Medha Agarwal of Defy joins Nick to discuss AI Revenue-Model Transformation, Building an AI Moat, and The Arrival of Agents. In this episode we cover:
- Revenue Models and Business Models in AI-Driven Vertical Software
- Challenges and Opportunities in Vertical Software
- Defensibility and Market Differentiation
- Founder-Investor Relationships and Early Decision-Making
- Sourcing Top-Tier Talent and Seed Investing Trends
- Future of Work and AI’s Impact
Guest Links:
The host of The Full Ratchet is Nick Moran of New Stack Ventures, a venture capital firm committed to investing in founders outside of the Bay Area.
Want to keep up to date with The Full Ratchet? Follow us on social.
You can learn more about New Stack Ventures by visiting our LinkedIn and Twitter.
0:17
Medha Agarwal joins us today from San Francisco. She’s a General Partner at Defy, an early-stage venture firm investing in exceptional founders at seed and Series A. Medha has invested in companies including Livekit, Whatnot, Synthpop, and Tend. Before joining Defy, Medha worked at Redpoint and Bessemer. She was also named one of Business Insider’s Rising Stars of Venture Capital, recognizing her as a standout investor shaping the future of tech. Medha, welcome to the show!
0:48
Hey, Nick, thanks for having me. Yeah, it’s great to have you. So what did I miss? Give us your kind of your background story in your path to venture Yeah.
0:56
So I started my career at Bain in consulting and but had actually always wanted to be an entrepreneur, so I went down the founder route, initially in my career and and started two companies before I moved over to the side of the table to become an investor, and I’m really grateful for that experience, both in both of Those companies and experiences taught me a lot about company building and having empathy for the seat that the founders we work with and talk to day in and day out, for what they’re doing and how hard that job is. I think it’s the most rewarding job in the world, but also the one of the hardest, if not the hardest there is. What types of companies did you build? One was a scheduling platform, and then the other one was a social, consumer, social,
1:45
interesting. And do those inform kind of your thesis now? Or you do you specialize in in similar spaces or or not? Both were really hard,
1:54
I guess one what? I don’t wish scheduling on anyone. I don’t know if you knew, but there’s one county in Arizona that doesn’t follow daylight savings as an example. So there’s just a lot of nuances and intricacies and what scheduling means. Calendaring means just are very different things to different people.
2:14
So is the short answer? No, you’re not investing in those spaces. I’m
2:19
not doing a lot in in scheduling software these days, but if someone figures it out, they should come find me, because it’s a hard problem to solve.
2:26
Cool so I assume you work with Neil over at defy
2:30
I do, and I know he’s been on your podcast a couple years ago.
2:33
He has, he has. We’ve looked at many deals together over the years, still trying to find one to work on together, but refresh us. You know, what’s the thesis at the firm, and what things are you focused on? Yeah,
2:44
so we’re doing inception through Series A, investing in founders who have a right to win in markets where they can build large, vendor, scale businesses. And we’re currently investing out of our third fund, mostly leading and, you know, pretty hands on, working with our portfolio companies, helping them inflect growth for the next stage.
3:05
Great. And about how many deals do you do per year as a fund, as a fund and in yourself, you know, yeah, as
3:13
a fund, I’d say probably eight to 10. And as an individual, two to three, two
3:20
to three Perfect. Well, I’d love to jump into some things on revenue models and business models. I know that you’ve written a lot about the topic, and you you know, you’ve backed breakout companies across different verticals, live kit, whatnot, synth pop. So, you know, as AI is rewriting pricing and revenue models, where do you think the real money will be made, you know, beyond kind of the traditional SaaS subscription,
3:45
yeah, so this is something I think about a lot. Vertically software has become very popular and exciting these days with the advent of AI, and my personal opinion is that it’s because there’s now an opportunity for verticalized software companies to tap into labor budgets, which are enormous in most of these companies. It’s the largest line item cost line item versus just going after traditional software spend. And so the TAM that we’re talking about, and that is addressable. The market that’s addressable for these companies is is completely different from what it could, could have been, and was five or even 10 years ago. So I think that’s the biggest change in opportunity that that we’re seeing in these companies, and a lot of that is driven by AI, because of what the LeapFrog and what technology can do, and how technology can now, if we look at some of these AI enabled workflows, they are getting pretty good at completing tasks or close to completing tasks end to end, whether they’re very complex ones or even simple ones, I think the opportunity to either replace humans in. Doing entry level work that no one really wants to be doing or and freeing them up to do more intellectually interesting or complicated tasks. Is is bigger now than before, or is bigger is exists in a way that it never has existed before. And I think that’s really, really exciting, and which is why I’m spending a lot of time around this trend. Will
5:21
these revenue models be multi pronged? Is it a few different types of source revenue coming in, or is it one main is it usage based pricing? You know? How do you see that evolving? Do you
5:33
mean, are the pricing models? Are they charging in a different way for different products that they’re selling? What do you mean by that?
5:40
Yeah, are they charging in different ways for different products or for one product? Are they, you know, going to charge in multiple different revenue schemes depending on usage? You know, we’ve seen subscription models that are seat based. We’ve seen subscription models that are volume of usage based. We’ve seen AI models that combine the two you have to pay one or both in some you know, form or fashion. Like, how do you see that evolving with kind of these new products? Yeah.
6:09
So in my mental model, I break it out into four different categories, actually. So the fir fic first is our traditional fixed pricing, which is often per seat or per location, depending on who your customer base is, and that is, was always thought of as the most optimal high value revenue, right? It’s fixed. It’s recurring. It’s predictable. You can your customers. Can predict it. You can forecast your revenue. The Cash Conversion Cycle is very compelling. It’s often contracts that are signed for a year, paid up front. But the drawbacks exist in a particularly in an AI first world, especially if we’re going after Labor spend, is that if we’re making people more efficient, it might mean that use, their usage, will far outpace the growth in people or seats and head count. And so there is the risk, depending on what you’re doing, that you are creating a lot more value than you can capture. And so a company might double triple in revenue, but not in the number of seats. And so the usage has gone through the roof, but you’re not capturing that. So that’s kind of the pro versus a con of the traditional SAS C based pricing, but, but there are a lot of pros. There’s a reason why, historically, in the public markets, SaaS business models have gotten about one and a half to two and a half times revenue multiple premium for those models, because they are more predictable in long when you look in longer time horizons, I think it’s also really helpful to slot in if they already have existing software spend allocated for the specific use case, right? If you already have marketing automation software line item budget, it might be easier to just slot into that same soft software spend. So those are kind of some of the the ways to think about whether that could be the right or the wrong model for you. The second model that I’ve started to see emerge is what I call input based pricing. So these are models that are charging their customers for based on consumption and usage. So for example, how much, how many terabytes of video, how many pages you’re uploading and processing, the number of tokens or queries that are being processed. And I think this is the closest to what traditional API based business is, or usage based pricing has been in the past, where we’ve seen people charge it being charged for the amount of storage they’re using, or compute or API calls. And I think these work really well. It when, when the inputs are fairly predictable and knowable in advance for your customers, so that they can very clearly be able to anticipate the general range of what the costs going to be, and also for you to be able to calculate ROI when you’re making the case for why they should purchase your software and to justify your pricing if it’s going to be extremely variable, it’s really hard for both parties to be able to predict that, and for company, for your customers, to be able to make a business decision on when and if to use your product. When is it worth it? When is it not so maybe that is, let me know if you have any questions, but that that’s how I think about input based prices, pricing. And then actually, maybe one of the most interesting ones that I’ve seen emerge with the advent of AI being applied to vertically software is output based pricing. So we’ve seen companies start charging for outcomes, which is really exciting, to show that that technology can do things like we were talking about. Take, take something end to end. Take a workflow from end to end. So how many units of work did you did the tool produce? Or how many tasks did it complete? How many summaries did it right? Yeah, and I think this pricing model most directly translates to being able to calculate the relative ROI on doing this manually, via human labor versus an AI completing that task. And so for companies that are rare, very, very focused on throughput and where they can quantify the value of the those outputs. This is a really great model. So for example, legal there are full of companies. You’ve probably met some of them that are charging, that are creating medical summaries or or case packets for lawyers, and it’s pretty clear to lawyers what the value of that is for them, and generally how many hours it will take their paralegal and themselves to put that together, and so you can charge for that output. And the ROI is very clear,
10:46
well, it sounds like there’s an established pricing methodology for the output to so that you can figure out how to price it effectively, so that they can measure the the trade off.
10:58
Yeah, it’s it’s very clear how much it would cost for a human to do it. I see, right? We can calculate, hey, on, on average, it takes us 10 hours to put this packet together, and at $30 an hour, that’s $300 and if you were to do it in an automated way, and you only charge me $100 then that’s very clear ROI me and I can free up my paralegal to do stuff that’s much higher value than versus summarizing or copy and pasting and finding information in various documents and copy and pasting it that’s not super high value add and go through their backlog of more interesting work. And then the last one I’ll say, is so a the pro of these usage based so input and output based models is that which is really exciting is when you can tap into labor spend. I think the potential for revenue from any single customer is so much greater than if you were just they were just seeing it as software spend, because the cost savings are so much greater. The downside is it’s much harder to predict, right? I think we saw this in the last cycle with transactional based models in public markets, and what happened to them is, in good times, you see it’s all up into the right and usage increases. But also it’s unpredictable, and very quickly, in a down, in a downturn or downside scenario, usage can, and therefore revenue can, fall off a cliff, and so it’s much harder to predict quarter two quarters out. And so what I’ve seen companies do, in a very thoughtful way, is to create a hybrid approach, where you say, Hey, we’re going to have base platform fee that you have to pay every month, and that includes X amount of usage, so this amount of terabytes uploaded, or this number of case summaries that that we will create for you. And if you do more than that, that’s fine. We’ll, you know, bill you for the overage at the end of the quarter, the month or the quarter. And if that happens consistently, we can talk about the next year and the next year, but it’s more predictability for everyone, which I think is can often be a really smart strategy for people. I
13:04
love it. Love it. Yeah, I was just thinking about, like, one of our own workflows the other day, just internally as a firm, you know, we’re, like, doing diligence. We’re doing research. So we have a, you know, a subscription to chat GPT, for example. So we’re paying a subscription, right? But then we have to do some API calls, you know, with our CRM and stuff to update it. And so we pay for some of that. And then we have Zapier doing some zaps so that I don’t have to ask my associate to copy and paste the market analysis into the document. It just does it automatically. So then we got the ZAP fee right, and then at the end of the day, it’s like, you know, how much does it cost for an associate to write a Deal Memo, you know? And if the final output is the Deal Memo, like, that’s kind of another value stream. So you think about all four revenue models in more of, more of an agent context. I guess if we had a fully functioning agent, I wouldn’t have to stitch together chat, GPT, with the API, with the zaps, you know, with the associate creating the finalized Deal Memo. It’s just done automatically, all at once. And I almost feel like agents are already here, because we’re kind of, I see a lot of my founders stitching together these things, and like the agents are being created as we speak?
14:22
Yeah, totally. Well, first and foremost, I want your process. That sounds super efficient, but I totally agree with you. I think you’re it’s a lot of work for you to even figure out what whether those costs are worth it, and whether some steps are worth it and some are not. And then how do you calculate the value? How long it would take your associate, versus the relative value that that he or she could be doing by doing something else. I think it’s it’s much more complicated. But the if you had one agent that said, hey, for everyone, every Deal Memo I put together, you’re gonna pay me 50 bucks. I don’t know how much you pay your associate and how value. Valuable they are. But my guess is that that would probably be worth it for the other things they would free up that person to do. And so I think that’s where, when you start doing things end to end, you can say, Hey, I just, I just did this whole thing. Quantify. You want to do it on number of hours saved. You want to do it on the value of the work that you freed up that person to do. Either way, it’s probably going to be worth it. So, you know, you’ll be willing to pay me for it. I can imagine this is
15:24
like an interesting topic, because I think the calculus of what you just talked about, like, what’s the ROI? Because it takes, it takes a while to learn the systems and train everything and stitch it together. But I do feel like forcing the mentality into the organization by founders. Like, we’re all going to use this stuff native the compounding advantage that you get out of everyone, like, doing reps every day on this stuff, yeah, the first couple cycles are, like, painful and inefficient, and they probably cost more than they’re worth. But the value you get out of the compounding knowledge of like, Oh, now I’ve got a complex task, but I can whip it out in like 10 minutes because I’ve just, I’m used to using all these tool sets, yeah, and
16:10
if I turn it around on the other side, so that’s, I totally think you’re right for more consumer bottoms up, Product led tools like a chat GBT or Claude or something like that. But for these verticalized software companies, there’s often it’s top down, because you can’t just go pick up synth Bop, which is a portfolio company, someone in healthcare who’s parsing referral packets, can’t use that out of the box there. There have to, they have to do integrations with your EMR and other internal systems, there’s some training on your data and your process for it to be valuable. And I’m seeing that more and more frequently with these verticalized software companies, which I actually think creates defensibility, which is something that we is coming up a lot in this in this category, but it creates defensibility for these companies, but is a lot of work up front, and so we’re seeing more of a forward deploy model that Palantir really made common and as a result, and it’s taken me a while to get back to your to respond to your comment. But the reason why I think that’s really important is it’s even more incumbent on the vendor and these companies to make clear what the ROI can be and look like when, when that product is been in, when it when it’s at its end state. And there is a lot more consultation, I think, that is required in sales processes today than there has been in the past, to help these customers dream the dream, if you will, of what their lives could look like on the other side of this implementation, if they were using this tool, which is a slightly different skill set than just, you know, if we’re going back to marketing, selling, kind of marketing, AB testing software, 100%
17:51
so you gave me the perfect prompt. Let’s jump ahead, like, talk to us about defensibility, rapid, rapidly evolving market. You know, tech tools, you know, we have vertical SaaS, we have vertical AI, you know, what are you seeing some of the best in class examples doing to kind of create an advantage and maintain that differentiation and advantage over time? I think
18:13
it’s a really good question. It’s something I’ve spent a lot of time thinking about, even as I talk to and look at companies in what my mental model should be. And I think I it’s still something I’m thinking about. Because I think there are categories that was kind of, you know, traditional thinking in venture and in startup land, that these are really crowded, it’s really hard to break out. And kind of good luck to you if, if that’s what you’re going to try and go after but in a lot of these categories, there are breakout companies who have really figured out a new product or go to market that’s really resonating with people. I mean, even coding assistance is a great example, like the common thinking was, what engineers ever going to change their IDE Good luck, and that’s not the case anymore. And so I think it’s a super exciting time. I think where I am landing on it, or right now, how I’m thinking about it, is that it’s often the best products combined with the most thoughtful go to market that wins. And it’s not usually one or the other. Sometimes one is more important the other depending on whether the sale is top down or bottoms up. But often what that looks like is not super obvious to me or to you, because we’re not deep, especially in a vertical that you and I are not deep in, what the workflows, motivations and constraints are of the end user is not super obvious, but to a founder that is understands that market and that user and therefore has a right to win in a way that maybe others don’t, and it’s not as obvious, they’ve really figured out what, what it is, what how those people think, and how they can add value. And those are the products and companies that I’ve seen really break out. On, how do you
20:00
strike that balance between product and go to market? You know, I remember years ago talking to investors on the show and talking about founder versus, you know, market, and the answer is, both. Right? You want both, but often you’re evaluating an opportunity. And you know, there’s, there’s some founders that are just excellent at product and go to market, maybe not as good, and vice versa. So like, how do you strike that balance? And how do you work with founding teams? Post investment to, you know, give them the best shot based on their strengths and and weaknesses. Yeah,
20:35
I think, honestly, I think the founder has to have an insight on which one is important, more important in that market, and then we would make the investment if we have the same point of view on the on the market, I think for a lot of these verticalized solutions that are top down, the way we were talking about, where the you have to help someone dream the dream in order to sign them up, because the value is not going to be very clear on day one. And maybe they can see it in the demo, and it looks really cool, but the reality is, and everyone knows that there’s going to be work that’s required to get there go to market is really, really important, and product is important, but only important if you can convince people to sign up for your solution and and invest the time and doing the integration, and then the cus and the and training their models to make it as useful as it can be, whereas, if it’s a bottoms up, if we go back to coding, if it’s a bottoms up, Adoption Model, products probably more important than distribution, because people have to be able to sign up and love it, and then you can kind of figure out how to get it to be viral and share, but if they love it, they’ll share it. How about like,
21:41
top down enterprise motions, where the buyer is many seats away from the user, right? Maybe you’re, I don’t know, selling some operational software, and you sell to the COO and the CO is going to save money, but it’s somebody on the shop floor that’s using a day to day. Like, how do you, how do you think about the sales motion on those models in designing a product that can appeal to both, you know, the buyer and the user,
22:11
I think you have to do both and the and the the company has to be aware of that dynamic I’ve seen, and especially in a world of AI, the world of AI where AI has taken everyone’s imaginations by storm, and there’s so many people who are thinking about how it can transform their lives and their work, and all these budgets have been created to to into for innovation and to test AI top down, where people are trying lots of things, and then they’re only going to keep the things that are really making a difference and users are adopting. I think it’s really important for companies and their go to market teams and implementation teams to really think about what that looks like. Of how do you drive adoption for the folks that are going to be using your tool every day, and also, how do you speak to them in a way that it doesn’t scare them, but shows them what it can do to enhance their lives? Are
23:07
there certain signals or behaviors or modalities that you’ve seen expressed in some of your best product founders? You know, as you think about some of the best products and businesses that you’ve, you’ve worked with, are there some themes that surface
23:24
in terms of how they are selling the solution, or how they building a product,
23:31
in terms of how they’re building the product? Yeah, I think it’s the
23:35
people that really, really understand the motivations and desires and constraints of their end user. I feel like, time and time again, I’ve, I’ve seen products, product decisions, and maybe go to market decisions that are not super obvious to me. I wouldn’t, as someone coming in new to this market, I wouldn’t have been like, oh, you should totally add a, you know, even though you could automate the whole thing, this entire workflow, end to end. You should add a like a confirm or deny or a check mark X mark next to the next to each decision you’re making. And if this is a, you know, maybe simplistic, simplistic example. But I was just talking to one of my founders yesterday, and he was saying, you know, for at first, we were just automating the whole thing, and we were doing it in 30 seconds and and, you know, it was very clear very quickly that that didn’t work for end user. They were freaked out by that. They’re like, we don’t understand why you’re making the decision you’re making, and it’s almost too fast, like we write, if it’s that fast. And so honestly, what we did was we slowed it down a little bit and kind of added some of the thinking right the way, like the foundational models will, when they’re reasoning, they’ll they’ll write out their reasoning, and then we gave them the ability to accept or reject our changes, even though we know that they’re right because we’ve trained enough on their data that we know, because we have a confidence score, we know the. Likelihood that it’s correct. We’re still giving them the right to the ability to do that, and they really liked that, and they really appreciated it, and over a few because it’s an industry that doesn’t have a ton of technology in today that they use, and so over time, as they continue to trust the solution more and more, we’ll automate it away, but on day one, we’re that’s how we’re going to present it to them, and that makes a ton of sense. It’s not something that I would have done outside in in this company operates in some kind of industrial market. It’s not necessarily something that I would that would have been intuitive to me, but they live and breathe the workflows of their end users and and I think that’s really clear in that interaction that we had.
25:40
Are there particular sectors or verticals that you think are sort of primed for transformation with, you know, what we’re seeing in AI, you mentioned legal before, but maybe some others,
25:51
yeah, I honestly, I think what is so exciting about vertical eye software is, you know, you can do the top down analysis in the economy on terms in terms of buckets of spend and industries and the economy and and which are the biggest, and therefore maybe have the potential to have the big, biggest out outcome. But I, I think the most interesting part and fun part of investing in vertical software, to me has been the the opportunity to meet founders who are really subject matter experts in what I would call kind of random verticals and markets that I have no expertise in and haven’t thought about before, but they can dream the dream, and have enough experience with technology in that industry to say, Hey, this is what we could do to transform this use case or wedge product, And this is where we can go from here and and this is why we’re building this, and we have a lot of conviction. And so that kind of dreaming the dream and learning about new verticals by investing in this kind of broader category is what I find so, so fun about it. Maybe if you to answer a question more directly, categories that I think are the most interesting today are verticals where there things, there were a bunch of things you couldn’t do with software before, because the without AI, because the data existed, but it was unstructured or too hard To access. And I think without data, it’s often hard for AI to learn, and if there’s bad data, for it to learn and actually give you good answers. But previously, there were a lot of industries where there was a lot of data, it was just really hard to access. How do you you had to ask humans to give you the data. You had to build a ton of integrations to do it, and it was really hard to structure it, and so the cost to serve and the complication just made it too hard to build a true, truly meaningful and impactful solution in a way that’s possible today, in my opinion. And
27:51
what would be some examples of those? And then to your earlier point, maybe, what are some of the random ones that you’ve you’ve observed, you’ve seen? Yeah,
28:01
so healthcare is a great one. I think it depends on where in healthcare. I think there’s a lot of bad data or incorrect data in healthcare, and so being very thoughtful about garbage in is garbage out. So being really thoughtful about is the data that I’m getting and the data in this market, do I believe it? Is it real? Is it accurate? Is important, but I think in aspects of healthcare, that’s definitely the case. It’s structured. It’s accurate, you just have to figure out a way to access it. Synth pop, a portfolio company of ours, is doing that for back office healthcare tasks. So they’re taking data from from individuals health record that has been structured by a doctor or a nurse or an NP, and they’re but it’s just not in a readable format. It’s in PDFs or faxes or something like that. They’re extracting it, they’re structuring it in a way that’s readable. They’re confirming it with other sources. I think that’s super interesting, and was really, really hard to do before AI and llms. Another example is in supply chain and logistics. I was on the board of a company previously that was trying to create visibility in supply chain, and I still believe in this mission. To this day, I think pre LLM and I meet a lot of companies trying to do this, but pre llms is really, really hard in the amount of effort and cost to get data across emails, in people’s heads, in WeChat, in the in someone’s NetSuite instance, like all over the place, is just so complicated and having to extract the data From a PDF, structure it and input it and compare just really hard. And I think that’s more possible now than ever. And so I’m pretty excited about what’s going to happen in supply chain and logistics. So maybe two broad examples for you.
29:53
That’s great. So you’ve been in the founder seat. You know, I’d love to hear kind of your general approach to where. Working with founders. But then, even more specifically, maybe you know, what are some of the early decisions you see that kind of make or break startups on their journey?
30:10
Yeah, so my approach, I think what I tell founders is, yes, there’s a lot of tactical things that hopefully i and we can help you with. But my number one goal and job is to be someone that you can call and trust and be what I like to call the first phone call because I’ve been your seat. It’s really lonely. The CEO job is really lonely. You only it’s not that obvious until you’ve been in that seat. And so to have someone you can trust and call about things that are not going well, or questions you’re having and insecurities without feeling judged, I think is very important, and especially for people’s mental health, but also to be able to kind of talk things through, and kind of, I was often so deep in the weeds, it was hard to kind of come out of it and think strategically. And so to be able to do that, I hope, is helpful. But then on top of that, I work really hard to add value for my portfolio companies, whether it’s making customer intros or helping recruit candidates, interview them, things like that. I try and be pretty hands on and in terms of the most critical early stage decisions they make to it all comes down to hiring. In my opinion, I think the job of a sales but the job of a CEO and a founder is to be a salesperson. At the end of the day, you’re you’re selling people to join your company before anything exists, when, on a risk adjusted basis, no rational person would actually do that. You’re selling customers to sign up to use a tool that doesn’t really exist or work very well, and to design, to partner with you, to use this kind of hokey thing, and ahead of when it’s going to be super useful and meaningful, and they’re going to have to invest time, and then you’re selling investors to give you money to get to a place that when to dream the dream with you, Ahead of the ahead of any of that existing and so I think being a salesperson is I think sometimes founders don’t always think about it that way, but that’s really the their job at the end of the day. And people, it’s all about people, because ideally, you’ll fire yourself from enough things over time and hire people that are doing better at it than you do. We are better at doing, executing and doing those things than you are. So two mistakes I’ve seen people make is one, not having the bar high enough. Like, you know, a people hire other A or A plus people, and I think B people hire C people, and that can permeate through an organization really quickly. So that’s number one. And then second is speed to hiring. I think the biggest mistake I made, and I got advice on, was we were I remember being like, so overloaded with all the work that I had to do, and someone came to me and said, like, didn’t you have an open, a rec open to hire for X role and and my reaction was, yeah, but I don’t have time to go hire because I’m trying to execute move the business forward. And they were like, well, that’s exactly why you have to drop everything. Go hire that person, so they can take over some of this work, and you can free yourself up to do everything else. And the fact that you haven’t done it yet when you’re feeling the need means you’re six months behind. And in now that’s going to be a create, creating a choke point in your process. And I think that’s exactly right it takes. Can take six months to hire a really good person, convince them to join, wait for them to come on board, and then get them ramped and so hiring before taking the time and taking a step back to hire someone great ahead of you, feeling that need in this moment really severely is really important to not slow down your speed of progress. Do
33:50
you have any insights on how to source you know that top tier talent because often it’s not looking right. So if you just put up a job post, you might not find it. So how do you identify them? And then, you know, start that, that process of recruiting?
34:07
Yeah, it’s a good question. I think the I think your network for the first five employees, I think the most efficient way I’ve seen companies and founders hire is through their network. For that reason, people are usually not looking you have to be able to sell them on doing this, something that’s super irrational, and getting them over that hump is easier when they know you, they believe in you, or they know someone that trusts you and believes in you, and they can, they kind of helps with that first hurdle, and making that leap, of course, there’s lots of other ways to find people, and I’ve seen people successfully do it through recruiters or cold outbound or whatever it is, but on a consistently, I’ve seen that to be the most efficient way to do it. Perfect.
34:54
How do you think a seed investing has changed? You know, in the past couple years? Yeah. And what advice would you give to founders, kind of in the the new raise environment that we’re in? Yeah,
35:07
I think there’s more capital now than there’s ever been, and there’s a lot of opportunity to raise money for a startup if you are, if you are feeling inspired and excited and wanting to devote the next 10 years to a specific problem space. So I think that’s really exciting and really great for innovation. I don’t know that my advice to founders has changed in the last two years. I think it continues to be, maybe what I wish people had told me when I was a founder and I was fundraising, which is, trust your gut and work with people. This is a long term relationship. If you are lucky enough to have options, work with people that are that you feel like you have a good working rapport with, that are going to be able to help add value and and just people that you could trust, because it’s going to be a long, hard journey, and hopefully it’s up into the right all the time. But I I’ve yet to meet a startup and a founder that has said that their company was up into the right from day one. Maytha,
36:07
if we could feature anyone on the show, who do you think we should interview and what topic would you like to hear them speak about?
36:13
So the, I think there’s so many people, but the, funny enough, the first person that come to came to mind is my partner, Brian. He is the expert in growth. In my opinion, I go to him for all questions, to get his thoughts on on all things growth related. So I think his expertise is second to none, and a lot of founders that listen to your podcast could benefit from some of his insights.
36:35
Awesome. Amita, what book, article or video would you recommend to
36:39
listeners? So I heard a I saw a YouTube link for Sam Altman talking about the future of what foundational models will look like and what the future of work will be. I don’t remember the exact title, but it’s fascinating to think about how someone like that, how they think about what the future will be, a question I think about often, and if you have an answer, I’m curious is, what would I suggest my kid be or study to be when they grow up? And what are a couple occupations that I tell them to, you know, stay away from, because I’m worried they’re not going to exist in 20 years. And I have some thoughts, but I don’t know the answer. So if you do, let me know
37:19
I have no answers. That’s why I asked the questions. I can tell you this, my son beat me in chess last night, and he’s seven. Wow, that’s amazing, all right? And he’s not a savant, he’s smart. But I was like, what just happened? It’s probably a little distracted, but all of a sudden, there goes my queen. Yeah, it was crazy. What do you think? I mean, do you have any do you have any thoughts about what the future looks like for our children and future of work? I
37:49
have thoughts, but the world is changing so fast that it’s really hard to know. I’ve had really interesting conversations with people with you know, will software engineering look like the way it does today? I don’t think so. But will all software engineering go away? I also don’t think so. Will college still exist? Does the role that college play change over time as the labor force and the distribution of what people do changes? I think they’re all really valid questions. If 10 years ago you told me that those are questions that I was really debating in earnest, I would have laughed and thought you were crazy, but the world’s changing so fast, and so I think it’s an important thing to keep thinking about and reevaluating awesome. Mitha,
38:31
what book, article or video would you recommend? Oh, we did that. I’m sorry. Mitha, do you have any habits, tactics or behaviors that are a force multiplier?
38:40
Well, I hope this is but in getting to know companies, I try to add value along the way. I’m very as someone who’s been a founder before, I’m very aware that I don’t want to just be asking for their time as we’re building a relationship. And so whether it’s customer intros or higher intros for potential hires, I try and make some of those connections while I’m getting to know companies ahead of a raise or during a raise. It it helps me. I don’t think there’s a better way to diligence a business and get to know the market, but it also hopefully helps the company. And at a minimum, even if we don’t invest they’ve signed up a new customer.
39:17
Awesome. And then finally, here, what is the best way for listeners to connect with you and follow along with defy email
39:22
LinkedIn, sign up for you can follow me on on x as well. Awesome.
39:27
She’s Metha Agarwal, and the firm is defy Metha. Thanks so much for the time and the insights. Thanks. Was a lot of fun.
39:33
Super fun. Thanks for having me. Awesome. Thank
39:40
you. All right, that’ll wrap up today’s interview. If you enjoyed the episode or a previous one, let the guests know about it. Share your thoughts on social or shoot them an email. Let them know what particularly resonated with you. I can’t tell you how much I appreciate that some of the smartest folks in venture are willing to take the. Time and share their insights with us. If you feel the same, a compliment goes a long way. Okay, that’s a wrap for today. Until next time, remember to over, prepare, choose carefully and invest confidently. Thanks so much for listening.
