#6 Jared Sleeper: Partner at Avenir

0:04 Today I'm excited to be joined by Jared Sleeper, the axe on software for Fenwit, Harvard undergrad, Putinham as a research analyst, matrix code 2, and now back focused on growth as a venture partner. The reason I wanted to have him on is simple. He has a unique perspective given all his experience at top funds. More recently, he's been doing public software tearowns. Peloto, Tyler, Atlassian, names that investors know well. And what I really appreciate, he's going company by company, business model by business model, trying to dig into what really matters and where the market may be getting it wrong. So today, we're talking everything software, AI, and what it all means.

0:45 Jared Sleeper, welcome to Liquidation Nation, my friend. >> Jared Kubin, great to be here. >> So, let's get right into it. I love your public company tearowns. you've now done 50 of them. The market is clearly saying these business models are impaired and the terminal value distribution of outcomes is much wider than even 12 months ago. >> What do you think is the biggest thing the market is getting wrong about AI's impact on software? >> Yeah, it's it's a great question. Um, you know, having done these 50, I think the biggest thing is just treating software companies as a monolith, right?

1:20 Um, and even the people who go a step further and call out a few software companies that they think are relatively safe will often add, you know, but the rest or but the long tail of, you know, garbage, you know, uh, point solution software is all in trouble. And whether or not that's true, you know, I really wanted to force myself to look at each company individually and figure out, you know, what that long tale looked like.

1:50 And so I just went through a list of 50 public software companies one by one. And you know, I don't know that that long tale exists in the public markets per se, right? I'm sure there's a bunch of private companies where you could make that case. Um, but that's my biggest gripe with the discourse, right? is a lot of people making these big broad brush statements about software or or different pockets of software and really hesitating to get into the specifics of you know this is a specific software company that I think is going to die or struggle or see xyz happen to it um on a defined timeline and so you know I don't necessarily call myself a software bull I think I'm just someone who believes in nuanced takes in investing um and I really enjoy debating specific speific predictions about how something's going to play out. Um, not broadstrokes. And so, yeah, I'd say when I did the, you know, actually when I, you know, did this, another thing that just really struck me that's worth talking through is this like paradox between long-term risk and short-term acceleration. So, the narrative that you'll read on X is, well, these businesses need to accelerate to prove that they're not dead. Um, and last week we had Figma um, accelerate in a pretty meaningful way. As I went through and did these profiles, I was looking for the degree of AI risk. Are there startups that are tackling this company that have real traction and real revenue that could pose an issue for it over time? And there was a pretty tight correlation.

3:24 The better the chances of the company accelerating, i.e. the more AI traction a given public software company had, you know, today in its business, the greater the intensity of the competition, right? Because logically they're close to these valuable AI use cases like AI for code or AI for creative. Um, and so the irony is that the companies that are not accelerating probably are safer on a five-year view because AI is less disruptive to the workflows and to the work that they do than the companies that actually might accelerate over the next two to three quarters. And I find that really interesting because I don't think a lot of people are thinking that way in the discourse today.

4:04 >> For the last 20 years, you and I have looked at the software space for a long time talking to all these companies and especially the companies that were buying software, right? The default answer to enterprise software was always buy first and not build. AI is obviously making internal software creation much cheaper or at least the uh vibe coding aspect of it. Where do you think that actually changes behavior for companies and on the other side where do you think investors are dramatically overestimating the death of software?

4:39 Yeah, I mean I think one, you know, I I disagree with the premise a little bit. I think there's a lot of homebuilt software out there that companies have built themselves and a lot of the world's largest companies have both very large development teams internally and then also work a lot with consultants to build and maintain these kind of Byzantine custom homegrown systems. It's certainly true that SAS have been gaining share relative to those systems, but still lots of software that have been purpose-built um sitting around out there. When I think about what's happening, you know, vibe coding is so powerful and I, you know, am writing clock code every day. Um where I haven't yet seen it really break out is people vibe coding multiplayer applications, right? Um, so you know, you've got your kind of core systems of record that everyone has engaged with and agreed to.

5:32 You know, there's this amazing story I tell about our internal CRM where uh someone on my team decided that he was going to change a few fields in it uh for his own benefit. Uh, and he pitched it to the group, but in a Slack message he sent out one day and I literally ran out of my office and was like, "This is going to be a disaster. You have to roll this back." And sure enough, within like 12 hours, we had people on the team losing their minds because one of their workflows had been broken. Right? And so that's exactly the kind of dynamic you would expect to happen if people started vibe coding on top of systems that other people use, right? So much of software development is the coordination of well a thousand people are going to be using this. How do we cater to the needs of all of them, right? In different countries, geos, backgrounds, etc. And so what I do see happening that's really exciting is it's never been a better time to be an individual user with access to your company's data. You can build exactly the tool that you want, right? Uh what's been harder to see so far is people vioding stuff that's going to be used at scale, right? Or even at a, you know, beyond an individual scale.

6:43 uh someone came in here to Aanir and pitched me on re rebuilding a CRM for us and I was like great so are you going to maintain it you know are you going to be the one who handles it when there's an outage like will we will we call you wherever you are and figure it out um and so I think the best-in-class companies are really going to focus on freeing up the individuals to build micro software for themselves or maybe for small teams while organizing around the organization itself around what are the tools that we can give them to help them do that that stay safe and kind of you know managed and agreed to by the whole firm. that like one thing I've been thinking a lot about recently is especially during this age of SAS right like kicked off with Salesforce and Workday obviously Service Now a lot of these companies for the longest time have tried to ingest as much data as they could and almost put up like these walls around their data as we kind of think about like the next 10 years it's almost going to be the opposite where I think they're going to try and tear those walls down and try and get as many agents or people or subscribers riers or other companies to try and connect to their data- which is a completely different paradigm for some of these companies to adjust to.

8:00 >> Totally. I mean it's you know it's the best case for how you could see some sort of innovator's dilemma situation show up in software which is every software company fundamentally is a database or a set of databases and a set of user interfaces. Now, I'm not a maximalist like some that the user interface is going to go away entirely for sure, but there are definitely some pockets of user interface that can go away. You're probably the same. I know you vibe code a bunch. You know, I have a GitHub login. I haven't logged into GitHub in like two weeks except to check my street counter and screenshot it once to post it on Twitter, right? So like you know the GitHub user interface you know has really become moot and everything is getting pushed you know out of cloud code and terminal there and so you know you can see you can imagine a world where a lot of software companies have a version of that dynamic happen to them where their data is useful uh it's being queried that is valuable and they are kind of serving as this kind of like core layer where it's all organized and the agreed upon definitions live and all the data pipelines come into and some of that data comes from a user clicking some boxes and filling out a form in a UI and some of it comes from you know trans call transcripts being ripped down etc.

9:23 What's interesting is you know today most software companies charge per seat not all but most right and that model actually would look a lot more like what data bricks or snowflake do where they store your data for you they charge you some diminous amount to do that and then they charge you a consumptionbased fee as that data is used. Uh that model exists it works great. In fact, data bricks and snowflake are, you know, relatively beloved uh as far as software companies go by investors and by customers, right? And so, yeah, you could imagine a world where, you know, the Salesforces and HubSpots of the world are effectively really nice purpose-built databases with, yes, some UI touch points, but that they're mostly being consumed in the same way that someone would consume Snowflake or data bricks credits. And a lot of these companies are starting to move that way.

10:15 There's a really interesting question of do you put a credit model on top of your subscription uh and see where that goes and try to make it incremental? Do you convert your entire subscription model into credit system uh somehow? Uh so there's a lot of pricing and packaging innovation going on. It feels like so far the equilibrium is stick with the subscription model, add the credits on top and you know it can it can prove incremental. Um but we'll see. It's gonna be really exciting. uh to see how it develops.

10:44 >> Well, what's really interesting right now is these companies are almost fighting like a threefront war. Uh you mentioned the technology obviously. Number two is the financial models that are going to emerge from this. Like I don't think anybody has any idea. We have a lot of you know good guesses on what they could look like. Then three is obviously on the marketing about how these teams and these IRS and CFOs explain their models so people can actually understand it. Like if we rewind analyzing a license and maintenance model is one kind of math.

11:19 >> Then we kind of move to changes in deferred revenue. Then we kind of move to like usage and RPO type models. And on that point like you and I used to share a lot of charts. We used to look at gross profit dollars, free cash flow, growth kers, and we used to kind of look at the world on this rule of 40 type math. >> How are you thinking about these metrics and how to analyze these companies in a world of AI? Do you think new metrics have to emerge?

11:50 >> Yeah, it's a great question. I mean, I'll dial back and say that this has been an immense personal frustration of mine, right? If if I take my career uh I've been doing this for 11 years now and I jumped right to the buy side and started doing public markets at Platinum which is where I met you and about half of those have been good years for SAS and half those been bad years for SAS right like somewhere right in the middle everything changed and it went from what whale rock used to call God's gift to capitalism in its shareholder letters uh to something different something much worse and I really do believe and I've been disappointed in in the way that management teams and investors alike have adopted to this new paradigm. Uh and the bottom line is that while the market perceived this to be a very high quality business model um and it was it is a high quality business model, recurring revenue, high gross margins, you know, often strong network effects or brand effects. Um, the companies kind of gave up on like financial literacy.

12:58 That sounds so provocative, but it's true. I remember seeing there was a year where I saw like 20 SAS IPOs at Putinham. It was like one of the heydays for SAS IPOs. Must have been 2015 or 2016. And almost all those IPOs would have a slide at the end which was like our target long-term operating margin is a 25% non-GAAP operating margin. And I was like still kind of investing purist back then. I mean still am, but I was especially a purist back then. I just come out of undergrad, right? And I was like, "So what do you think your stockbased compensation's going to be in that out year?" And they'd be like, "Oh, you know, 20 20% of revenue." And I was like, "So you're telling me that your long-term margin is 5%?" like that just doesn't make any sense. And a a couple of those guys became my friends down the road. And it turns out that the bankers just copy and pasted the same slide for every single IPO road show. No one wanted to even think about making an argument for why they'd be a higher margin business than that non or even like considering stockbased compensation. And so that was what was allowed for a long time. And frankly, the management teams of a lot of the world's leading software companies are brilliant product people or brilliant salespeople. But because the business model is so rock solid from a recurring revenue and theoretical margin basis and was treated that way by investors for at least a stretch, they kind of left the finance piece as like a kind of third class citizen in the organization, right? And so now like there's a real struggle for folks to articulate, you know, core principles of their model.

14:33 Uh, you know, talking about gap, you know, versus non-GAAP. I don't even think it's even a lexicon of a lot of SAS CEOs to this day, right? Um, so to your question on new metrics, I mean my number one ask would be for everyone to convert their non-GAAP operating margin target to a gap operating margin target and just build religion into their organizations that stockbased compensation is a real expense and fungeable with cash. one management team, Bentley Systems, God bless them.

15:00 Uh their last names are Bentley still, which tells you something, said this explicitly on their earnings call and I gave them a shout out because it it brought me tears tears of joy. Um but most software companies have not done that. Um and a lot of the big tech companies like Netflix and Neta have, right? And so I would love to see that. In terms of new metrics to analyze the businesses, the the honest reality is not enough has changed yet on the financial side. These are still by and large subscription business models. Their gross and net retention if anything have slightly improved as a group over the last three or four quarters and I think those are still really great metrics. Like for me, the litmus test for whether something's going wrong in software over the next two or three years will be if we see a software company watch its gross retention really degrade from like mid9s into the high 80s for something like that because it's getting ripped out and that I think would be an amazing thing to dig in and understand and pro if that happens it's bad for the whole sector but especially bad for that company. Um, and I would want to deeply understand what's what's happening there. That's not happened yet. Like I don't have a good example for you where you've seen that kind of degradation.

16:15 >> Think about it. Like I'm sure you and I are going to talk about this probably sometime over the next 12 or 18 months. But you can imagine a scenario where a company comes out and reports earnings where gross margins are down, operating margins are down, deferred revenue falls off a cliff. Yeah, >> but somehow net dollar expansion really accelerates and the stock's down like 25% after hours and people are really confused on what's going on.

16:40 >> You could see a lot of confusion. I mean there's, you know, the transition to a fully usagebased model if anyone does that could be very messy, you know, in both directions. Um, and I think I mean take even a step back from all of this, right? The software companies today are just so bad at storytelling. Um because they haven't had to. I think you know they they didn't have to historically. It was taken as such a given in Silicon Valley that software was an amazing business model and that you trade it 10 times sales in the outright in the in-out case out year, right? And so, you know, we've got some private portfolio companies here at Aanir that I think do an excellent job of explaining to the world and their customers and their employees that they are on the cutting edge and they're out there and you can see on X like ramp in our portfolio. Amazing company, right? You know, they have AI agents playing Railroad Tycoon. You know, they write about internal tooling that they've built and how widely it's used. And the investor community understands that they're operating at a very high level. And I'm sure that that sort of marketing encourages the employees at RAMP to operate at a high level and makes talented people want to join RAMP, right? To be part of that.

17:58 Um, I don't see the public software companies doing a good job on average of marketing themselves that way. And I suspect we're actually seeing less of the exciting things that are going on inside them than exist. Uh, in part because of that. Now you've got like Bill Staples from GitLab who's recently like picked up the pen and gotten active on X. I love that. Uh I think that's exactly the right thing to do. Um because this is a narrative war and these SAS companies will not be able to fight it with metrics uh for some time to come if if I'm right about the kind of way AI will will kind of you know spread through the world. And so they've got to fight it by convincing investors and employees and customers that they're not dead. Uh and you can't do that by being silent or having your head in the sand.

18:48 >> Interesting. Le let me ask you about a historically a dirty word in software services. >> So Sequoia obviously came out and put out their big piece on the big market opportunity. You have to be seeing a lot of interesting companies in like the growth phase. Do you think that that this kind of services thesis, this forward deployed engineer, is that showing up right now? And h how are you looking at that opportunity >> over the next couple years?

19:19 >> Well, look, I mean the objectively best performing public software company of the last three years from a financial basis has been Palanteer and it's not even close and they are the ones who coined that term, right? Maybe it was their agency, but they did a great job. Um, look, I think and and I will may couple this like I'm certainly culpable, you know, and I know you remember these days well like there was a time when services was a dirty word in software and investors considered services to be a low margin business and we wanted software companies to have the highest possible proportion of subscription.

19:54 services looked at a stance and a lot of software companies reacted to that by dropping their services arms and outsourcing them and you know I I saw probably a dozen plus examples of a software company getting up on its earnest call and saying you know what our revenue is going to be a bit light but it's because we're shifting our services to thirdparty partners and you know subscription revenue is going to be and the stock will go up right people are like oh great you're just making yourself a pure business right um and you know I'm not sure that we could have seen it coming but in retrospect I think that was a huge mistake uh because unfortunately it separated software companies from their customers um in a pretty profound way and those customer relationships and that un institutional understanding how to implement complex solutions within customers is exactly what's needed right now to glean maximum value from AI now that might change over time right like there may be an era in five years where it makes the most sense to buy packaged AI solutions and certainly if you sell to SMB customers they're never be able to afford services anyway. So like you know they'll have to wait for it to be a packaged solution but yeah I think it's it should by rights come back as something that software companies really want to lean into because we're dealing with a brand new basically alien technology, right?

21:20 Um, and so customers are going to need help figuring it out. And as an institution, you want to be learning from like the experiences that people are having on the ground. And so, yeah, amazing position for Palunteer to be in because they were kind of services maxing way before it was cool. Uh, huge credit to them. And I think the software company should play catch-up. And I actually would be surprised if the market would look at it negatively if a SAS company said, you know what, we're beating our revenue estimates, but it's going to be gross margin a little bit, you know, um, detrimental because we've decided to hire 50 FTEEs to like for deploy them into our customers. Now, there's a question of could they hire the right people and how do they operationalize that, which I think is a very good question. Um, but I don't have a doubt. I don't doubt in my mind that if they could hire the right people, that'd be the right thing to do. Right now, >> I'm actually waiting for the first large software company to say, "Hey, you know, for the last 10 years, we've been um we we have a lot of partners. We've been pushing a lot of services out. We're actually pulling it back in house right now. There's a very large one in the state of Washington that would be primed to do something like this. And I'm just waiting to see who the first one to do this is going to be."

22:33 you know, it's um there's always a couple folks who really lead from the front, right? And you know, we know the folks you'd expect it to be. I mean, I will say, you know, to their credit, you know, we were talking about CrowdStrike just before we jumped on, like CrowdStrike has always had a services business, right? and they've always prided themselves on having elite cyber services and they've talked on earnings calls about how many dollars of net new AR they generate per services engagement when that team goes in and helps a company that's in deep trouble get out of it. Um, and so yeah, we'll see. I think you're right. I think we'll see a move in this direction over the next couple of years. And it's good because that's also a way to employ people who won't be needed to write packaged software anymore. So, uh, world stays in balance that way. Oh, I remember when they were on their IPO road show, George came in and kind of told the whole story. I actually brought his um I think it was called Hack Seven.

23:26 Uh you might be able to find a copy on Amazon right now, but I brought one of his books from way back in the day to our IPO meeting. He signed it. He signed it for me and uh it's fantastic. So, George, love you. Congratulations on everything. >> He's so responsive on email even to this day. Uh it blows my mind. And I'll tell you a story that I think is just helpful in this context for some of the folks who are listening who are, you know, maybe software investors, maybe not. But in December of 2019, I went to a 20 person lunch to discuss CrowdStrike. And this is like four or five months after the IPO. And if you remember, it didn't do too well after the IPO. Um, and there were 19 Crowd Strike Bears at that lunch and one bowl. And it was like a shooting gallery. Like we went around and person after person was like endpoint's a terrible category. The pricing compresses to nowhere. End users don't even know what's on their laptops.

24:23 What's sticky about that? Um all these like fairly true reasonable things. Um and obviously that was completely and utterly wrong. And so if you're ever doubting whether the entire market can be wrong about something at once, you know, these are the smartest software investors on earth. uh almost to a man or woman wrong about one of the great companies um in the history of software. And so it's always good to have that epistemic humility baked in and remember what that feels like. You know, >> other just interesting things I find about the industry and we talked about how services was a bad word and honestly security has been a bad word in tech for a long time. I mean all the way back to fire eye. But how are you thinking about the world of security in the age of AI?

25:14 >> Yeah, it's a great question. Uh I think already two years ago if you surveyed CIOS and CTO's and asked them about their top spending priorities, security was number one and it wasn't close. Um and that was before what's happened recently, right? Um there's at least two separate vectors that I think are really interesting. The first is when you have agents running around in your environment, so many problems get an order of magnitude more complicated than they were before, especially around identity and data security where some of the basic assumptions you could kind of make as a CISO that an end user would apply some common sense don't apply anymore. Right? Um I just tweeted about this yesterday, but I am, you know, vi coding a little app for myself for fun and claude deleted my production database. Um, and it was not backed up and thankfully the only user was me. So, it wasn't too big of a deal, but I did lose my like 12day streaks on the little games I was building on that. And there's no way to get them back. They're gone for good. Um, and so, you know, that how do we manage agents in an enterprise is actually a very complex problem. Uh, because agents like people cut across all the different substrates.

26:32 So you might say, well you could use a browser and if someone goes into chat in the browser and pastes the wrong thing in you can catch it there. But then what if they download the desktop app and they do it on the desktop then crowdstrike has a lens on it but Crowdstrike doesn't have perfect visibility into the browser right and so it'll be really interesting to see I think every CISO on earth is trying to figure this out and what I think is going to be particularly interesting is the wave of AI adoption we've seen to date has been mostly individual users.

27:00 Uh, and I think every company on earth knows that there's been a lot of security protocols violated in furtherance of letting that happen. Um, but for enterprises to adopt it and sanction it and bless it, they need to get their security ducks in a row. And the reality is committees of people just always move on timelines that are like months to years, not days to weeks. And so over the next two or three years, everyone knows they need to do this, but I think we're going to see some really interesting shifts in how organizations do security. And then there's the code security stuff with mythos, which from everything that I hear is a genuine like massive moment. Um, you know, it's a very natural extension of AI for code, right? Like AI is exceptionally fluent and capable at code. And so why shouldn't it be able to look at an application and start chaining vulnerabilities or weak points together uh into full breaches? It's like it's so logical when you consider it that way, right? And that just means that like the attitude towards code security in particular is going to have to change fundamentally. Um and that's I think completely separate from the agent security thing. Uh so yeah, in short, a security is going to be an amazing place to be from a business and investing perspective. the public company multiples do reflect that, right? Um, but you know, I think they reflect it for a good reason. Um, and it's really easy to see how it could be a share gainer from here relative to SAS.

28:25 >> Are you seeing anything innovative in kind of like the private company growth stage in the security space as it relates to AI? >> Yeah, we're seeing a bunch. Um, so I'd say I'd take through some companies I think are are interesting. You know, there's um a bunch of companies in the code security space like Changengar, I think is a fantastic company. Um it's hard to explain, but they kind of flip it on its head and instead instead of scanning your code for vulnerabilities, they give you packets of code to use that are vulnerability free. That's my attempt to make that um groable to a wider audience. Uh, and so there's some limitations because your your engineers can't just do whatever they want, but whatever they do do is kind of guaranteed to be secure at least based on what's known publicly. Um, easy to see how in a mythos world that might be a very appealing way to develop software uh for large folks. So I think that's really interesting. And then there's a bunch of specific companies, but I think you know dealing with data security, right? Uh, Veronus is a public company.

29:29 There's companies like Sierra and others. Whiz has a product that are private. Um data is interesting because it's the um choke point, right? So I just gave an example of agents sometimes living on your desktop, sometimes living in the browser, sometimes living, you know, as an API call, you know, sitting on a server somewhere. It's hard to gra to grapple with all of those. But as an enterprise, the one thing you know for sure is you don't want them taking your sensitive data or deleting your sensitive data. And so there's an argument that the control plane for how agents operate is somewhere at the intersection between identity and data and there's a lot of interesting startups like Sara and a ton of identity startups uh that are trying to figure out how to tackle that. Uh so yeah, interesting space.

30:13 >> Well buddy, this has been fantastic. I'm sure people are going to want to follow up to this conversation. You and I could talk probably for days on some of these topics. Let let me close with this question for you. For young investors out there or let's say young founders entering AI right now, what should they spend the next five years obsessively learning about that you you think people are ignoring? >> Yeah, I've got two answers to this question. I think I don't know. I always hesitate to say people are ignoring something, but I'll just give you my my strong takes. I think the first is the way that investing operates has changed so fundamentally in the age of AI and everyone says that it's like a trope but I think the disadvantage that comes from being part of an institution that is not AI first and allowing you to use the latest tools has never been greater. And so I was at Harvard giving a talk uh to the financial analyst club that I used to run. And I've always given the advice that I think people are best served chasing their passions early and working for companies or investing rather than going through the traditional banking consulting routes. But but I've never felt more strongly that you as a student or as a you know a new entrant into investing need to be interviewing the organizations that you're considering working for and finding out if you are going to have access to cla code. if you're going to have a token budget, if there is an incredibly restrictive compliance policy around you or not, because man, if you spend two years editing PowerPoints without an AI plugin while the rest of the world is learning how to adopt the most powerful technology ever developed, you are just going to be behind. You're going to and in a way that I don't know that has ever been true before. And so that's my first piece of advice, which is you just must be part of a forward-thinking organization. And the best case is your organization's pushing you to use AI more. An okay case is you're free to use whatever you want. But the worst case is your organization is actively slowing you down. And I genuinely think that a lot of the world's largest organizations, and it's not their fault, they have compliance, you know, regimes that are strict and inflexible, etc., but they're actively slowing their employees down and that to me is an untenable environment for a young investor. So that's one thing I think and and hopefully in using all these tools they will get sufficiently excited about what's coming and how much opportunity there is as an investor when the world is changing as quickly as it is. But the flip side of that is I think I'd encourage them to really focus on business fundamentals, right? Uh there's a lot of hype out there in the world. um a lot of people who are trading same day options and buying shares of things they know nothing about and there's a lot of business model innovation happening right uh we've got companies with 0% gross margins uh that are growing really quickly that are you know have a story for how the gross margins will be higher uh but you know this wasn't really a facet of the SAS uh investing wave the company's always had relatively high gross margins right you know we have companies with very high gross churn that have a case for how they're going to improve that churn over time. But in general, there's just this like level of new business models emerging with AI.

33:34 And I think there'll be a lot of payoff for the folks who stay anchored in the first principles of what makes a good business and can suspend their disbelief about the technology and believe what the technology can do but also apply a hard-nosed understanding of how economics work and how capitalism works and that you can never be successful without other people coming for you and you know build their own concept of what moes look like in this world and what a good business looks like because when everything's grown growing two, three, 400% year-over-year. And in my world, we see a lot of that, right? It's going to be the ability to discern which ones are the lasting businesses and which ones are not that make for good investment picking. Um, and so, yeah, that's what I'd be telling any young person to be focused on today, and that's what I'm focused on. So, I'm right there with them. I >> I think it's fantastic advice for everybody, and uh, appreciate you coming on, sharing your knowledge with everyone, and I'm sure everyone is going to want to hear a part two to this. So maybe we'll circle back in 6 months and see what kind of acceleration and deceleration and crazy chaos is going on in the markets at that time.

34:41 >> Well, I always love chatting with you. Thanks so much. Appreciate it. Take care, John. >> Awesome. Awesome, my friend. Thank you.