2024 Annual Investment Outlook

Listen to our 2024 Annual Investment Outlook Panel with four top VCs based in Seattle and the Bay area. Get their perspective on the investment climate in 2024 and their plans to invest specifically in AI/ML based startups.

Tim Porter, Managing Director, Madrona Ventures
Tim joined Madrona in 2006 and invests in B2B software companies in the Pacific Northwest. He is currently interested in intelligent applications and SaaS, cloud-native software infrastructure, the modern data stack, machine learning, DevOps, and cybersecurity. He is also a board member of numerous Madrona portfolio companies.

Sudip Chakrabarti, Partner, Decibel VC
Sudip is a partner at Decibel where he works with companies that are transforming the enterprise, cloud and infrastructure markets.  Sudip's past investments include Exabeam, Heptio, Serverless.com, Datacoral, Spotnana, Temporal, Tesorio and Polly. He had also been a board observer at Databricks, Digital Ocean, Actifio, Mesosphere and Samsara.

Kellan Carter, Founding Partner, FUSE Ventures
Kellan Carter is a founding partner of FUSE where he focuses on early-stage investments in intelligent software in both horizontal and vertical categories.  He currently sits on the Boards of Zuper, Pictory, Avante, Owl, Xemelgo and PDM Automotive.

Suman Natarajan, Senior Principal, B Capital
Suman Natarajan is a Principal at B Capital’s San Francisco office, where she invests in startups powering the enterprise, across domains such as cybersecurity, cloud infrastructure, DevOps, data and ML/AI, and B2B SaaS.

Brought to you by TiE Seattle
Hosts: Shirish Nadkarni and Gowri Shankar
Producers: Minee Verma and Eesha Jain
YouTube Channel: https://www.youtube.com/@fromstartuptoexitpodcast

Transcript

SPEAKER_00 0:01

Welcome to the Startup to Exit Podcast, where we will bring you world-class entrepreneurs and VCs to share their hard-earned success stories and secrets. This podcast has been brought to you by Thai Seattle. Thai is a global nonprofit that focuses on fostering entrepreneurship. We encourage you to become a Thai member so you can gain access to these great programs. To become a member, please visit www.seattle.tai.org.

SPEAKER_05 0:44

Hello, welcome to Startup Exit. My name is Gary Shankar. I'm very excited about this episode. My co-host is Sherish Natkarni. Sherish and I talked to four prominent VCs for this episode titled 2024 Investment Outlook. They gave their predictions, they gave their outlook, and it is a very interesting one. Shirish and I serve on the board of Thai Seattle, and this podcast is a production of Thai Seattle. It's available on all platforms. I'm very sure you will enjoy this episode as much as Sherish and I enjoyed it talking to the four VCs. Now, on to the episode. Over to you, Sherish.

SPEAKER_09 1:28

Thank you, Gauri. Welcome, Tim. Very very pleased to welcome Tim Porter, who is the managing director at Madrona Ventures, one of the uh premier uh and longstanding uh VC firms out of uh Seattle. They also have a presence uh in San in the Bay Area. Uh Tim joined Madrona in 2006 and invested B2B software companies in the Pacific and Northwest. So welcome, Tim.

SPEAKER_03 1:57

Thanks so much for having me. Great to be with you both.

SPEAKER_09 2:00

Sumanna Frajan is a principal at B Capital's San Francisco office, where she invests in startups powering the enterprise across domains such as cybersecurity, cloud infrastructure, DevOps, ML AI, and B2B SaaS. Uh previously she was a senior vice president at City Ventures and an investor at Intel Capital. So welcome, Suman.

SPEAKER_10 2:24

Thanks for having me.

SPEAKER_09 2:25

Sudeep is an investor and serial entrepreneur. He is currently an investor managing director at Decibel VC. Prior to Decibel VC, he was a partner at Madora Ventures. And prior to that, he was a partner at Andreessen Horowitz and Lightspeed Ventures. So welcome, Sudeep. Kellen Carter is a very experienced VC at Fuse VC, which is a new VC firm that uh spun out of ignition capital. Uh, he is a very experienced investor, has invested in a number of firms like iCertis uh here in Seattle. So welcome, Kellen.

SPEAKER_06 3:08

Thank you for having me. It's an honor to be here and great to see you both two days in a row.

SPEAKER_08 3:14

Thank you.

SPEAKER_09 3:14

So let's start with our first question. Um I'm sure most of our audience is familiar with uh all of you, uh, but it would be great if you can talk a little bit about your firm and your investment thesis.

SPEAKER_03 3:29

So Madrona Venture Group has been around since 1995, and we invest in early stage technology companies, predominantly in the Pacific Northwest. We're currently investing out of two$700 million aggregate funds, so$700 million in aggregate. About 60 to 2 60% to two-thirds are precede to Series A, and about um a third are what we call acceleration stage, which is sort of things series B or C. And um for those later stage companies, we'll look at companies across the country for the early stage. About 80% is in the Pacific Northwest. And so I work with about 14 companies right now. Um two are in Portland, one is in California, and the rest are in Seattle and Bellevue. And we've invested a lot in AI and ML, but we invest across the software stack, uh, applications and infrastructure, and we also invest in consumer software. So happy to get into all of those things in more depth, but that's a little bit about Madrona and the stages and themes that we invest in.

SPEAKER_02 4:30

Yeah. So uh B Capital uh is a multi-stage uh global uh investment firm. Uh half the team is based in the US, half the team is based in Asia. And as a firm, we invest in seed to pre-IPO across a number of sectors: enterprise, uh, healthcare, industrial, uh tech, um, and climate. Um and uh I sit on the enterprise growth team. What sets us apart is um in addition to having our own platform team, we have a partnership with BCG where all of our portfolio companies are uh able to work with partners at BCG across a number of different project uh areas and for go-to-market advisory. Um, and in terms of our investment thesis, especially on the enterprise growth side, we're looking for companies that are starting to see signs of product market fit as early as the series A and B stages, um, all the way to those companies that are at uh pre-IPO. And we are uh you know looking to lead uh in those uh types of opportunities with check sizes ranging from 10 to 20 on the low end all the way to uh 50, uh 50 million plus, um, and averaging about uh 20 to 40 there in the middle. Um yeah, and I'll I'll pause there.

SPEAKER_01 5:50

So our firm is called Decibel, uh, Decibel Partners. We started about four years ago. Uh the thesis behind starting Decibel was very personal. So we all here at Decibel have been founders before, and in general, technical founders before. We started companies, we raised venture money, we had wonderful investors on our boards. But their help was not as impactful because they were more of a generalist and they lacked deep knowledge in our sectors. So when we started Decibel, we really started with specialization because we wanted to be helpful to the founders we are going to back. So at Decibel, we only invest in one area that is enterprise IT infrastructure software. That's all we do. And then we have been able to build a number of different founder services to help our founders go from zero to one. So think of us you know bringing your customers, helping you design your go-to-market machine, help you uh recruit your first 25, 30 people, and also you know, help you tell a great story. So all of those founder services are led by different operating partners here at Decibel, and we bring that to the table when we invest in a company. And finally, we are focused on seed and series A. So nothing is too early for us. Uh and then we are usually investing in companies founded by great technical founders who sometimes you know might have an open source project, sometimes just have an idea on an app-in. And then we invest anywhere from two to ten million dollars. So our current fund raised uh earlier this year is a$285 million fund. So we can go long with our founders as they build these companies. Um I hope that gives you a sense of where we focus and what we do.

SPEAKER_06 7:32

Yeah, uh we started the firm almost four years ago, February 24th, 2020. And we were so blessed to have the prior mentorship and tuition of working at Ignition Partners and then the support of the partners at Ignition to help us launch Fuse. But we kind of joke that we've crammed 40 years of experience into four years between starting the firm right before COVID, really excited, then figuring out how to raise money over formats like this, Zoom, Google Meets, et cetera, and then managing a firm throughout one of the biggest tech resets we've seen since the GFC in the bubble. And so it's been a lot of fun. It's been a lot of hard work, but so blessed to work with great partners and uh great founders and great LPs at Fuse. But the whole premise of starting Fuse was that Seattle is emerging as one of the most important software markets in the world. Outside of this new shiny object called AI, cloud was one of the most important technology things. Well, having Microsoft and Amazon being both based in Seattle plus Google Cloud, plus Oracle Cloud, Seattle is in position A of being one of the most important technology cities in the world. And now with this whole AI momentum that we're seeing, it's also in position A. And so our thesis in starting the firm in early 2020 was that there are going to be a lot of very promising companies and founders starting exciting businesses here. And that by building a firm with our ages and our longevity of wanting to go do this for the next 30 years, we would be the right time, right place, right city. Uh, it was all coming together. And so we started the firm and we had a thesis of uh of in building a firm that, you know, why fuse? How do you have edge? How do you reverse engineer really good returns by building the best product for founders? And so we built this LP community, 300 strong of executives of every single company you can name here in the Pacific Northwest. And the trick is that we engage our LPs to be supportive to our founders. And one of our LPs described it as the best form of collusion he've ever seen. But the whole purpose is bringing the best product to founders in the form of talent, in the form of customers, in the form of advisors, so we can be part of those companies' journeys and be one of the most supportive partners. Uh, we also founded the firm on the premise that enterprise software is one of the most important categories. And then that's what we're going to be investing in. All B2B software, all vertical, and that includes vertical software, includes horizontal enterprise software. And yes, both of those include AI. And been really, really fortunate that was able to raise fund one of 173 million and just close our second fund of 254 million.

SPEAKER_08 10:22

Great.

SPEAKER_09 10:22

Thank you. So um, you know, as you know, uh 2022 was a banner year for uh investing and uh startups. Um I'm wondering what was 2023 like from an investment perspective compared to uh 2022, and what was the pace of your investments uh in uh 2023?

SPEAKER_03 10:43

Yeah, absolutely. You know, and I gotta say, I could just say it up front, it's it's so much fun to be on with both of you. And uh I've been at Madrona now since 2006, so 17 years, seeing a lot of upcycles and down cycles, but uh the two of you were two of the very first people I met in the startup ecosystem in Seattle. Shiri, she won't even remember this, but I was in my summer of business school in 2002, um I was asking Greg Gottesman, who's at Madrona at the time, about any companies I should try to talk to, and he introduced me to you when you were running Team On Systems that you successfully sold to Rim. And then when I was at Madrona and what was it, 2006-7, one of the first companies I got to work with with our founder, Tom Alberg, was Single Point that Gowry was running. So I'd like to thank you both for all the different things you've done for the ecosystem. I've learned a lot from both of you, Gowrie, we went through a lot of things together. And so now as we're reflecting back on the current cycle and 22 versus 23, it's just fun to think about that history and and say thanks to you both. The um so 23, um, you know, it's interesting. I'd say for 2023, the overall pace for early stage was pretty consistent with where it's been for us uh in recent years. We continuing to invest. We were investing at a at a pretty rapid uh or at least consistent uh pace. There was a big flavor around AI and ML companies. So probably, you know, half to two-thirds of the new investments that we made in the last year, you know, were either in AI or in applied AI. And it was definitely an early stage focus. So I think that is also representative of the broader VC market. Um, you know, early stage AI was hot and everything else kind of was not. And so um, you know, that's the pithy, you know, summary of it. The growth stage deals, um, as has been well documented, you know, in both 22 and 23 were very, very slow. Um and so I mentioned at the beginning we have an early stage fund and acceleration stage fund. Our early stage fund is basically we raised it and started investing in September of 22. So we're about 15 months in. And that is basically on pace in terms of what would be a typical historic, it's our fund nine, so historical investment pace, sort of, you know, 12, 15 new investments a year. And our acceleration fund is a little bit slower than would have been paid. And I think we've made an investment. I think we have now we'll have two. Um, but that's a little bit slower than would be typical. And I think that's representative of the broader market, which, you know, happy to dive into the dynamics around that more as well.

SPEAKER_10 13:22

Yeah.

SPEAKER_02 13:23

23 for us, especially on the growth side, was much slower, uh, especially compared to 22. Um, the you know, we saw a lot of companies that um were still had uh quite sizable valuations uh at the early stages, but were still catching up on the growth side to kind of meet a new multiple uh requirement that uh we kind of saw as an industry driven also largely by the interest rates in public markets. Um and therefore we were being much more selective. Um and in fact, uh, you know, we're we're much less active uh overall uh in 23 than uh than we were in the previous year.

SPEAKER_01 14:08

So, first of all, I think 2023 was really a tail of two cities in some ways. Uh what I mean by that is for the mid to later stage companies, I think 23 was a really rough year, particularly if you haven't quite figured out the fundamentals of your business to raise around while you are sitting on a valuation that was kind of inflated from late 2020 to 21. It was hard because you are sitting on an inflated valuation, your business is still early, and money definitely dried up. So a lot of our friends in growth stage venture funds hardly did you know more than one or two investments this year. On the early side, it was a very different picture because I think a lot of that was because of old excitement we are seeing around you know AI, AI powered companies. So it was a it has been a crazy pace uh in 23 on the early side because a lot of great founders who were sitting um on the sidelines thinking about maybe they will start a company someday. Suddenly they saw this amazing tectonic shift coming in, and they were, they didn't need any more conviction. So they jumped into the game and ended up starting companies. And we saw that happen again and again, these amazing teams coming out of all these amazing companies, big or small, and really starting their companies and getting a lot of investor interest. So you we saw like crazy activity on the early side as a result, some crazy evaluations. And a lot of the latest stage venture funds also came up to do more early stage uh investments because they were not doing, you know, they were not being as active on the growth side. Hence, you know, the early early sector, early stage market has been really crazy frothy, you know, mostly in a good way. So that's really what 23 has been for us. Um I would say one other thing, and this is more personal and related to decibel. Like if you had asked us back in 2022, I think there was a little bit of a concern around the table about some of the infrastructure software market segments being kind of mature or done. So we were seeing a lot of companies in certain segments that were almost like splitting value in terms of uh uh splitting hair in terms of what value proposition they brought to the table. Whereas I think in 2023, certainly because of the new AI stack, the new technology, a lot of those segments that were mature are again up for grabs. Because now you can go and displace a lot of the incumbents because you now have a very fundamentally different tech stack that can do things very differently, can offer a very different cost model, and probably a different distribution as well. But certainly a lot of the segments that we thought were kind of mature, were done, had incumbents, are up for grabs. I think that has created a lot of opportunities around the table, particularly on the early side.

SPEAKER_06 17:17

Well, as you might imagine, things slowed down. And it takes a while for the NASDAQ and then the growth equity firms, then the mid-stage firms, and then finally the earlier stage firms like Fuse to feel the impact of what that new rate environment meant. But now it in 2023, it really started to make its way into early stage software companies. And the reality is as capital just gets more expensive. And one of the things I'm most pleased with as it relates to our founders and the portfolio is their understanding their response to a higher rate environment and knowing what it takes to now really exceed the high watermark of the last round and build an enduring company. But the reality is rates go up, capital is more expensive, which means you have to put more proof points on the board. But what that meant for new companies is that a lot of founders just went into bunker mode in 2022 at the end of 2022 and at the beginning of 2023, knowing they needed to go put more points on the scoreboard before going out to market. So now what we're then seeing is companies with much more meat on the bone, more customers, more revenue, a tighter reporting of the KPIs and financials. Uh, the companies are just a little bit more mature relative to where they were in 2022 and 2021, but still going out and raising the same amount of capital of those. And so I think you just see better businesses get built because capital is more expensive. And the reality is founders don't want to dilute themselves to zero. You end up funding the business with with revenue. And so uh I think you'll see a lot of really big businesses get built in this era. There's obviously this big platform shift that's now happening, but because capital is more expensive, founders are getting their business models much more dialed earlier in their journey.

SPEAKER_09 19:14

Great. Um, so the other thing was in 2022, uh valuations uh were sky high. A lot of companies even got you know 50 to 75 times uh revenue, which is crazy to think about. Uh, how were valuations impacted in 2023?

SPEAKER_03 19:31

Yeah, valuations um rationalized and came down to some degree. You know, valuations actually peaked into the public market peaked on December 15th, 2021. And so 22, you started to see, you know, decline, particularly in these late stage, you know, investments that really slowed down. Those really high multiple, high dollar amount, high valuation, you know, slowed considerably through 2022, and you continue to see that in 2023. As I just was saying, almost very, very few sort of late stage deals happened. On the early stage, I would say in 2023, you saw a return to what I would call more a normal level and pace of diligence. There was less of the, hey, we just met, this is hot. You know, the next day send a term sheet. I think valuations for things in hot areas like ML and AI, I wouldn't say they, you know, they were still robust uh in in many cases, but I'd say the overall it came, you know, down to you know, more traditional levels, although I certainly wouldn't say low. Um so we we saw them dip, uh, but certain sectors continue to be hot.

SPEAKER_02 20:43

Valuations were absolutely effective. So um the as we saw multiple compression in the public markets, the lower multiples, um, which is basically a multiple of revenue, uh, which was kind of our kind of determinant for valuation, we saw that uh drop considerably. Whereas in 22, we could see multiples that were you know 50, uh 25 to 50 plus uh X on uh NTM, our next 12 month revenue, um, or ARR. This year we're you know, or last year we were seeing much lower, uh, you know, 10 to 15 or even lower was kind of uh the the compression. And so you can see as companies, if they didn't hit that revenue mark, uh the you know, the multiples and the valuations uh consequentially were much lower as a result.

SPEAKER_01 21:37

Yeah, uh that's exactly the case in terms of valuation. So early saw on the early stage side, valuations were pretty rich in general, particularly if you have if you're coming out uh of a company that you know has been building in the AI space, has been in the LLM space, and you Have a strong technical background in that, I think you pretty much commanded really premium valuations. No question about it. Like there were companies that we got excited about, you know, eventually got done by someone else because the valuation was way too high for our own taste. And when I say way too high, I mean like a cold start seed company, you know, raising at 65, 70 pre. So which is hard to do, honestly. I would say the other thing that also contributed to high valuations is the size of the rounds. So this is actually really interesting. And I think for the first time we are probably seeing this. Like we didn't see this with the shift that happened with internet. We didn't see this with the shift that happened with mobile. But now in AI, I think there is a fundamentally interesting trend we are seeing, which is like the world is getting divided into two buckets. One that we internally call GPU rich or GPU hungry. These are companies that need to raise a lot of money to even get the compute they need before they even roll out their first product. And then there is a second bucket of companies that are GPU poor. They probably don't need to raise nearly as much. I think the GPU rich companies, you know, what we have seen in 2023, have raised humongous amount of money, you know,$35 million seed,$55 million seed. So that is also very unique in 23. And I don't think we saw that before with internet and mobile.

SPEAKER_06 23:31

Anything in AI infrastructure, if you're an elite founder spinning out of Google or Microsoft or Amazon or OpenAI, it's really hard to diagnose what the frameworks are for valuations. I mean, things are off the charts and it's exciting, but there's still going to be a lot of dead bodies in the next couple of years. The market is rewarding high quality business models. And we have our founder 2024 outlook actually for our founders next week. And if you look at the top decile of public stats companies, they still trade at 15x. If you look at the bottom core tile, as it relates to the rule of 40, they trade at like two and a half X. So the market is really still rewarding companies with a high degree of operating leverage where they can grow fast and capture market share, or they can quickly be cash flow positive. And the ones that aren't are getting highly penalized. The other companies that are getting highly penalized is calling something that's not software software, the market is now calling it is what it is. So services businesses trying to market themselves to software businesses, that's gone away. That's totally gone away.

SPEAKER_09 24:44

Great. Over to you, Gonry.

SPEAKER_05 24:47

Oh, thanks. Um, let's uh dive a little deeper into your uh AI investments, right? So of the investments you made, what were the broad thesis as a firm did you guys have making that investment in 2023?

SPEAKER_03 25:03

Yeah. So we we think about the the Gen AI or applied AI almost as a stack. I mean, it's you know when we're evaluating individual companies, of course, it's about what's the customer pain point, what's the customer pull, what's the market opportunity. When we segment this, we sometimes think about it as more of a technology stack and where are the how do you address the customer problems across that stack? So at the top of the stack, it would be called intelligent applications. And this has been the biggest area of investment. I think it'll be continue to be the biggest area of investment. And these could be um areas in you know vertical industries, it could be line of reinventing line of business applications, um, places where you're using machine learning and AI to create you know new and differentiated user experiences using your own data and third-party data to create new insights. And so that that's been the biggest theme. You know, we've done this conference last few years called the Intelligent Applications Summit, and we had a huge group of interesting companies and people brought together to talk about that. In the layer just below, we call AI platform services. Um these are essentially tools and middleware that you know that help people um create these applications, orchestrate the applications, um, include you know things like security, um, etc. Another big area that fuse that infuses all of these um intelligent applications uh and ML models is data. So the modern data stack and different ways of doing data management continues to be a big scheme, a big theme and would be sort of the next layer in the stack. And then the place we haven't made any direct investments is in sort of the models themselves. You know, so obviously OpenAI ushered in, you know, this current age of Gen AI and large language models, et cetera. Um we haven't invested directly into model companies uh that were just providing horizontal model service. But what we really believe in is that the the proliferation of open source models, smaller models, fine-tuned specialized models for company's own data, et cetera, we'll continue to see lots and lots of innovation there. And at the bottom of the stack around physical infrastructure, you know, GPU companies, GPU clouds, those have been they've received a lot of funding. There's this chip shortage, et cetera. We haven't played in that area. We haven't made any semiconductor investments, it's a little bit out of our uh expertise and sweet spot. So I'll just as an example, the companies that I invested in personally in the couple of deals that I led in the last year or so, you know, one was a company called Runway ML. This is the a company that is um video and image generation from text. They actually have built some of their own models. They were part of the original group that created um stable diffusion. But it's really an integrated application, you know, that creators uh can go in and um prototype and create movie scenes, create commercials, create uh sales videos, all the reasons you'd want to create a video and create very powerful video creation and editing tools in the hands of uh of a of an individual user. And then another company is called Number Station, and they're think of them as next generation uh BI and analytics, where you can chat your own data. And this is a team out of the Stanford AI lab, and they use uh open source LLMs that they will fine-tune and train on your company's own data, so all the privacy and control of your own data, but then allow uh really flexible natural language-based uh querying and data exploration to go just directly to insights from your data. So sorry for the long-winded answer, but that a bit of a flavor of the type of things that we're doing in general and the ones that I've specifically been working on.

SPEAKER_02 28:47

Yeah. Um we we were really slow on uh on deals overall. There was a deal that in the AI infrastructure space that's not been uh announced. Um, but uh our thesis was really around looking at um kind of proprietary technology. Right now, the foundation model companies are have so much momentum and runway, and their models are only getting uh better. And so we're seeing that opportunity to for value creation is um is a little bit lower. A lot of we saw a lot of companies come up that were just wrappers around kind of the main foundation models. And so we were really focused on what was kind of the unique technology or moat that uh a company could bring to bear that was building uh AI or enabling or uh uh securing uh around AI. Those were kind of the kind of peripheral and enablement areas that we were really focused on. And those continue to be kind of our core focus areas uh going forward as well.

SPEAKER_01 29:52

And I'll give you a little more color uh because maybe you know one of the things that will be interesting to your listeners is what are some of the broad thematic areas you know, someone like us, you know, um have been focusing on. So for us, um, and this was true for 23, and this will be true going forward. I would say broadly, there were three thematic areas for us. So, first one, which is natural for us, is the infrastructure stack. So the whole infrastructure stack obviously is getting almost reinvented in some ways uh because of AI workloads. I think you know, even some of the cloud computing stack we know uh today is probably going to get reinvented. So we definitely have paid a lot of attention to that. I mean, obviously we all know you know about vector databases and you know, all the infrastructure you need for rags and all of that. I think you know it fundamentally requires different kinds of stack, and we have paid attention to that. We haven't made a ton of investments in that area just yet. We have had made uh at least one uh that I can think squarely fits in that bucket. But I'll tell you why, because it feels still, it still feels a little early in terms of what what that stack might look like when it is settled. So I would say we are paying a lot of attention to things that are baking out and ended up making one investment. But you could expect us to do more in the future as the tech stack settles or standardizes around certain uh components. So that's one big thematic area for us, infrastructure stack. The second one for us I kind of touched on earlier, uh, which was okay, like you have this in a different existing infrastructure software market segments. How can those be disrupted with AI-powered uh solutions? And I think that will that is going to be a big focus for us in terms of investments. Like we have made investments. Um, there's actually a company in Seattle called DropZone that we uh led an investment in. And they are doing security, uh, think of them as SOC analysts in a box. And that whole space was kind of done, but now suddenly is up for grabs because with AI and LLM, you can offer a very different capability, very different cost model. So you'll see us doing that in different infrastructure software segments again and again. So that's number two for us. And the third one, I think we are very early in the cycle, but we actually have already made a couple of uh uh very interesting investments there. And this is something uh I think we are personally very excited about. So we have this internal joke that we are going from a SaaS world to a SaaS world. Let me explain. So when I say we are going from a SaaS world, I mean software as a service. And then we are going to a uh SaaS world, I mean service as software. And this is something uh we actually have even blogged about a couple of weeks ago. And we believe that a lot of the uh roles or jobs are going to be significantly automated with human supervision, but still significantly automated. The mundane parts of the uh job of a human are going to be automated. And we'll see a lot of these companies that might look like a services company from the outside, but is powered by software largely inside. And I think you know, we'll see that quite a bit. So we have made uh one investment that is think of it as a more like a you know financial research analyst uh in a box, again, with human supervision. Uh, we have made another investment that is like think of it as an IT services in a company uh in a box, uh, but again, powered by AI. So I think you'll see a lot of uh those investments from us uh coming forward, going forward.

SPEAKER_06 33:43

Well, we've been investing in AI for quite a long time, even dating back to my days at admission, where we're lending investments in a company called iService, which is emerging as one of the leaders in enterprise software, enterprise contract lifecycle management company. And the fun fact about iService, we closed that investment leap day of 2016. So the company's really like two and a half years old or year and a half years old. Um but the uh the premise about that investment was there's a ton of data inside of contracts. And yes, it's enterprise contract management, but the insights using AI and machine learning that you can extrapolate out of those contracts was the longer-term play and vision that we had at Fuse and that the founders had at iService. And so I just want to use that example that this AI shift is really not that new. It's just a lot more prolific in the market, and it's in every pitch deck. But where you're seeing us invest in AI today is in a lot of vertical-oriented solutions. AI is part of the value delivered in generally what is vertical markets. That's an area that we really like. And a great example of that in Fund2 that we made an investment in earlier this year is called Quandry. And they're doing RPA for the insurance brokerage industry, helping with dock automation, renewal automation, uh, extracting data in the insurance documents and making sure that data is synced with the existing systems. But it's very vertical specific, which makes it harder for one of these broader platform LLMs like Anthropic or OpenAI to do anything because it requires domain-specific knowledge, domain-specific workflows, domain-specific data. So we really like uh vertical solutions with AI is a big part of the is a big part of the component. We did just make an investment that was leaked to the press in a company called GripTape AI, ex-Amazon uh executive, third time he's founded a company. And what they are uh developing is software for instrumentation of AI inside the enterprise, enterprise grade. The reality is, in most AI scenarios right now, all the usage is done with SMBs or in innovationslash curiosity budgets inside the enterprise. It's not really becoming trusted infrastructure. And if you look out there, many of these companies, especially banks, are essentially eliminating anyone's ability to use Chat GPT or OpenAI because they're concerns about what people are doing with the data. And of course, you don't know what some software algorithm might do if it hallucinates. And so there's been a lot of restrictions. But what Kyle and the GripTape team are now doing is helping bring AI models into the enterprise with appropriate and proportional guardrails. Who can do what, with what access to data, and with what models with a higher degree of accuracy with the specific model. And so it's essentially helping bridge that trust gap for real business scenarios that just doesn't exist today. So that's one example of a superstellar founder spinning out of an elite company here, Amazon, and helping bring AI into the enterprise in a highly thoughtful way.

SPEAKER_05 37:10

So thanks for breaking down the stack um so nicely, Tim. So uh this uh when you take at the bottom, the GPU companies where you have not yet made an investment, and then the LLM companies, right? So where you are focusing, seems like there's two uh thesis uh emerging. One is around modeling, and the other is around uh intelligent applications. Then there's this GPU hungry and the maybe the GPU light footprint. Are there theses that that of the of the pictures that you've seen, are the theses that you are uh seeing or you have formed and say this is our outlook and focus for 2024? When you when you look at your stack and what the market's dictating or would buy.

SPEAKER_03 38:00

Yeah, and you know, venture capital and in Madrona strategy specifically is all about trying to have a prepared mind and think about some of these themes and you know, understanding or trying to think about where the puck is going to go, but then also trying to find the best entrepreneurs. And especially for us, find the best entrepreneurs in the Pacific Northwest, and especially at the early stage, you know, we we want to find founders that just impress the heck out of us and how they you know understand a customer and understand a market. And so everything we do doesn't have to be neatly within these investment themes that we spend a lot of time thinking about. That being said, you know, we think the themes are pretty important, and a lot of times, you know, investments happen where you know themes and founders come together. But I just want to make that point that we're not purely just you know trying to you know sit here and think that we can predict everything. You know, it's all about founders and the founders, you know, hustle and grit to go solve customer problems. So with that caveat, I the answer is still you know yes. I and I'd say or I would frame it slightly differently, you know, that these intelligent applications, as you said, biggest theme. The other piece I would just I would call kind of tools and platform services, right, versus modeling companies. And I it it helps you with some of the the models and using them. So but I we use you know slightly different terms there. Um I would say that on the on the intelligent app side, this one is just has the broadest surface area, right? And uh and most often we're gonna be um looking for founders who have an insight into a market and a customer affinity, etc. And so we think vertical applications, you know, whether it's AI tools for e-commerce, AI tools for you know industries from financial services to insurance to um construction to you you you name it, you know, things that we haven't thought of. There's gonna be lots of opportunities. We'll love to back those kind of founders. Um we have thought quite a bit about um, you know, line of business applications, you know, legacy applications, where Gen AI or new data sets or using old data sets in novel ways might create, you know, opportunities. So we sort of have, I won't bore you with going through all we sort of systematically think through the different parts of the business, you know, from office of the CFO to, you know, we've already made investments in the in the sales stack, we've made investments in the legal stack and you know, in-house legal. So continuing to think through those. Um we continue to think productivity apps, you know, for for like the information worker that there's, you know, from personal agents that help you do various things, you know, in the workplace is gonna be a big ongoing area. We've that's always been an area of investment for us from companies like you know, Smartsheet, you know, 12 years ago to Coda more recently to what's going to be the new thing, you know, in productivity in the future work. So we think about that area a lot. Um and then we think about automation broadly. You know, we were investors in UiPath and you know, Soma spent a lot of time with UiPath, and Ted Cummert is back at Madrona having been there. And and what are all the different places where the the paradigm of something like RPA, what's the next gen of that? And how can we create, you know, bring more, you know, really strong ROI-based automation into different businesses? And it's often about human augmentation, you know, not human replacement. So one of the companies I work with is called Outbound AI, and they help revenue cycle workers in healthcare uh more effectively um manage and navigate claims, explanations of benefits, eligibility of benefits and verification, et cetera. This is really manual phone-based work, and most so how can AI help those workers get a lot more efficient and uh do their job more effectively? So those are some areas in intelligent apps. Um I could keep rolling on the platform services and tools, but I don't want I don't want to talk past the question.

SPEAKER_05 41:48

Good. Uh so 2024, uh uh from a macro perspective, there's the economy, there's the elections, and then there's uh, you know, the wars. How has that brought into uh or how do you think that'll affect uh A uh VC investments and and because you're a later stage VC yourself, how will that affect your investment thesis?

SPEAKER_02 42:17

Yeah, it is um look, I think it gets kind of absorbed into the broader macro market uh environment. I would say I don't think those are um areas that are directly impacting an investment on a case-by-case uh basis necessarily. Um so from a thesis standpoint, in terms of the domains that we're focused on, it does not affect uh much materially. Um, in terms of the you know broader um implications of all of those different events, uh, you know, it's creating, it's already been creating a ton of volatility in the market uh overall, um, which means that we have to be even more um careful when scrutinizing opportunities, uh, just to understand if what we're seeing is um a spike that is going to actually be durable and enduring, um, and is not just a result of some um kind of temporary micro uh opportunity that might have been created based on the broader macro. Um so I think we're still really focused on um, you know, fundamental drivers. You know, what are what are enterprises uh looking for? What are their key priorities? What are their budgets? Where is consolidation um you know in favor versus a best of breed opportunity that might lend itself to a startup uh opportunity kind of gaining a foothold uh and growing from there? I think those are still kind of the the main drivers. And as those macro trends impact though that decision making at the enterprise level. That's how it kind of uh flows downstream into kind of the investment decisions that we'll make. Um, but we're not necessarily building our theses around necessarily those uh like around the election or the wars uh per se. Yeah. We see there are a number of companies that have raised a tremendous amount of capital. And look, 90, 80 to 90% of that is just going to GPUs. It's going straight to uh NVIDIA's pocket. NVIDIA. Uh so that's something that we're trying to be really careful about. Because I think what we want to understand is how can these companies build something that is sustainable uh over time? And what is kind of that ROI uh for that model? Um, because uh in a sense, you are kind of putting uh forward this huge capex, um, but some of these uh companies are still like pretty formative in terms of the in terms of what they're building and are treated are more so like seed companies in terms of you know the stage at which their ideas and products have been able to be really tested uh against the market. Um so we're really focused on uh you know how are those companies who are building able to still solve problems and not just uh focused on kind of the you know the kind of uh the moonshot uh opportunities. I think there are some investors who are great about making those uh moonshot bets. Um but I think our, you know, we take a little bit more kind of risk-adjusted stance where we want to understand how is this actually going to help uh, you know, in the enterprise, uh, you know, drive a huge amount of ROI and um enable organizations to accelerate whatever different function they're um trying to achieve.

SPEAKER_01 45:51

Yeah, so I will be a little um, I would say almost provocative here. And I'll say that in general, by and large, we are we are going to stay away from the GPU rich or the GPU hungry investments. Um and you know what they generally look like is they are kind of building or training their own foundational models for you know, maybe you know, image generation, maybe you know, whatever you know, the use case might be. And to even build your first product, you really need to invest in your GPU compute. So probably half of the round that you raise is literally going to NVIDIA, you know, the next day, uh if you can get it, that is. So I think generally, in general, uh a decibel, we are going to stay away for most part from those kind of investments. Um our focus is really going to be the GPU poor. Um, and that's sort of like you know, all the uh investments we have made in 23 kind of fit in that bucket. I think, you know, even on the infrastructure stack, there is a GPU poor play. I mean, think of it as the picks and shovels play, right? Where you are not on the hook to build the best foundation model, but you probably are the infrastructure that enables others to run open source models, you know, in a fine-tuned proprietary models. So we are big believers in that kind of movement. We keep calling this our small model movement. So there's large models, obviously the LLMs, and then there's the small models. We believe that, you know, yes, open source models are not there yet in terms of you know quality, accuracy, and so on, but they will get there. I mean, they'll the that's probably a safe bet to make. And I we also feel like a lot of the enterprises will want to train or fine-tune their own models based on open source Lama 2 or whatever, you know, the next one that comes through, and then really have control on how they run those. So the infrastructure stack that enables that is a big opportunity for us, and it's it fits the GPU poor um physics.

SPEAKER_06 48:04

Well, there's a couple things, and there's a lot to the question, so I could go on all day long. Um one is you've never seen the incumbents move faster or announce new products faster than Microsoft and Google and Amazon have in the last year. Haven't seen a reaction or a behavior like that from incumbents with vast distribution building and announcing and creating noise in the market, or creating actual real valuable products. And so one of the things that we're constantly talking about that fuse in with our companies is what is the what do the incumbents mean to you in the future? You know, are you part of the product roadmap or are you highly durable and defensible against an announcement from a Microsoft? And so it's beware of the cuments is kind of the theme for 2024. And how do you have defensibility and durability with your customer base as you're going to market? The second piece of that question is people are excited and scared at the same time. It's a fun place to be. And figuring out what are those early scenarios of use that might not be the long game of what you want to build, but you can show demonstrable value with AI, with solutions with your customers and start to build trust. I think the key word, if you're a customer doing anything in AI and a vendor doing anything in AI, is how do you establish and build trust? And trust is earned, it's not granted. And so, what are the early solutions and value you can deliver, albeit knowing there's a longer-term play, to build that trust? I think the reality is there's going to be a lot of mistakes made. There's going to be a lot of scenarios where software did something that you know essentially pisses off the customer. So building trust is a really key element to go to market in 2023 and extending into 2024. The last thing is we're guiding our founders is just assume budgets are still going to be tight. Procurement's kind of the superhero right now, and that sales cycles are going to take longer. And so we're working really shoulder to shoulder with our founders to help make sure these budgets have the appropriate assumptions, the forecasts have the appropriate assumptions, and just making sure we're all sober and staying close to the metal on how enterprises are bringing AI and adopting software, especially in 2024.

SPEAKER_05 50:31

So looking at 2024, it's an election year. The economy is doing what it's uh what it's doing. Uh, with your focus on AI, uh regulation is around the corner in some fashion, right? And then there's uh drama all over uh uh open AI. And uh so given that, what's your outlook for uh the macro uh economic uh thesis around 2024 and how it affects what you guys do?

SPEAKER_03 51:07

Yeah, great. Who knows, right? Actually, in in general, um you know, we when we invest, we invest for the long term and we assume it's gonna be you know eight or ten years, you know, cycle on an investment, and you know, the a fund is 10 or 12 or 14 years. And so we try to keep the bar the same. We try not to, you know, let kind of macroeconomic um crystal ball gazing, you know, impact our investment decisions year in year, and we want to back the kind of founders and trends and markets that are gonna be durable and that figure over the course of that investment we're gonna have some upcycles, we're gonna have some down cycles. And yes, being smart about how you as a company time some of those things and how much you spend and how you push on the gas versus the break, etc., are important, but we try not to get too prescriptive around thinking we can predict the future. That being said, I mean, I think some stuff that we're gonna see in 2024, I think we'll see, I think we'll actually we'll see more um later stage rounds. I think they're gonna start to happen. I think you'll probably see quite a few down rounds, you know, within that. It won't it won't all be. There'll be these companies that raise a lot of money in 2021 um and have been working feverishly to hit their milestones, get efficient, grow into their valuations. You know, some or many have, and you know, and and now that it's time to raise more money, I think they'll be able to do so. And then there'll be, you know, some other ones who haven't, and those will be harder situations. But either way, I think a little bit, 2024 is a year where those companies are gonna have to, you know, come back to market. And so there'll be some really positive pieces of that with more growth financing, but I think there'll be some other companies that don't make it, you know, because of that. So more positive, but some some negative, and we've seen some of that in the last few months already, right? Some companies like hitting the end of the road. I think we'll see more MA. Um, and I think some of that'll be what I call good MA, you know, where it's a win-win outcome for the founders, for investors, and for the company that's acquiring them because they have a strong thesis for how they can go take it to the next level. There hasn't been a lot of that, very little over the last few years, but I think that you'll see um even if the really big acquirers um are dealing with antitrust discussions and things like that, there's maybe a little bit of muting on those. I think there's a whole crop of companies um who are at scale who you know will will start to see some more positive MA including in private equity. And you know, before we turn the record button on here, we were talking about rover, you know, dot com here locally and just uh are in the process of the take private by Blackstone. But I think that's an example of good MA. So I think we'll see some more of that, you know, finally. Um and um I think the IPO markets are open, but it's a really high bar. So I think we'll see a few more IPOs, you know, because there's some really good companies that have been stacked up waiting to go out, but I don't expect exuberance. Uh in the customer side of things, I feel like we've hopefully are past the bottom, you know, of the really big riff uh radical cost cutting, cost rationalization. I don't think it's gonna turn to exuberance where all of a sudden spend, spend, spend. But I think especially in you know strategic areas like MLAI, like cybersecurity, like data, I think you're seeing incrementally, you know, some hiring and some spending. And so I think it's not saying selling's gonna get easy. The selling market has been hard the last couple of years, especially for selling into both businesses and consumers, but I think we'll see some incremental thawing there as well. Got it.

SPEAKER_05 54:36

So so because you are uh making later stage investments, the enterprise market is still vibrant, although the the conversation is all around AI. Uh how has the investment dollars uh and therefore the pace changed because AI is sucking all the oxygen out of the room? As you said, every dollar raised, probably 80 cents now is going to NVIDIA. So how is that and you guys make large bets? So it's not like uh you're making a couple of million dollar bets, right? So that's uh the how do you see specifically enterprise investments in enterprise software change because AI is sucking up so much oxygen from the room?

SPEAKER_10 55:26

Well, I think you know, there are a couple of things.

SPEAKER_02 55:29

One, AI as a category, I think we've seen a flurry of investments. That pace has been slowing, and I don't think it's just because it's the end of the year. I think uh there is a little bit more um scrutiny that the entire industry is putting on uh all of these AI companies that are coming up, just as we're seeing a crunch across um kind of the startup ecosystem broadly uh speaking. But I think what's happening just as quickly is that you know the utility of AI is getting uh you know implemented across all of these different software tools. Um and so being an AI company is almost saying, you know, is almost the same thing as being a software company at this point. And so that I think that demarcation is, you know, becoming less and less uh stark and is kind of blurring around the edges. And so um how we actually categorize, I think these kinds of uh opportunities might slowly kind of converge um over time. Um, AI, I think is going to still remain a very healthy category for investment. Uh that all being said, I don't think it's uh going away at all. We're still very much uh focused on it as well. I think we're just trying to be um very prudent about uh what people are actually trying to solve and if this is a research project versus a product versus a platform versus a business, right? Like we want to be very careful in what we're thinking about. Um and quite frankly, there are a lot of companies that pre uh predated the AI, uh the you know, ChatGPT uh-led AI boom that are now able to incorporate AI and augment their offering considerably as well. And those kinds of companies, especially at the growth stages, can become really interesting as well because now they have, if they can implement it better than the incumbents that they're looking to immediately uh displace, then there's a huge kind of opportunity and uh displacement and kind of grow into that next stage uh as well.

SPEAKER_01 57:37

We actually are kind of bullish about 2024 because uh towards the end of this year, I think we are seeing some budgets open up again. So, and we are seeing that through our portfolio companies. I feel like in 2024, it's it's going to be less about our investment thesis, more about companies that have something to sell. I think they will have a better year. That's our you know, that's my view. Uh, in terms of investment thesis, I think you know what's going to happen, you already see this. I think, you know, particularly on the AI side, I think, you know, we are going to get some realistic lessons. So this year was extremely frothy. Everyone was shipping a demo product, but that last mile has been super hard. Like actually putting AI-powered products in production at a large enough scale is super hard. And I think a lot of people are kind of uh realizing that, particularly on the enterprise side. So I would say in 24, you will see this theme of going from prototype to production. So a lot of focus will be on how do you take that looks very cool in a demo to actually production. And I think our lot of our investment thesis is going to be uh in investing in companies that enable that, like enable that last month. So that's that's sort of like the impact on our investment thesis because we do feel like the budgets will open up for any of this product.

SPEAKER_06 58:57

Well, the election year is always the variable, it's impossible to predict. And it always causes, you know, humans like certainty. They don't like uncertainty. Elections create uncertainty, and the natural natural reaction when there's any uncertainty is to generally pause. And so uh just prepare for enterprises and companies to just move a little bit slower because of this thing called the election, which creates uncertainty. The other thing that I think is really important is I think we're getting close, if not have hit the ceiling of where interest rates will stabilize. But I think for a lot of the excess spending, whether it's the investment and the crazy uh cycle we just went through, plus the government spending, I think rates will have to stay fairly high for a little bit longer. They might not go up, but they're gonna have to stay at a current rate or close to it throughout most of 2024. Do I see us getting back to a 2% interest rate environment or 3% interest rate environment by the end of next year? No. And so preparing for rates to stay where they are, um, for that to, of course, impact the capital markets the way it does, especially private equity, that's a good, safe starting point. Prepare for that, but hope it gets better.

SPEAKER_05 1:00:18

So, given that pieces, how will it affect the pace of your investments uh in 2024? Just as a VC, you guys plan number of uh investments. How is that going to affect your pace of investments? Is it changing, same? What is it?

SPEAKER_06 1:00:35

We're we're staying the course. And a big part of that, Gary, is when we're investing in a company, the journey, if we get it right, and we're super fortunate, is generally longer than the U.S. marriage, the U.S. average length of a marriage. It's like 10 years. We have no idea what the markets are going to look like 10 years from now in 2033. We just have no idea. And I think keeping the steady course uh and the pace of investment is something that we're planning to do throughout 2024. We are very much in business. We will be very active. Uh, we're excited about the uh resourcefulness mixed with the innovation cycle we're in. I think a lot of great companies are gonna be getting built and started in the next few years.

SPEAKER_05 1:01:20

So, quick rapid fire. What will be your focus for uh 2024 from a pure investment perspective? What are you looking for? And you'll say, yeah, that's our bet.

SPEAKER_06 1:01:31

Vertical software enabled by AI. That is something we're really, really keen on. It's where there's gonna be more durability, more pricing defensibility, and you're gonna compete less with the scary incumbents.

SPEAKER_05 1:01:44

So what's the one prediction you'll make for 2024? And we'll come back in 2025 to validate your prediction.

SPEAKER_03 1:01:52

So what's the one prediction you would make that um VC investment dollars and number of deals and MA dollars are all up in 24 versus 23?

SPEAKER_10 1:02:06

Yeah.

SPEAKER_02 1:02:07

My prediction for 24 is um I am pretty optimistic about the market overall. I do think that it's gonna remain tough. Um, but I am uh my prediction is that we're gonna start seeing at the maybe at the very end, and this is a hope just as much as it is a prediction, um, a couple of big enterprise IPOs. I know that everyone has been pushing things out to 25, 26. Um, but you know, I think I think things can get better um sooner.

SPEAKER_01 1:02:42

There will really be a real competitor, and this comes from my own inner bias towards uh open models, open source. Like I spent a lot of my time there. I think there'll be a real competitor in terms of an open source model that can give the open AIs and the anthropecs really run for their money. I don't think that is true today. In 12 months, I think that will come true.

SPEAKER_04 1:03:09

We just wrote it off a prediction, that's all. I come from a land of astrologers, so I love predictions.

SPEAKER_06 1:03:16

Um beware of the incumbents is the theme of 2024.

SPEAKER_05 1:03:23

Great prediction. Uh we we think that uh you it swings back to normalcy, or at least to what it is. Thank you very much for making the time. Appreciate it. It's been a pleasure having you on our podcast. Thank you very much.

SPEAKER_03 1:03:38

Thank you very much. I really appreciate it. Fun to hang out with you guys. Thank you very much.

SPEAKER_08 1:03:41

Thanks so much. Thank you so much. Thank you. Thank you so much, guys.

SPEAKER_06 1:03:45

It was really fun. It's always an honor to have the time with you too.

SPEAKER_05 1:03:50

Take care. Thank you for listening to our podcast from Startup Exit brought to you by DICE Seattle. Assisting in production today are Isha Jain and Mini Verba. Please subscribe to our podcast and rate our podcast wherever you listen to them. Hope you enjoyed it.