James Lindenbaum on founding Heroku and Heavybit, why devs matter more than ever and what it takes to win
James Lindenbaum is the founder of Heavybit, a 9-month program dedicated to developer-focused entrepreneurs. Heavybit’s program is designed to help post-seed companies like CircleCI, Rainforest QA and Iron.io and their older counterparts like Stripe, PagerDuty and Meteor get traction with developers, grow their team, harden their technology and create and refine their go-to-market strategy. Prior to founding Heavybit, James co-founded Heroku, a developer-centric platform-as-a-service (PaaS) company which was acquired by Salesforce.com in 2010. Over 5 million apps have been built on top of Heroku and the platform currently processes 8 billion requests per day. He has written about and advised dozens of companies on developer-focused strategies.
Given James’ breadth of experience and expertise serving developers and entrepreneurs who are serving developers, we sat down with him recently to chat about trends he has seen in developer-facing businesses and why now is as good a time as ever to build one.
On founding Heroku…
“In the early 2000s, there was a renaissance in programming languages and web application frameworks like Django and Ruby on Rails, which were meaningfully easier to use — primarily because they were focused on developer productivity. This opened the door for a broader set of people to build software. As a result, there were awesome products being built, but we noticed that deploying and running software still sucked. So there was this this extreme juxtaposition of how easy it was to build great software vs. how terrible it was to run and deploy that software. Simultaneously, you had the beginning of Infrastructure-as-a-Service. AWS was new and you had the sense that this stuff was going to be super powerful but really needed to be simplified, particularly deployment. So we started Heroku to fix that problem, but specifically with the view that developer experience and productivity mattered, with the design of that experience being the focal point.”
On why Heroku succeeded…
“Early on in PaaS, you had a bunch of guys who focused on managed hosting or creating tools for AWS, and most of these guys viewed the deployment problem as a systems engineering or architectural problem. We believed you had to take a holistic view of what it meant to be a platform for developers. If you had to choose just one reason why we were successful, it was our focus on the developer experience — to a fairly extreme degree. We felt that things like whether the APIs were beautiful, the tools felt immediately responsive, or it was two commands vs. one command to get something done would have a material impact on winning, and it did.”
On what’s changed between Heroku being founded and now…
“Right now we’re in an interesting transition because everyone knows that developer experience is table stakes. Now everyone is creating purpose-built tools for developers and we’re amazed by the breadth of tools being created and use-cases being served. The issue now is one of decentralization and tool sprawl, so there’s a lot of confusion. There are a lot of interesting new problem areas that aren’t well defined yet, so there are folks nibbling around the edges but not quite sure what the core issues are. For example, we’re seeing a ton of tools emerging around the area of continuous delivery of software — continuous integration, developer environments, deployment tooling — ultimately with the goal of trying to create parity between development, staging and production environments and encourage rapid iteration. There are dozens of companies all with slightly different approaches and no one has really emerged yet. So there’s a ton of interesting work being done but also a ton of noise.”
On what it takes to win today…
“To win takes a combination of two things: 1) you have to have a strong point of view on what the problem is and the right way to solve it, and 2) you have to be correct. That’s fundamental and not a lot of people understand the core developer problems or are solving them the right way. The other issue is understanding the subtleties of the go-to-market. There’s a really big shift happening now in how enterprise buyers learn about and buy technology. Understanding that zeitgeist is absolutely critical for success. The big paradox around developer businesses is that the people who unquestionably build the best products (developers) are also the people who have unquestionably the least knowledge of what the enterprise buying cycle looks like. A big part of what we’re doing at Heavybit is trying to address that specific problem. Luckily, that’s mostly an education problem.”
On founding Heavybit…
“After Heroku, many entrepreneurs who were starting developer tools companies came to me for advice, and it became clear that there’s a very specific set of common problems all these companies face, mostly around go-to-market. There’s really only one good way for these companies to go to market, and many just don’t know what it is, or more specifically how to execute on it. Helping these entrepreneurs understand the strategies and tactics and actually coaching them through it — giving them the right resources, connecting them to the right specialists in their specific domains — is really critical. I’ve seen many companies that should have been successful because of the product and market go through a lot of pain and burn through a lot more capital because they just don’t know how this stuff works. So that’s the main goal here: to provide the education and resources to make these companies successful. Secondly the goal is to provide community, because it’s clear that the founders of these companies are not talking to each other as much as they should be, given how similar the sets of problems are.”
On the emergence of the bottoms-up adoption model…
“The reason everyone has converged on the bottoms-up go-to-market has to do with large secular forces. Developers have become incredibly important and influential in the buying process. You can’t just sell something to the CIO over steak dinners and have it shoved down the developers’ throats. Even if you could, it’s too expensive; it’s not competitive. If there’s a bottom-up, high-efficiency go-to-market strategy possible for a given market, anyone who does it will kick the ass of anyone who’s doing an old-school, top-down, long-cycle, high-cost sales model.”
On the challenges of bottoms-up go-to-market…
“It’s really hard to understand the difference between your “organic” product — the product that your developers use and love — and your business product, which needs to be, effectively, a separate product. Whether they look like different tiers of the same product is just packaging. It’s like crossing the chasm: you have to understand at what point you start the customer discovery process over again and develop that business product. Understanding the relationship between your organic product and business product, how your organic product drives leads for and sales of your business product is critical. Understanding how to do product management becomes a huge pain, because, again, the organic product is being built by developers for developers, so in the beginning you have amazing product intuition. Over time, as they’re successful and their tools are adopted in the enterprise, the buyer drifts further away from being people like them and pretty soon that chasm is so large that you have no intuition for what you need to build. That’s when you have to start doing things that look more like traditional product management. You have to talk to customers, ask questions, and that’s a process-oriented thing that engineering-led organizations are allergic too — it’s a cultural challenge. That’s one of the critical things we see companies struggling with, particularly between the Series A and Series B. Other major issues include how to build a sales team, how the sales team functions, and how the culture of the sales team fits together with the culture of the engineering team. All those are common problems that are key to navigate.”
On why developers rule now…
“All companies are software companies now. If you buy the ‘software is eating the world’ thesis, the way a company does anything today — any kind of business process — it’s all software. And so as a business, no matter what your size, your agility — your ability to roll out a new product, change processes, manage your people, etc. — is equivalent to your ability to develop and change software. So your software development velocity determines your competitiveness. So developers and their productivity become the most critical resource of any company. That’s starting to be realized by a majority of the world.”
On the saying: “you can’t make money selling to developers,”…
“There was a time when developer tools were very valuable. Then some crazy things happened, most notably Microsoft killing Borland. Power became consolidated in a few platforms (Microsoft, SUN/Java), and those platforms realized that giving away really good developer tools would strengthen their positions. In doing so they commoditized the market and destroyed the opportunity for anyone else to build really good developer tools. The lesson that taught everyone, particularly enterprise buyers, was that developer tools are not something that you pay for, they’re something the platform company gives you. It also taught entrepreneurs and investors that dev tools are not strategically defensible. But that was a very long time ago. Since then, all those platforms have been broken down and decentralized — linux, open source, etc. There’s no one who controls the stack to the degree that, say, Microsoft used to. In the meantime, because of Moore’s Law, open source, cloud, etc. developer productivity has become the most valuable asset for companies. So in reality, dev tools — even pure tools — are one of the most valuable things there are and will only grow to be more valuable as more companies become software companies. I believe there’s a market dislocation now because the world has clearly changed but people’s perceptions are still stuck in the past. As perceptions become unstuck, the value of these products will get corrected rapidly. I believe we’re beginning to see the unsticking.”
