. If investors hold a majority of the company after just four years, you may lose the focus of the founder, and that’s a big problem for everybody.

My guidance on this is like you have to be reasonable, like a lot of the folks that come to me, let's say on Twitter or wherever, I’ll say, "Look, in a seed don't give up more than 10% or 15%."

So the valuation has to reflect that reality, no matter where the company might be revenue wise or with its product.

Would you say it's fair to say maybe as a rough guidance to that kind of a question: be long term greedy as opposed to short term greedy?

That's exactly right. When I was raising a Series A in 2002, one particular general partner who was given the process of issuing us a term sheet was trying to convince me that having a really high evaluation was a bad thing. I'm like, ‘yeah, right, well he's a venture capitalist.’

Actually after spending an hour with him on it, I think there's validity on having an evaluation that's too high and an evaluation that's too low.

What areas of business operations do you see opportunities for AI or ML to disrupt?

I think there's interesting opportunities in decision support and customer support with the advances in natural neural nets for text processing. There've been a lot of announcements recently on other bots, Facebook, Microsoft. I think a really good use case today are web-chat windows where, arguably, 80% of the exchanges could be automated and maybe just have the last 20% of that supported by humans.

I know Facebook, for example, as a platform has built messaging into its product in a significant way. Do you see that going to the enterprise?

The question to ask about enterprise: What are the real problems that could be solved? I do think that end-user support and maybe internal IT support have reasonable use cases. Most scenarios have humans that predominantly answer similar questions over and over again. Maybe there's a way to automate that and make that easier. I don't know how that'll play out.

Last question: What would you say is the biggest mistake that you see entrepreneurs make as operators and when trying to secure funding?

I think on the fundraising side, entrepreneurs need to be thoughtful about who they think the right partner is at the firm. Does that make sense? As much work as you can do to understand what he or she has invested in before, then kind of tied to that is spending time with that partner. Nothing pains me more when I hear the company tell me, “I’ve got a term sheet that expires in 24 hours.” It's like exploding job offers. I remember I used to deal with this at Twitter and say, like, "Look, guess what, a day later people are still going to want to hire you". That's the same as the other case. It's more important, really, because you can't fire an investor. Both parties should want to spend some meaningful time together to make sure that you want to spend the next decade with this person.

On the operating side, I see a lot of communication inside startups break down as they get bigger. What works at five people doesn’t work at 30. I've been very fortunate where I've gotten to see the building of six engineering teams. Now I get to see more actually grow and the same issue happens at so many companies. Let's say at 25 people, if you don't have a new communication mechanism, or a little bit of structure organizationally, someone will come to you and say, "Ugh, you know what, like this doesn't feel like a startup anymore. We're really big." But their complaint has nothing to do with the fact that you're really big, that you're 25 people, but it feels big because the communication is actually not occurring.

It was occurring naturally before.

Exactly, where it's not as intentional. Like hey, we're all in one room now, we're sharing it over Slack or whatever. Where now you might have forum in all hands or you may have to carve out time to walk around and say, "Hey, how are things going?"

You're saying as entrepreneurs actually introduce those kinds of structural ways of facilitating communication it actually makes the company feel smaller.

It does and it’s a constant challenge when scaling. Another big thing that often challenges founders is ceding some actual control to their leadership team. That can be hard. But if founders don’t do it, their leadership team feels like they're being micromanaged and they sense that you don't want to let go. But if you don't let go, you’ll limit your growth. It's tough, it's a very hard thing to let go, because that startup is like your child.