General Catalyst’s Clark Talks Opportunistic Investing in Tech

Deep neural networks, distributed work, compute at the edge, and low-code/no-code have been catching the venture capital firm’s attention.

Joao-Pierre S. Ruth, Senior Editor

April 4, 2022

7 Min Read
photograph with Venture Capital written across front and multiple icons to support
Aleksey Funtap via Alamy Stock Photo

With the next evolution of compute, data, software, and cloud still taking shape, venture capital firm General Catalyst looks for opportunities to back innovators who might speak to the new needs of the market. Companies might be angling to bring workers back to the office but not every staffer is going to let go of the pandemic-driven paradigms of remote operations. Furthermore, companies who embrace the cloud may have new considerations that drive them to seek varied service options as they push more resources to the edge.

General Catalyst has invested in such companies as Glean, a work assistant and knowledge platform, and, deep learning AI platform for talent management. Quentin Clark, managing director with General Catalyst, spoke with InformationWeek about strategies that guide the firm’s investments. Clark, former CTO of Dropbox, says a mix of thematic and opportunistic funding makes up General Catalyst’s playbook.

What are some guiding philosophies behind General Catalyst’s investments in technology?

We have to balance thematic with what we refer to as opportunistic work. We have to pay attention and engage with companies that get referred to us through our founders and other parts of our network. There are other incubator functions--that is important for us to engage in because we don’t necessarily see everything as we view things thematically. It’s just impossible. We do some of our very best work when we are being more intentional.

An example of that, Ro is a [digital health] company we invested in a long time ago that came out of our Rough Draft Ventures program. That becomes opportunistic. But something like Glean or Eightfold, which are investments we made, those came out of much more thematic work.

Intentionality isn’t just about the themes. Part of being intentional is about understanding the kinds of companies we’re getting involved with regardless of the domain space. General Catalyst has this focus on responsible innovation, which means you’re building companies for good, growth, and aligned to impact on stakeholders. It’s not always easy but we have that mission to participate in that.

What areas are you spending time on?

These are not permanent. One of them is a continued evolution of job domains, so heads of HR, finance, sales, and support -- there’s a shift happening and not just because of SaaS platforms. We’re still in early innings on SaaS software touching all those domains. We’re seeing a shift from machine learning as just great algorithms to what’s closer to AI.

There is a technique called deep neural networks, which allows us to build more like a semantic model more so than a smarter algorithm. I can build a machine learning algorithm to do a better job for candidate-matching in a tracking system, but what Eightfold has done is build a semantic model over all your people data. Your employees, your org chart, their reviews, all the applicant information, and interviewing information. It allows me to do things that were not possible before like, “What skills is this department going to be short of in six months?” “Show me three candidates for this open role that are in far-flung departments away from where the role is that actually have high propensity to be successful in this job.”

So regardless of what the hiring manager says they need, how they characterize the job, something else can look at this semantically and say what’s happened in reality is people who’ve had a tour of duty internationally in a support function end up being great at this job. That prompt can help people see that is a good point. That’s what I mean by deep neural network, this next generation of ML. That’s hitting all these job domains. We are systematically looking at other domain spaces.


Another area is dynamic workforce, which is a little fuzzy. I fit things like, Awardco, and Hopin into these things, as well as things like Loom and Glean where it’s not just the tools end users are using because they are much more project-based than they used to be. Now it’s like, “You’re going to do this project and when that’s done, there’s another one. Maybe you do two at once and the teams you work with are different.”

It’s a different system that we’ve put in place. Distributed work is permanent now. We will get back in the office one, two, three days a week -- or not. All that is changing how employment is thought about and the deal between employers and employees. Employees have never had a larger voice at the companies they work for.

There’s more specialized tools, more things you have to operate between. That’s not going to get easier without software. We want more specialized tools. We want a midmarket treasury for a high physical goods SaaS company to come and help that finance department. That’s a good thing. But to take that finance person and make them constantly copy and paste and change roles, that’s a bit of a problem. We have to find ways to embrace that diversification of tools while giving people back some sanity of knowing where they are.

The third [area] is a distributed, data-centric forthcoming cloud platform. There’s a set of trends, one of them being multi-cloud, where businesses are waking up and saying they can’t afford to be single-sourced to AWS and need to have some bargaining power. We’re also seeing a world where shift-left, shift-right stuff where developers have more empowerment. They’re employing a lot more tools and more stuff is being distributed. There’s more of a need for performance at the edge. We’re pushing compute and data out to these edges. How do we secure those edges?

The other thing that is having a massive effect is more and more data is being born in the cloud in the first place or being moved there. These are huge forces at change and what does the data ontology end up looking like? How does that fit in with this serverless, more distributed, more edge compute world? That’s an area we’re spending a bunch of time on.

Has the cycle of assessing venture capital opportunities changed in response to current events?

I got here a few months before the pandemic hit. The nature of venture capital very much accelerated in 2020 and 2021. Things got a little more rational lately, but the pace was super fast. A lot of it was because of the efficiency of meeting companies over Zoom was way different than in person. We’re not waiting these long periods of time. We just needed to order the work differently than we did in the past. It’s the same net amount of work.

Have there been areas of technology that stood out or are surprising?

One area around low code, composability -- sometimes when you dig into stuff, you realize it’s something you took for granted, turns out to be a really important unlock. We as an industry have had a need to put APIs behind every SaaS product there is, like a Coda, or Notion, or Airtable, or Zoom. There’s been a need to put cloud APIs against those products. Mostly people assume interoperability is good and the thing that pushes this over the line to get done has to do with IT, regulatory compliance, and legal holds. That’s what's required for them to be in compliance with the law as a company.

We dove into this area of low code/no code, and I didn’t appreciate that change over the last few years has been this massive tailwind to allow people to build these automation systems. Five years ago, they were running around trying to convince these SaaS companies to give them a little scrap of an API to work with. Now these things are being born and they know they need to do this from the beginning.

What to Read Next:

What Bain Capital's David Humphrey Sees in Hybrid Cloud

Can AI Lead the Way in Low Code/No Code App Development?

AI and Data This Year: Bigger, Bolder, and Business-Focused

About the Author(s)

Joao-Pierre S. Ruth

Senior Editor

Joao-Pierre S. Ruth covers tech policy, including ethics, privacy, legislation, and risk; fintech; code strategy; and cloud & edge computing for InformationWeek. He has been a journalist for more than 25 years, reporting on business and technology first in New Jersey, then covering the New York tech startup community, and later as a freelancer for such outlets as TheStreet, Investopedia, and Street Fight. Follow him on Twitter: @jpruth.

Never Miss a Beat: Get a snapshot of the issues affecting the IT industry straight to your inbox.

You May Also Like

More Insights