Where top VCs are investing in open source and dev tools (Part 2 of 2)

In part two of a survey that asks top VCs about exciting opportunities in open source and dev tools, we dig into responses from 10 leading open-source-focused investors at firms that span early to growth stage across software-specific firms, corporate venture arms and prominent generalist firms.

In the conclusion to our survey, we’ll hear from:

These responses have been edited for clarity and length.

Gaurav Gupta, Lightspeed Venture Partners

Which trends are you most excited about in open source/dev tools from an investing perspective?

Back in the day, developers had little budget or power and they relied on IT (or business people) to drive purchase decisions. But things have totally changed in the last few years.

Developers are now the key influencers/decision-makers for a whole host of new software categories where we see bottoms-up adoption models succeed. This is happening in areas like security, observability, design and collaboration tools, AI/ML, data analytics and visualization, and cloud infrastructure.

As these areas become more developer-centric, they are also getting disrupted by either open-source or cloud-centric offerings. Over time I believe we’ll see traditional commercial on-prem software vendors slowly die away. This makes it a great time to be a startup founder and disrupt the incumbents across many of these categories.

How much time are you spending on open source/dev tools right now? Is the market under-heated, overheated or just right?

I spend roughly half of my time on OSS and developer-related tools. The reality is that in many areas of enterprise infrastructure, having an open-source business model is pretty much table stakes to be competitive. The flip side is that sometimes open source is a bad choice for startups, especially when a product requires a top-down selling motion and has a high cost of customer acquisition.

The successful IPOs of Elastic and MongoDB — and the multi-billion-dollar private valuations of companies like Databricks, HashiCorp and Confluent — have created high expectations around revenue multiples and, as a result, valuations have skyrocketed. Things definitely feel overheated, but I remain optimistic that there is still a lot of room for innovation.

Any other thoughts you want to share with TechCrunch readers?

I’m a big believer that most open-source startups need to really be hyper-focused on adoption in the beginning. Too many startups rush to monetize too early, feeling the pressure from large VC financing rounds.

The reality is that if you lose your developer community in the process, it’s almost impossible to recover. Navigating the trade-off between monetization and adoption is an area where I believe investors and founders need to make sure they are absolutely aligned.

Julia Schottenstein, NEA

Which trends are you most excited about in open source/dev tools from an investing perspective?

There is a huge opportunity in data engineering. Getting data in a useable format is still a nightmare for most companies and the complexity grows as companies scale.

We’re excited to partner with startups that simplify the data management life cycle from workflow, access and privacy, to queries, to underlying infrastructure.

How much time are you spending on open source/dev tools right now? Is the market under-heated, overheated or just right?

I spend a majority of my time in open source and dev tools. Developers are still key decision-makers in companies and the growth of the developer ecosystem is impressive (10M new developers on GitHub last year).

Open source/open core have become the essential model to get distribution and adoption quickly. These open-source “side projects” often become widely adopted and are proof of a healthy market need.

Are there startups that you wish you would see in the industry but don’t?

The app integration layer is bigger than most people understand since a lot of the connective tissue is solved today with human power. On the spectrum of a prosumer and simple solution (Zapier) to an enterprise-ready and robust solution (MuleSoft), there’s room for an open-source platform to play in the middle, providing something more complete but still easy to set up.

Any other thoughts you want to share with TechCrunch readers?

Companies are maintaining less and less of their development stack. Innovative, large tech companies have built a lot of internal tools for their developers to improve the software development process and lots of these tools are being open-sourced now so the whole ecosystem can benefit. Uber, Airbnb, Facebook, Netflix, LinkedIn, Spotify and others have spawned really interesting open-source startups.

Peter Sonsini, NEA

Which trends are you most excited about in open source/dev tools from an investing perspective?

We’re seeing almost the entire software stack become open-sourced. Traditionally, the best open-source investments have been in infrastructure or lower in the stack.

While we continue to be excited about those opportunities, we’re seeing success of open source in the application layer as well, particularly in messaging, data analytics, RPA, customer data platforms, content management systems, design systems and more. Not all business applications will make sense for open source of course, but we’re excited how the pie has grown.

How much time are you spending on open source/dev tools right now? Is the market under-heated, overheated or just right?

We’re very focused on open source and dev tools and have been investing in this market for over a decade: Databricks, Nginx, MuleSoft, MongoDB, Elastic, Sonatype, Pentaho, Sourcefire, etc. We continue to be bullish and we bring experience and team in scaling tools into platforms.

The number of $1B outcomes for developer-focused companies has grown significantly over the years and we don’t see that slowing down. There will always come along a shiny new and better framework, language or library to make a developer more productive, so the prospects for building $billion+ outcomes in this category will continue to present themselves regardless of any point in the VC cycle.

Are there startups that you wish you would see in the industry but don’t?

We try and see as many new projects as humanly possible and move quickly to become the best partner VC for true innovation and world-class entrepreneurs, which by definition, just don’t come along all that often. NEA has been closely involved in so many successes in this category over the years that we tend to “know it when we see” it and move very swiftly.

There’s been a lot [of] investment in the design space recently and I still think the hand-off between designers and developer is broken. We’d love to close that gap and make the design and development process way more integrated. We have an early-stage investment in Plasmic that will be a part of that solution.

Any other thoughts you want to share with TechCrunch readers?

Adoption is the most important metric at the earliest phases of open-source companies. Stars on GitHub is a nice vanity metric, but if an open-source project has a strong community, healthy contributors and lots of downloads, then they have a shot of building a valuable business.

Developers can be hard critics, but they are also loyal to great tech. If you’re building an open-source company, focus first on adoption, then on revenue. It’s never too early to come talk to us. We’d love to hear from you!

Salil Deshpande, Uncorrelated Ventures

I spend all my time on infrastructure software, both traditional and decentralized. And much infrastructure software is open source. So although it’s not the only monetization method available, I do still care a lot about the flywheel of engineers implementing their innovations as open source, battle-testing and hardening the software in the process, and providing additional non-open-source features or services to monetize, while still maintaining and improving the open source. This virtuous sustainable cycle met the needs of the community, the enterprises who wanted to pay for advanced features and the authors of open source.

Amazon Web Services (AWS) and cloud infrastructure providers perturbed this cycle a few years ago. For a couple of years now, I’ve been calling attention to a situation whereby, the definition of open source gives Amazon the freedom to take open-source code, run it as a commercial service and give nothing back to the authors of open source.

But Amazon is not to blame. Amazon’s behavior toward open source is self-interested and rational, and Amazon hasn’t needed to worry about whether its behavior is sustainable. Amazon is playing by the rules of what software licenses allow. Blaming Amazon for this situation is like blaming leopards for preying on gazelles.

The blame lies with the ossified stances by industry associations who believe that unless Amazon has the freedom to take your code, run it verbatim as a commercial service and fully intercept the monetization, your code is not open source. At the heart of this ossified stance is the tenet that open source licenses cannot restrict any specific uses — even if that use is one that might strip-mine the open source and make the cycle unsustainable.

As a result there is still no standard way for authors of open source software to express a simple concept:

“I do not want my open source code run as a commercial service.”

Because if you try to express such a thing, then your code is deemed not open source – due to the sacrosanct definition of open source.

Fortunately, the industry found a way around such ossified stances. That way is so-called source-available licenses. More than a dozen high profile open source companies have switched from open source licenses to source-available licenses in the last year. Sidestepping the baggage and myopia of ossified definitions of open source, these source-available licenses do not require that Amazon have the “freedom” to take your code, run it as a commercial service and give nothing back to you.

Way more often than not, I now meet early stage infrastructure software companies that plan to give their interesting features either a source-available license or a closed-source one, rather than open source. This wasn’t the case before. It’s a shame because “open source” had a good brand; the phenomenon of innovative companies moving away from open source will weaken that brand. Not to mention the license proliferation.

All that said, infrastructure software has been and still is a wonderful area to be investing in. Half a trillion dollars of value has been created from scratch by infrastructure software startups, just in the last ten years and just from exits over $500M (because those are easy to tabulate). We are in a renaissance period for software, and infrastructure software in particular; and nowhere near the end of the period.

I’ve been lucky enough to be able to invest in many of them including MuleSoft, DynaTrace, DataStax, Tealium, SpringSource, SysDig, Frame, Redis Labs, Quantum Metric, Compound Labs, Mobile Passport, Dropcam, Sonatype, Netdata, Opsani, Dgraph, Gradle, Fossa, Rollbar, StorageOS, MakerDAO and others; and wish I had invested in many other great ones that I missed. I’m looking for another trillion dollars of value creation in infrastructure software over the next decade.

Ethan Kurzweil, Sakib Dadi, Jenny Gao & Mike Droesch, Bessemer Venture Partners

Bessemer has been actively investing in developers’ tools and OSS across the entire stack since 2008 when we first invested in the seed round of Twilio. While there is certainly more noise than there was 10+ years ago, it has never been a better time to be a developer as more investors and companies are focused on building products that provide incredible developer experiences.

We publish our thinking on this roadmap more broadly in our laws of developer platforms, which you can read about more here. We are most excited about developer technologies that not only enable developers to become part of the product development process, but also empower other functional groups within the organization ranging from DevOps to product managers to designers and occasionally even front of house functions like marketing, sales and support.

This democratization of development is something our portfolio company LaunchDarkly has enabled by making adopting feature flagging drop-dead simple and expanding to a platform that allows for the smooth rollout of features by cross-functional teams.

As part of our work on developer-focused businesses, we have also spent a lot of time thinking about open-source software. While we are still interested in backing technologies addressing core infrastructure problems, like our investment in HashiCorp, we are becoming increasingly excited by OSS moving into different verticals and up the stack to the front-end and even the application layer.

Bessemer has backed an OSS front-end testing company, Cypress.io, due in large part to the amazing developer experience and strong OSS community they have cultivated. We continue to spend time focused on trying to identify the strongest OSS communities as we believe that this is one of the most essential ingredients for building a successful open-source project and commercial entity.

There is a lot of complexity around taking a project, whether it be from a big tech company with stringent IP restrictions or a small group of individuals working remotely and developing a company around it. As a result, we’ve been focused on ways to measure healthy community engagement and continue to look for companies who understand the importance of cultivating strong, healthy communities.

Lonne Jaffe, Insight Partners

At Insight Partners, our focus is long-term investing in companies that have already achieved product-market fit but are still growing very quickly — we’ve been calling these fast-growing companies “ScaleUps,” as opposed to startups. Many of these ScaleUps are built around open source or have a major open-source component, especially the ones that are delivering software that helps people create more software.

ScaleUps are substantive enough to apply material resources to software development and go-to-market, however, they are still focused enough to have a mission-oriented culture and this can make it easier to attract and retain top open-source leadership organically and via acquiring smaller open source-centric startups.

For example, our portfolio company Aqua Security — the ScaleUp leader for cloud-native, developer-centric security software — supports the open-source projects Trivy, CloudSploit, kube-hunter and kube-bench. When CloudSploit and Trivy joined forces with Aqua, it wasn’t an “exit” — rather the leadership of CloudSploit and Trivy were able to fuse with a border platform with lots of growth ahead and culture of respect for open source.

One domain we’re actively looking at for new investment is ScaleUps that help developers more easily include prediction and machine learning capabilities in robust production-grade software systems — in a way that preserves both security and privacy.

Building powerful, immersive software that is differentiated versus the “built-in” capabilities of large public cloud companies takes time, and curating an open-source developer community and an application ecosystem takes even longer. This is why developer-centric ScaleUps need patient investors that are willing to invest over long-term time horizons and support visionary founders and executives with practical, hands-on growth expertise.

Jai Das, Sapphire Ventures

Which trends are you most excited about in open source/dev tools from an investing perspective?

We are spending a lot of time looking for opportunities in DevOps tools. According to IDC, the DevOps market will be $15 billion by 2023. The number of developers globally is expected to increase from 23.9 million in 2019 to 28.7 million in 2024.

The volume of software written by these developers, the demands for a faster release cycle and the increasingly complex layers of dependencies and security requirements of software are increasing the need for DevOps tools. We are looking to invest in companies that are focused on all aspects of developing and deploying software including writing, compiling, testing and securing software.

We are also looking for startups that help developers collaborate and be more productive. Open source is a GTM strategy for DevOps tools and infrastructure software. Therefore, most companies we are evaluating in these sectors have an open-source strategy to increase adoption with developers.

How much time are you spending on open source/dev tools right now? Is the market under-heated, overheated or just right?

The infrastructure team within Sapphire spends the majority of our time focused on two areas. First, we focus on tools to enable developers to be more productive, to collaborate and to write, compile, test and secure software better and faster.

Second, we focus a lot on the API economy which provides all the Lego blocks (e.g. payments, AI/ML, customer experience) needed to quickly build software applications that can scale rapidly. Enterprise software companies that are growing rapidly with efficient SaaS metrics are in high demand and get high valuation. DevOps companies with these characteristics therefore also get a lot of attention and high-revenue multiple valuations.

Are there startups that you wish you would see in the industry but don’t?

One area of DevOps that is still nascent and where we would like to see more startups are in software testing and in developer productivity. Also self-generating software based on AI/ML is another area of untapped innovation.

Any other thoughts you want to share with TechCrunch readers?

Every company is now a software company and there is an insatiable demand for software developers.

There are just not enough software developers being trained even though computer science is now the most popular undergraduate major in U.S. colleges. Therefore, automating the software development process as much as possible will be key to help make every company into a software company.