A Conversation with Benchmark’s Chetan Puttagunta – Part One
In Part One of their conversation on the AI opportunity for venture capital, Benchmark General Partner Chetan Puttagunta and Adams Street Partner & Global Head of Fund Investments Brijesh Jeevarathnam discuss the substantial market opportunity and AI’s power to redefine workflows with the potential to generate significant value across industries.
Chapters
00:21 Introductions
01:31 An unprecedented market opportunity, a spectacular moment in the innovation cycle
3:32 A $1 trillion annual enterprise IT and IT services total addressable market: “We’re going after the whole pie with this AI stuff”
5:41 The impact of AI on GDP and global productivity, shifts in consumption patterns as the market grows
7:34 Case Study: How Benchmark portfolio company Sierra is transforming customer support with AI
10:52 Revenue model experimentation in AI applications
12:49 Case Study: How Legora is upending legal workflows by giving lawyers AI superpowers
Transcript
Brijesh Jeevarathnam: Hello and welcome. Our guest today is Chetan Puttagunta, General Partner at Benchmark.
Benchmark is an iconic Silicon Valley venture capital firm. They’re renowned for backing such marquee companies as Twitter, eBay, Uber, Snap, Discord, and many more. They’re known as well for having a culture that is flat and equal as a general partnership, and Chetan is one of those general partners.
We are fortunate, Adams Street, to be a limited partner with Benchmark going back 30 years to 1995, starting with Benchmark Fund 1.
So it’s great to have you here, Chetan. You’ve been a venture investor now since 2011, and you’ve had tremendous success along the way, including backing companies and building companies like MuleSoft, MongoDB, and Elastic, and many more to come, I’m sure. Anything you want to add to the introduction?
Chetan Puttagunta: No, we deeply appreciate the support of Adams Street this entire time, and you’ve been great partners to us, and we’re primarily focused on the early stage business, and we’ve been consistently focused on that, and so we deeply appreciate your support.
Brijesh Jeevarathnam: Thank you. So let’s dive right into it. Let’s talk about the subject du jour, which is AI. Let me start with a broad question. There’s so much AI hype, froth, whatever you want to call it. Is it real? Is it different than prior innovation cycles? Why is AI so at the forefront of the zeitgeist today?
Chetan Puttagunta: I would say that it’s absolutely real. It’s perhaps the biggest innovation cycle that I’ve been a part of as an investor. Whether you want to measure it on ROI, whether you want to measure it on efficiency, increased revenues, et cetera, these technologies and these software companies and these models are delivering remarkable amounts of value very, very quickly.
And while the activity and the funding rounds and the scale of everything seems unprecedented, I think the direct analogy is the internet. I think it’s just if you go back and look at some of those companies, and how they scaled, and their funding rounds for their time — inflation adjusted, of course — there’s a lot of analogies to that.
And in the moment, I think it’s hard to know exactly who’s going to persist and who’s not, but you can sort of back up and kind of look at who’s delivering real value to customers today, who’s establishing data gravity, for example, in B2B software. Like, the principles have always been like, value is created when you capture a lot of data and you create data gravity around your piece of software. And you can look at these companies and start to think about who’s actually doing that right now, who’s getting very sticky customers, who’s getting a lot of customers in the trial phase, and who’s getting customers in production phases.
And as you examine that, you start very quickly realizing this is a pretty spectacular and special moment in the innovation cycle. And as an early stage investors, you know, as you know, being an LP of ours, we’ve been fully in this world for the last couple of years.
Brijesh Jeevarathnam: If I could frame just that what you said, let’s start with the macro level first. The addressable market for AI is seen as much bigger than prior technology, potentially much bigger slice of the GDP, including service economy. What do you think of that?
Chetan Puttagunta: If you look at the consumer world, you can look at that TAM and it’s like, what is exactly the TAM of this thing? And I think what we’ve seen already with ChatGPT, and all of the other consumer products that have gotten significant revenue scale and consumer usage scale, you can very quickly look at the sort of prevailing business models there, whether it’s advertising or subscription, and you can start to pencil out how big the TAMs are they’re addressing. I mean, these things are some of the biggest markets in the world and they’re going after them with amazing amounts of disruption.
On the B2B side, if you think about overall IT spend, combined with overall IT services spend, it’s on the order of $1 trillion per annum worldwide, and we’re going after the whole pie with this AI stuff.
When do these companies actually make a dent in that $1 trillion market? I think they won’t seriously start making dents for a couple of years because I think relative to large software incumbents, the absolute revenue scale of all these companies combined is still small. It’s just the growth rate and they’re growing, you know, many multiples faster than the incumbents. I think there’s a lot of anticipation of when it’ll start to make big dents in the macro world, and obviously, ChatGPT has made a dent in the macro world.
And the moment for AI in the consumer world is here. It’s happening. And I think in the enterprise world, it’s also happening. It’s just we’re not at the point where we’re starting to see it seriously dent incumbents yet. But if you watch the trend, you can kind of see where it’s going.
Brijesh Jeevarathnam: It’s coming.
Chetan Puttagunta: Yes.
Brijesh Jeevarathnam: We’ll come back and tie the addressable market size, massive, with the valuations of the companies. Let me ask you one more macro question. A number of technology thought leaders, for example, the CEO of Microsoft, Satya Nadella, talks about how AI could drive global productivity and maybe GDP growth of 10% a year, etc. And that sounds obviously aspirational. It sounds amazing. Do you believe in that? And if so, in what time frame do you think is possible?
Chetan Puttagunta: Every time there’s been a massive technology wave, humans have done a really good job adapting to it and building around it. The big difference between internet, mobile and AI is just the speed of change, primarily because everybody’s connected now via the internet, and so AI is able to spread through the world much, much faster.
I don’t know that the productivity line is such a straight line. Maybe if you zoom out over a 25 to 50 year period, that’ll be the CAGR, perhaps, but I don’t think it’ll be smooth. I think it’ll be pretty bumpy because there’s going to be pretty interesting shifts in how consumption works as these technologies get broader and broader.
The overall pie will grow, and grow rapidly, but there’s going to be shifting allocations in the pie.
Brijesh Jeevarathnam: The slices will change very radically.
Chetan Puttagunta: And I think that’s why it won’t be linear. And I think as we look at where the effect has already shown up today, obviously it’s been in the hardware, power, and data center space. Software investors, we just talked about how it hasn’t really dented anything yet, but it’s certainly dented things at the semiconductor level, at the manufacturing level, at the construction level, at the power level. If you’re in those industries, you have felt it. If you’re removed from that, you haven’t. It’s not come to you yet, that GDP increase.
Brijesh Jeevarathnam: That’s a good segue to my next question, which is to make it real for our audience, can you share today’s examples from your portfolio, perhaps, of companies that are transformational for their customers in terms of, let’s say, for enterprises, driving revenues, reducing cost, service quality, et cetera? Would you mind?
Chetan Puttagunta: Yeah, I’ll give you two very specific examples. We have a company called Sierra, which we backed at formation. The co-founder and CEO of this company is Bret Taylor. This is our third company with Bret, and what they do is, they’ve created AI software that allows enterprises to plug into their systems and basically take over all of their customer support function.
And so they suck in all of your data and program their AI agents to basically interact with your customers to give them the best possible experience that you have delivered. But when you deliver a great experience, it’s like one out of 20 is great, or one out of 50, or one out of 100 is great. With Sierra, it’s like 100 out of 100 is great.
And so what ends up happening if you use Sierra is you see a dramatic increase in customer satisfaction, a dramatic decrease in customer wait time, customers getting annoyed with you, and all that stuff. And one impact on the enterprise is that you can drive revenue growth because as customers have great experience with customer support, that customer support channel can drive revenue expansion with the consumer.
And obviously, this is like dramatic cost reduction as well on support. Very few companies do support in-house. You usually contract with a third party vendor. It’s very chaotic. There’s massive turnover in this population. Nobody’s really trained for your products. Nobody’s super proficient in how you should handle customers, and so this is why you get one out of every 100 interactions is great.
With this software, you bring it in-house and you just do it with AI, and it delivers dramatic value to companies.
Brijesh Jeevarathnam: In this example, does the customer see that kind of step change very quickly? Does it take some time? Can you want to walk through the timelines of that?
Chetan Puttagunta: This company was formed in January of 2023, and they recently announced a funding round at a $10 billion valuation.
And I will tell you that the basis of the valuation is the very specific customer case studies that you can see from them about the amount of value delivered by the software. Not only in revenue acceleration for the company, but also dramatic effects on cost. To call it best in class would be an understatement. It’s in a class by itself.
It’s relative to SaaS, the numbers don’t make sense. There’s almost no incumbent that offers this. There are some upstarts that offer this, but the market is so huge that the upstarts are figuring out their own lanes, and Sierra is kind of running on its own in the enterprise space.
And as a result, I think they are more capacity constrained as a company than demand constrained, to be honest, and that’s a very different dynamic in software. That is not how SaaS worked.
Brijesh Jeevarathnam: And these companies — I want to go back to your second example as well — but are these companies charging not so much on a seat basis? So can you talk about the revenue model of these companies, too?
Chetan Puttagunta: For Sierra, I think they can be flexible on how they price, and they blogged about this, too, which is like, one of the interesting things when you do this is you can sort of price by the job being done, or the work, or the task being automated here.
And so the pricing model here is not pure consumption. It’s not pure, sort of, seat-based subscription. It’s more about, sort of like, work that’s getting done, or work that’s getting automated, and so these contracts can take on a different flavor.
I’ll give you an interesting example. We recently met a company that was in the sales space that was thinking about charging based on the deals that they helped their customers land. So, like, 20% commission on whatever deals we drive for you. That’s a really interesting business model. I think there hasn’t been that much business model experimentation in the SaaS world as an example, because it basically you either bought a term license that was a value based sale, or you bought a seat-based license, and then in the data products, you did consumption based stuff, because based on either the amount of data being stored, the amount of data being moved, the amount of data being transformed, whatever.
So these business models can have more experimentation. But I would say right now, there are a lot of models, business models in AI, that are primarily seat-based subscription, and then you can layer on consumption on top of it and stuff like that, because I think it just makes the buying decision much easier. Like you can just say, okay, well, first contract, let’s just do a seat-based thing.
Brijesh Jeevarathnam: Get in the door.
Chetan Puttagunta: Yes. And then as we think about upsell and the second contract, then we can experiment with the business model.
Brijesh Jeevarathnam: Is it fair to say the revenue model experimentation is because the value to the customer, the ROI, and the timeframe, they’re so tangible, they’re so measurable. That’s why this is possible versus the prior cohort of technology companies.
Chetan Puttagunta: That’s absolutely right. I’ll give you the second example I have is Legora, which is a spectacular company that sells software to lawyers.
So one of the things they do for lawyers is they have this incredible feature called Tabular Review, which allows lawyers to compare attributes of contracts across thousands of documents all at once. That software will process all this stuff.
So it’s a great case study of one of their customers. They’re doing an M&A case and they had just forgotten or missed analyzing 1,000 contracts. And so it was late at night, partners get together and they’re like, “do we wake up the whole firm and basically staff every associate in the firm on this thing, and it takes two or three days and should we just do that?”
Then one of the other partners says, “we’ve got this AI thing. We bought this AI thing. Why don’t we throw it at that and see what happens, and then tomorrow morning we can decide.”
They throw 1,000 contracts into Legora, they tell the software, here’s how we want to analyze, and they go home. And then the next morning it’s done. And each cell is cited with the exact passage back in the contract – it highlights it, it formats it beautifully.
And then they use the associates in the firm to review Legora’s work, and they find that it’s made fewer errors than what they would naturally expect errors to be made. And it did it faster, did it while everybody was sleeping, and what seemed like they missed something that slipped through the cracks, it just recovered.
So for the firm, they delivered incredible customer value. They didn’t pull all their resources to work on something for 72 hours and thereby basically slow business down everywhere else. So it was massive aggregate value created because they didn’t have to restaff on an emergency. Two, they were paying Legora on a term license anyway.
And after this, one of the partners was like, “we got to get this to everybody. Imagine how much time we’re wasting around the firm. Everybody should just use it.”
One of the questions that everybody had about AI for lawyers was, the legal business model, especially for large law firms in the United States, has been billable hours. So if you introduce AI into this, won’t you reduce billable hours? And is it against the business model of law firms?
What’s actually transpired is that, if you talk to big law firms, they’ll tell you that they don’t get paid for the rote work. No client of theirs, no serious client of theirs, is going to let you bill hours for document processing, document review, all this kind of stuff. The serious billable hours show up if you’re going to court, providing strategic advice, or doing actual negotiation. So there’s some human element where it provides strategic value. That’s where partners with really big books of business, that’s how they deliver value to their clients.
And similarly, Legora sells to law firms and to enterprises. So they sell to lawyers everywhere. And so for enterprises, increasing the efficiency and capacity of a legal team is really important, because they don’t want to create an in-house staff of thousands of lawyers. They want to use software to get more efficient.
So what’s happening is, because Legora is going to these law firms and giving all of their attorneys these AI superpowers, everybody starts to become more and more efficient. Everybody then also delivers far more value to their clients.
And what’s happening now is law firms are getting more profitable than they’ve ever been. And so it’s a really interesting dynamic.
And I think what’s also starting to happen — I just spoke to a senior partner at a law firm last week who was saying that, as a result of AI software, specifically Legora, he’s spending way more time with his associates on the apprentice side of law. And his view was, like, maybe we transform the model where you can actually test younger people earlier on if they can be partners earlier. And so, with software, could you see somebody going from a first year associate to a partner in five years?
Perhaps. Because they’re just really good at the client facing side of the job, and then the document processing part of the job is just done by computers. And so that stuff is dramatic, and it’s happening.
And so has it hit the GDP broadly of the United States? No. But it’s now being consumed by one of the biggest industries, biggest service industries, in the United States. And it’s getting more efficient. It’s getting more profitable.
Well, what does that mean? As these firms get more profitable, they’ll invest more somewhere, because excess capital will find its way elsewhere. So that drives growth and consumption elsewhere. And so these are interesting trends that you can already start to see blossoming, and you can sort of play this out, and you’re like, OK, well, extend this one year, two years, three years.
And I think there was a big question of like, in legal specifically, will there be a role for humans? This was like the big question in 2022 and 2023. And in 2025, I’ll tell you that question, in my mind, has been answered, which is specifically, there is human-to-human interaction, that’s how the world does business. That’s how the world conducts society and legal systems and stuff like that. And that’s where the value is now accruing because that becomes so much more critical, because everything else, like the rote processing, is now just going to be done by software.
Brijesh Jeevarathnam: It’s a good example because there’s so much concern around AI, quote unquote, replacing the human. And I’m sure there’s many more examples like this, where it’s really elevating the human’s productivity and value, quote unquote, in this case, billable hours. And it’s a true, I guess, in the sense of a co-pilot, if you will, to make the whole system more productive and more valuable. That’s a great example.
Important Considerations: This information (the “Recording”) is provided for educational purposes only and is not investment advice or an offer or sale of any security or investment product or investment advice. Offerings are made only pursuant to a private offering memorandum containing important information. Statements in this Recording are made as of the date of this Recording unless stated otherwise, and there is no implication that the information contained herein is correct as of any time subsequent to such date. All information has been obtained from sources believed to be reliable and current, but accuracy cannot be guaranteed. References herein to specific sectors, general partners, companies, or investments are not to be considered a recommendation or solicitation for any such sector, general partner, company, or investment. This Recording is not intended to be relied upon as investment advice as the investment situation of individuals is highly dependent on circumstances, which necessarily differ and are subject to change. The contents herein are not to be construed as legal, business, or tax advice, and individuals should consult their own attorney, business advisor, and tax advisor as to legal, business, and tax advice. Past performance is not a guarantee of future results and there can be no guarantee against a loss, including a complete loss, of capital. Certain information contained herein constitutes “forward-looking statements” that may be identified by the use of forward-looking terminology such as “may,” “will,” “should,” “expect,” “anticipate,” “estimate,” “intend,” “continue,” or “believe” or the negatives thereof or other variations thereon or comparable terminology. Any forward-looking statements included herein are based on Adams Street’s current opinions, assumptions, expectations, beliefs, intentions, estimates or strategies regarding future events, are subject to risks and uncertainties, and are provided for informational purposes only. Actual and future results and trends could differ materially, positively or negatively, from those described or contemplated in such forward-looking statements. Moreover, actual events are difficult to project and often depend upon factors that are beyond the control of Adams Street. No forward-looking statements contained herein constitute a guarantee, promise, projection, forecast or prediction of, or representation as to, the future and actual events may differ materially. Adams Street neither (i) assumes responsibility for the accuracy or completeness of any forward-looking statements, nor (ii) undertakes any obligation to update or revise any forward-looking statements for any reason after the date hereof. Also, general economic factors, which are not predictable, can have a material impact on the reliability of projections or forward-looking statements. Adams Street Partners, LLC is a US investment adviser governed by applicable US laws, which differ from laws in other jurisdictions.
This Recording includes extemporaneous remarks delivered in an informal, conversational format. Statements herein regarding characteristics of Adams Street Partners, its investment process, past, current and potential investments, and market conditions or forecasts, as well as other statements are necessarily high-level in nature and should not be relied upon as a basis for any decision, including an investment decision. Many such statements are forward-looking statements and are associated with numerous known and unknown risks, uncertainties, assumptions and limitations. Before relying on this recording as part of any decision, including an investment decision, a listener must contact Adams Street Partners to determine whether there are additional relevant material facts and/or contextual information that are necessary to understand the statements herein, as well as whether there are any risks or limitations associated with the information described herein. All such information will be provided upon request. All characterizations or comparisons are the subjective impression of the speaker and need additional context in order to be reasonably understood (which could not be provided contemporaneously due to the informal nature of the conversation). In particular, statements of differentiation, such as describing advantages, favorable comparisons, quartiles, deal quality and other statements should not be relied upon unless the same statement appears in written format in a document prepared by Adams Street Partners for its investors.
