.jpg)
A Product Market Fit Show | Startup Podcast for Founders
Every founder has 1 goal: find product-market fit. We interview the world's most successful startup founders on the 0 to 1 part of their journeys. We've had the founders of Reddit, Gusto, Rappi, Glean, Cohere, Huntress, ID.me and many more.
We go deep with entrepreneurs & VCs to provide detailed examples you can steal. Our goal is to understand product-market fit better than anyone on the planet.
Rated one of the world's top startup podcasts.
A Product Market Fit Show | Startup Podcast for Founders
Why new Carta data shows bridge rounds might be worse than you think. | Peter Walker, Head of Insights at Carta
Carta just released their report for Q4 2024. Peter is Head of Insights at Carta, and the person who owns their data practice. We sit down to talk about the largest trends he saw across fundraising, industries, graduation rates and even hiring practices.
Carta data shows that graduation rates from Seed to A are much lower for companies that have raised a bridge round. We analyze why that might be and what that could mean for early-stage founders.
VCs read and understand all this data. If you want to operate on equal footing— you should too.
Why you should listen:
- The role of AI in the venture capital landscape.
- Why there are a trend of larger funding rounds going to fewer companies.
- Why so much capital is being allocated to AI companies.
- Valuations for seed and early-stage companies are on the rise.
- Why bridges and extensions have become so popular.
- Why bridge rounds have lower graduation rates to Series A.
- What the data shows about how hiring practices are changing.
Keywords
venture capital, AI, fundraising, market trends, valuations, startup ecosystem, early stage, late stage, investment, venture capital, bridge rounds, seed extensions, startup growth, hiring practices, AI impact, early stage funding, market trends, valuations, exits
Pablo Srugo (00:00):
Just yesterday, Carta released their Q4 numbers for all of Venture for 2024 and actually for all of 24, kind of like the annual report. So we sit down with Peter Walker, the head of Insights, who leads the entire effort and practices at Carta of understanding the data better than really anybody in the world when it comes to venture because they have ground level data about cap tables across all of the US and globally of what's exactly happening. And this is private market, so it's always hard to get true data, but Cartas data frankly is the closest you can get to truth and not just what gets reported by headlines and TechCrunch and all these sorts of things. And we have the first live conversation about what happened through 2024, what happened in that final quarter? Where are the trends going, where are valuations at? Where are round sizes at? What's happening with team sizes? We talk about how teams are actually getting smaller even though fundraising levels are up. What are people doing with that money? This is trends that you as a founder have to understand because believe me, VCs are looking at this stuff and so if you want to be on balance with them, this is stuff you need to know.
Previous Guests (01:03):
That's product market fit, product market fit, product market fit. I call that the product market fit question, product market fit, product market fit product market fit product market fit. I mean the name of the show is product market fit.
Pablo Srugo (01:09):
Think back to the last few months, the last few years as you've been running the startup, how many different founders have helped you out? The reality is founders help each other out. That's just who founders are. They pay it forward. So help a founder out. Take literally five seconds, take your phone out of your pocket and hit five stars.
Peter, welcome back on the show, man.
Peter Walker (1:33)
Thank you so much for having me back. Appreciate it.
Pablo Srugo (1:35)
Well, looking forward to getting your takes on what's going on in Q4. I mean, frankly, as a VC and just somewhat in the startup world, the only thing I hear about is AI and more AI and bigger rounds and even bigger rounds and then a bunch of deepseek stuff, but mainly just AI and big AI rounds. So I'm curious what you're seeing in the data. Hopefully there's other things than just that.
Peter Walker (01:59):
Nope, that's it, man. That's all we got
Pablo Srugo(02:03):
<laughs> Done. All right, episode’s done, man.
Peter Walker (2:05)
Yeah, I mean this was easy. No, look, there are other things going on, although I'm not surprised that AI is dominating the landscape. I think as we looked over our data through all of 2024, so just talking about the full year that was, we saw about 20%. 18-20% more cash raised by companies on Carta in ‘24 than ‘23. So that's obviously a boost, but the actual number of rounds was down about 5%. So this sort of dynamic of bigger rounds going to fewer companies, I think it actually is true, especially at the earliest stages, early stage activity, the fundraising pace was effectively flat from ‘23 and ‘23 was not a great year by any means. So I do think that that feeling across venture, which is: there's some companies, typically AI ones, who are just speed running this thing and then everybody else is caught in this sluggishness. I think the data sort of does reflect that there's still either glut in the system where people are trying to figure out what venture looks like in a world of AI and lower round counts. I was expecting candidly that ‘24 would see a pop back in terms of fundraising pace much more than it did. I was pretty surprised by that number
Pablo Srugo (03:31):
And I know you're not necessarily a fan of splitting the data this way, but what percentage in Q4 or maybe throughout the years is going to AI companies?
Peter Walker (03:42):
When we look at the total split by round size, or excuse me, by round stage, so you've got AI in seed and series A, it's probably something like 30-35% of all the capital that we see or saw last year as going to AI companies. But the weird part is if you look at the late stage, it's actually much higher. So 50% almost of every dollar that we see invested into say series D and series E companies are going to AI companies and that's a dynamic that's very odd. Usually in a technological change it would start at the beginning and then a higher proportion of capital would go to early stage rounds and then those early stage rounds would graduate, et cetera. Again, AI has just kind of done this very quickly where now even the late stages are dominated by, we all know the names. There's like 10 companies that seem to generate gigantic, gigantic rounds and that swamps everything else.
Pablo Srugo (04:37):
Yeah, I mean part of it is that kind of distortion, whether it's openAI. I think another big one was Databricks just raising money and just kind of skewing the numbers that way
Peter Walker (04:48):
You take it at round count instead of just dollars. It's still a higher proportion in late stage than early stage, which I think is odd.
Pablo Srugo (4:55)
That is weird.
Peter Walker (4:56)
There's a discussion, I'm sure you have it as much as I do, about can non-AI companies raise in this environment? The answer is yes, definitely there are non-AI companies that are raising , but I don't know whether or not- when does AI stop being this hype cycle and a justification for a round and simply just becomes the technology that individual companies are using? It's got to happen at some point in the nearish feature
Pablo Srugo (05:25):
A hundred percent, and for me, it's already happened, at least in the way that I look at it. Maybe it's not in the data, but I just don't- people will ask me, what do you invest in? And I'm like, I can't say I invest in AI. That's just so stupid. It is just like saying, oh, in the nineties or early two thousands, I invested in the internet. Come on, who doesn't? Or in SaaS in the 2010s. Of course you do. Everything's just that way. If you're so these days, I mean there are core ai. I look at our companies and I'm like, okay, you're actually foundational. Or there's one we have that's doing a Cuda competitor, true foundational pieces of the stack that are core ai. And then there's people who are more AI agents that are leveraging AI for a bunch of different things, people that are leveraging AI in their products.
(06:12):
But again, it all starts to be this weird spectrum where it's like, are you AI? Are you not AI? And then to your point, it's kind of like, well, who cares? You're probably leveraging AI as much as you can. The only other side to that though is yeah, if you are starting fresh today, it is the inflection point that is the big change in the market. So there's a reason why many founders would do something that is ai, not just to be hot, be in ai, but just because of the opportunities that it might've opened up.
Peter Walker (06:46):
Totally. I think that this is the annoying part about the AI discussion is if you're a young entrepreneur just thinking about beginning a company, you should take advantage of the latest tools available to you. So in some ways, if you're not looking at AI at front, it's very odd. It's like, why are you just doing this as a contrarian thing that you just don't think that this technology is that helpful? I mean, fair, I guess that's a viewpoint that people can have, but many, many other founders disagree with that. So I think it's just going to become less and less the story of AI versus everything else. And then the story in ‘25 and ‘26 is going to be AI integrated into everything. So we stopped talking about it as a separate idea most of the time,
Pablo Srugo (07:28):
A hundred percent. What else? What other big kind of trends were happening in Q4?
Peter Walker (07:32):
Man, Q4 was a pretty good quarter in terms of valuations. We saw rising median valuations or seed stage for series A. for series B and series C, things kind of levelled off a little bit, but again, the story that has been basically the story we've been telling through this data for the last, I don't know, six to seven quarters here, is valuations continue to creep up on a median basis and round volume is either flat or a little bit down. So fewer companies, but they're achieving robust valuations when they raise. The other thing that I've started digging into, which I’d love your opinion on actually bridge rounds is this big- If you look at our data, there are so many more companies that are now raising extensions after a primary round. So if you just look at seed for instance, companies that raised something like, I don't know, 35 or 40% of companies that raised a seed round end up raising some sort of extension capital after their seed before they get to series A.
(08:38):
And when I looked at the difference between those kinds of rounds, there's two ways to raise a bridge round in our data. One is to do it on a priced round, which takes time, takes lawyers, whatever. And the other is bridges that are using safes or convertible notes. Those bridges have become a lot more common, the safe and notes one and they to series A at a wildly lower rate than the ones who don't bridge that way. And I have a feeling there's this kind of tricky feeling that I'm experiencing right now where I think there's a tonne of time and attention and money being spent on bridges that are going to end up being completely wasted. And I didn't really feel that way until the end of this year, but now I'm looking at it since 2021 and I think that there's a ton of bridges to nowhere that have been raised over the last three years.
Pablo Srugo (09:28):
That's very interesting. I have a lot of thoughts on bridges, but maybe just first on the data, so right now 30 to five to 40% of seed rounds have a bridge after, Is that right?
Peter Walker (09:38):
Yeah, 35, it's call it 30% of companies that raise a seed end up raising a seed extension of some kind before they get to an end.
Pablo Srugo (9:48)
And what's that trend? Was it always 30 or was it 15% before?
Peter Walker (9:50)
No, no, no. It's definitely gone up. It's gone up quite a lot. So look, and obviously the time factor comes into play here. If you raised a seed extension in ‘24, I don't know whether you're not going to raise an A yet. You didn't raise a series A in 2024, but that was kind of on purpose. You did a bridge. So we don't know what the ultimate value of those bridges are going to be, but if you take, I'm looking at data from say 2020, the percentage of companies that raised a seed round in 2020 and then have raised some sort of bridge capital since then on a safe or a note, only 4% of companies that fall into that bucket have gotten to a series a 4% versus 50% of companies that didn't raise a bridge round at all. So obviously if you're raising a bridge round, probably it's not the best thing, Hey, why did you need to bridge in the first place? So I get that we should expect the numbers to be lower, but safes and notes as this extension capital, the end result of those I think is almost always not good
Pablo Srugo (10:52):
It's interesting because I have a few different kind of perspectives on this and purely from my viewpoint, I mean one is I wonder we have that covid period, that ‘21 and ‘22 heart, which is so hard to pull out because in that time it is true that if you were let's say 12 to 24 months post seed and you couldn't raise an A in most of those cases, we're talking generalities here because there's so many exceptions to every rule, but as a rule, something was wrong because capital was flowing, it was going. So that kind of makes sense to me that foots at least again anecdotally, it's like if you had to raise a bridge round, like you said, nobody wants to raise a bridge round, everybody would rather raise the primary round. It's a higher valuation, it's more money. So if you had to raise a bridge round for that 2020 cohort, you were doing 21 or 22, it probably meant something was not really working.
(11:52):
And so then you are already in this, you have this adverse effect, whatever that's called, but you're already looking at the sample that's no good or not lower quality let's say. But then you fast forward and you look at the companies as an example that maybe raised the seed in late ‘21 or certainly let's say even in ‘22 they raised the seed round and then the rest of ‘22 was soft 2023 was also kind of soft. And now I think now, and we talk about this later, it's starting to get hotter again. But in any case, those companies, the bar first series A had moved up even just maybe from when they raised the seed and they thought about where they needed to be to raise an A, and that yardstick just kind of kept moving as they approached it. And those companies, and this is now where I'm saying based on my own, let's say things that I've seen myself either in my portfolio outside of it, but seen personally and many times raised seed extensions and I'm not sure whether the quality of those companies was actually worse.
Pablo Srugo (12:56):
I mean obviously if you compare 'em to the percent that just raised an A in general, obviously it's not as bad, but if you compare to the average seed company, you have the companies that raised the seed there, they're whatever quality, some raise an A right away, okay, they're top, some don't raise anything and then some raise a bridge round. That's where I'm less sure, especially because what I found is just like the jump in valuation between a seed of 15 million or so, 20 million probably median valuations to your 40 or 50 million valuations at series A. It's a big gap and just sometimes I have seen it again in these not so hot last eight quarters where it's like, you know what? You're actually on a good track, but you just need a little bit more juice to really be able to get there and raise that big great 10 million plus versus having to go to market and raise a price round, but maybe it's 5 or 6 million and you're almost in this no man's land.
Peter Walker (13:59):
I totally hear what you're saying. There's a part of me that wants that world to be true where a lot of these extension companies are really consolidating and going to be good candidates moving forward. The worry I have is that it may just be the case that many of those companies that ended up raising extensions are going to get swamped by younger companies that are moving faster and they don't have some of the cap table dynamics that the maybe seed rounds from late ‘21, early 2022, et cetera have. And so when we look at these graduation rates, for instance, hugely, everybody graduated in 2020. In 2021, it was too high and then 2022 it was very low. Even after two years, only about 16-17% of those companies had got from C to A, usually that's 30%. So a lot of companies got stuck effectively at seed and now they're doing what they can.
(15:02):
They get bridges, creative financing, et cetera. And then the question is, will a lot of them end up getting to A at any point or are they just going to get boat raced by the companies that are coming up behind them? They're growing, as you mentioned, series A metrics, what you need to raise a series A, has gotten what, conservatively two times higher, three times higher than it was three years ago and that’s just because a lot of these companies are moving so much more quickly, AI, etc on top of that. So I'm a little worried that a lot of those glutted companies are just not going to really end up being anything.
Pablo Srugo (15:43):
I think it's a fair point and it's totally a legitimate risk because again, you had, I mean basically, and I know there was GPT before chat GPT, but there's still the chat GPT moment, it ended ‘22 and in terms of the sort of things that people were building, it all changed through ‘23 and ‘24. And so if you had a 2020 or 21 company pre-chat GPT, you're probably not really AI at the core, you're doing whatever, 22 comes along, market sucks. You raise a bridge round late 22, early 23 and the world's changing behind you and then that new YC company or whatever that's born in mid 23 with all that knowledge just goes out and starts building AI agents and doesn't have any of the baggage that you had and in many instances can have an equally or more compelling product. Yeah, it's not going to be as feature rich, but they might have this one pain point thing that because the AI agent is just so well suited to do that and really the ROI is so high that you're actually trying to build that in while having everything else and you've got the cap table baggage.
(16:46):
I've seen it. I'm seeing it on both sides and the companies that are like they were the 2020 cohort and the new companies that are actually going to come against some much bigger companies but that are kind of pre, let's say chatGPT moment and it's an interesting dynamic. It's yet to play out, but I think it's an opportunity for the younger ones and a risk for the later ones. I guess that's the whole idea of tech disruption
Peter Walker (17:10):
A hundred percent. It's just funny that that happened to these youngest companies. They're like, oh, they're at seed stage, they're the disruptors and they're like already getting disrupted by the behind ones.
Pablo Srugo (17:20):
Exactly
Pablo Srugo (17:21):
Well, the older ones at least have distribution and distribution for this era is very helpful because AI in a way is actually easy to add. And so if you have distribution and you add this AI element and everybody gets efficiency, it's huge. But if you don't have distribution yet but you do have some baggage, sometimes it's better to have just a blank slate and just start from zero.
Peter Walker (17:45):
I kind of feel for the entrepreneurs who are kind of caught in that transition, it's not easy by any means usually, but I think it's even harder when there are these macro winds of: low fundraising environment plus the advent of AI plus all these younger companies that are coming up behind me. It's just a quagmire. So it's tough. I do think that there's one of the ways that founders that from our data are responding to that is- one of the trends that's been very clear since the beginning of 23- is they're just not hiring. They're just hiring way fewer people per unit of revenue than they used to. And that's great in a way. Obviously you've got ARR per employee that's going up, you've got these capital efficient businesses. Part of me is a little sad, it's super awesome to work at an early stage startup and I wish more people got the chance to do it, but I totally understand why companies are choosing that. It's just that's one of the clearest patterns in our data from 24 is: companies are staying smaller a lot longer than they used to.
Pablo Srugo (18:53):
Do you have numbers on that? How many employees the average seed company has and maybe even aRR per employee, these sorts of things?
Peter Walker (19:00):
Yeah, totally Not on the ARR side, but I can give you full-time equity holding headcount. So that's the first part caveat to this story is we can see very clearly equity holding full-time employees, but if their companies are doing a lot with consultants or fractional that don't get equity, it's a little bit harder to view that. But just on an aggregate basis, if you look at a company that raised a series A round in the first half of ‘21 on Carta, they had a median employee count of about 20. full-time headcount. Now at the same company raising a series A in 2024 that was more like 15 and the second half of 2024, that was more like 13, which yeah, the gap between 20 and 13 doesn't feel big, but those seven headcount represent a gigantic decline in the number of employees. So a lot of people are really scrutinising new hires that way, and I honestly don't know- Usually if this had happened, you would say, okay, as fundraising picks back up hiring, we'll pick back up. But in the AI age, I don't actually know that to be true, and I'm not sure where startup employment goes long-term that way.
Pablo Srugo (20:12):
I mean there's so many things too because you have the freshness of all the insanity that was 21, 22, and founder's minds, whether they lived it or their friend lived through it, they're like, the last thing I want to do is overhire and get my burn just too high. That's just so fresh in people's minds. And then on top of that, you have a way to fight that, which is like you said, through AI. I must admit as a founder, I think it's just awesome. Smaller teams that can still get the stuff done where you don't feel like everything's just followed by the wayside, you need more people. But where it's not resource constrained, it's by decision. It's very powerful because you do trade off as you grow your team, you need more communication, more alignment, more overhead. You can do more, but you can move. It's slower to move, it's slower to get things, make new decisions. So the ability of companies now maybe having the best of both worlds, a lot of money in the bank, a lot of ability of getting a lot of things done by leveraging AI, and yet that small team where communication overhead, et cetera, is it is very small overhead, very clear communication, very clear alignment is very powerful dynamic.
Peter Walker (21:21):
Alright, so as the VC in this conversation, what happens if those companies don't need the cash. they were spending the cash, used to be spending the cash on headcount primarily. Now maybe that goes to compute or other places, but the idea that compute costs would equal headcount costs, I think that would be a large assumption. What happens to companies that are coming up and they're saying, look, really, I liked the venture idea. I took a round, I have some outside investors, but I don't really need more capital right now. What happens to growth VCs?
Pablo Srugo (21:59):
That's a great question. I'm just happy I'm not a growth VC man. Really. Yeah, no, it's funny for us at seed we're so aligned with the founder from a cap table perspective because in the traditional venture model, we're going to be diluted four or five times to an exit and sit behind a prep stack, take on all that risk just like a founder does. So in a way a world where they talk about that one employee billion dollar company has this idea in your head that is still far from reality, but just maybe a north star sort of thing. The idea that a company could raise 1 to $5 million, let's say and just go and still grow exceptionally fast and still hit real scale but just become profitable a lot sooner, is music to my ear. To me, that's awesome. That's just the best case scenario.
Peter Walker (22:43)
Is it a best case scenario if then you don't get any markups for your fund?
Pablo Srugo (22:48)
If that's the case, then there is that interim period that's a little bit challenging. But we have other ways and frankly we started looking internally at measuring our performance by revenue generation, how many companies are on our path to 10? And so I think it is, I think the net effect is everybody makes more money, the founders, the investors, the LPs, and in the meantime, I think if you're a smart LP, you can see through the kind of net IRR as just the metric and look at the underlying portfolio of the company. And if I have a company, hey, we invest in this company, we own 20% and it was doing nothing and now it's doing 50 million ARR, yet it hasn't raised. But you don't have to be that smart and be like, okay, that's interesting.
Peter Walker (23:28):
If the LP doesn't understand that, maybe it's not the right LP.
Pablo Srugo (23:31)
Yeah, I agree with you there.
Peter Walker (23:33)
I do think that this dynamic, though, pushes. It's why, it's part of the reason, it's not the only reason. It's part of the reason why some of those multi-stage funds and late stage venture is getting so wacky these days. It's like we have billions and billions of dollars to be invested, but it seems like the candidates for that investment are lower than they used to be. So we're just going to plough some crazy rounds into these companies and we're going to do it again and again. And it really does suggest that there are two different parts of venture opening up and the gap between early and late stage has never been this wide. I dunno, I don't know what happens.
Pablo Srugo (24:19):
Do you see that? Does that show up in the data? Because intuitively it makes sense. It's like, hey, it used to be no marketplaces and customer acquisition and product and engineering and now it's a lot. Product engineering is still there, but the marketplace stuff, the really expensive Uber like models, they're not really happening. That money is maybe going to GPUs, but who's really doing that? I mean foundational models, the really big scale companies, everybody else is in need of that scale of spend.
Peter Walker (24:46):
Okay, look at the venture dollars raised in ‘24, which we all know is billions and billions of dollars, but very concentrated to name brand funds primarily. What return can those funds get on those billions and billions of dollars? You mentioned it, there are like eight, maybe five foundation model companies that can gobble up cash that way, although we can talk about what DeepSeek does to that, where does that money go? How is it spent effectively? I honestly don't know. It's a venture I think is going through this moment of: it is not this tiny kind of scrappy asset class anymore. It is professionalised. We probably need to start talking about it the way we do with private equity, whether there’s low, middle, high, there's different kinds of private equity that have different strategies. Venture is not a monolithic thing. It has different strategies and different investors are going to need to differentiate themselves by the way they talk about themselves moving forward. Because it is a totally different game to be deploying a $2 billion fund than a hundred million dollars one, it always has been, but it's even more different now.
Pablo Srugo (25:57):
And the other thing I wanted to touch on is there's this feeling, and it might just even be me, but it definitely feels like the winds are shifting and we're going from, and I think it's weird, the stock market has been on a tear for a decent amount of time basically since 2023. But in the world of seed stage venture, it's been for a while that the thing just gets repeated like, oh yeah, it's not the best time to raise a fund, it's not the best time to raise around as a founder or whatever. To me it feels like that story's kind of over, the winds are changing and it's actually a great time to be in business. But is there anything in the data the last few months that are showing any, basically it feels like things are getting hot again. Is there anything that sees that in early stage or is that just straight up in my head?
Peter Walker (26:46):
Let's see, what could be pointing towards heat in the early stage? I mean look, the number of rounds and dollars deployed in say seed and series A was basically flat to a little bit up from 2023, but the Q4 was a pretty good quarter versus say Q1 or Q2. So there's a bit of a ramp towards the end of the year. That's definitely true.
Pablo Srugo (27:13):
The round sizes or round prices in Q4 versus the rest of the year?
Peter Walker (27:16):
round sizes and round prices were- So let's go with seed stage, for instance, seed stage round, primary median valuations for a seed stage round on Carta were 15 million in Q4 after being 14 and a half in Q3 and 14 in Q2. So rising up, that's pre money. So then you add in whatever, 3 million raised or so and you get to pretty high valuations for seed stage companies. Series A kind of the same thing, actually a little bit flat in Q4. So yeah, prices and cash continue to rise pretty slowly, but they continue to rise. I also think that there is this dynamic of location where geography comes into play here. Whatever happened ‘21 into beginning of ‘23 where it kind of felt like Silicon Valley had lost a little momentum, that has definitely gone
Pablo Srugo (28:15):
Didn’t last long.
Peter Walker (28:16):
Yeah, SF is ripping. even in comparison to what it usually does. I think it's ripping and I don't know, that’s an AI story to a large degree. There's so much AI in the bay and in New York. There's less AI everywhere else, and I am waiting for that to change. And we haven't seen it change significantly yet, but I would imagine 25 is the year where AI becomes the dominant theme in the non-major ecosystems as well.
Pablo Srugo (28:47):
Yeah, I would have to think so. Any final trends? Maybe what's the top three? I mean we talked about conversions, we talked about prices overall, the year things are seemingly getting better. There's definitely this dichotomy of the really, really high big huge rounds and then things at the early stage more let's say normal, and then the opportunities that people have to do more with less. Anything else?
Peter Walker (29:10):
I think the other thing that we didn't touch on yet is exits. So Q4 of ‘24 saw the most acquired companies off Carta of any Q4 in our database, even higher than ‘21. So now a caveat, we don't actually know for what price those companies were acquired in most cases. So we don't know whether or not those are acquihires and fire sales versus acquisitions that really make people a lot of money. But we do see a trend towards increased M&A. I think that's going to be true in ‘25, probably some bigger deals than were under consideration in the last couple of years for sure. And then we're all just sitting around waiting for those IPOs. when ServiceTitan’s going out. There's a couple other ones on the schedule. It's just, again, I think it's always underrated, the impact both in terms of morale and actual recycled capital that IPO has.
Pablo Srugo (30:04):
What's funny is if you look at the names I think about that are, I guess they're not that recent anymore, but Duolingo, Reddit are two names that come up in my head at least. And they're crushing it. They're up 5x since IPO. It's insane.
Peter Walker (30:18):
I thought that that was going to be more of a starter gun, but it didn't really materialise into a big bump in IPOs. But I mean that is the conversation when I speak to a lot of later stage VCs these days is like, when is liquidity really going to start opening back up?
Pablo Srugo (30:34):
Makes sense. Cool, man. Well dude, appreciate you sharing all of that with us. It's always great to have you on the show, man.
Peter Walker (30:42):
Absolutely. Pablo, great to see you, man
Pablo Srugo (30:42):
Listen, when you go to a restaurant, you eat a nice meal, maybe a fancy one, maybe not. Do you leave a tip? I assume you probably leave a tip. You probably leave a tip 100% of the time. Well, guess what a review is just like a tip, and I know you haven't been leaving one. So just like the waiter that doesn't get a tip after hours of great service, I'm getting a little frustrated. So take your phone out and leave a review. It helps the show move up rankings. It helps us get better guests. It doesn't just help me. It helps weigh more founders. Thank you.