Freemium Is a GTM System, Not a Pricing Tier

Machine-generated transcript from the episode video. Speaker labels are not included and some names and product terms may be transcribed phonetically.

[0:00] I've learned that the best founders are always selling. Welcome to the GTM Vault podcast hosted by Rick Kleta. We uncover the strategies, the pivots, and the breakthroughs turning startups into giants. Let's crack open the vault and find out how. Today's guest has sat in every GTM seat from VP of product to CEO to board member and now he's leading the charge on PLG and AI powered growth. Dave Boyce has sold companies to Oracle, Amazon and Ora launched products that reach profitability without a sales team and teaches the first ever MBA course on productled growth. He's the executive chairman at Winning by Design, a board member at Forester, and a trusted adviser to founders building the next generation of SAS. In this episode, we dive deep into what real PLG looks like, how AI is transforming RevOps and GTM teams, and the new playbooks founders should be running in 2025. Let's get into it. Dave, you've led multiple products to profitability through pure

[1:08] PLG. What are the biggest misconceptions founders have when they try to go PLG? You're not wasting any time. You're getting right to it, Rick. It's um I mean, look, one thing, yes, yes, I've built multiple products and um and also one thing we have to understand is everything's changing right now. So, I'm going to say some things about like how it's worked in the past and we're going to talk about how it's working in the future. But, um you know, the the number one thing for me, I think a lot of founders get it, Rick. Like I mean if you think about like right now is like an AI it's almost like an AI native founding kind of environment and when I spool up a a company I mean this was true you know when I was founding companies like 15 years ago but every single platform that you use to spool up your company from you know from you know Google Docs to HubSpot to Carta to like it's all it's all self-service. You never talk to any humans when you're spooling up your own company with your own platform. So, why would you be building a platform that requires something different that you're selling to your customers? That said, the biggest thing that I've seen is that founders skip the founder selling mode.

[2:20] They're like, "Hey, I'm going to build a self-service product." Yeah, cool. I'm going to make it really easy to discover and use. Yeah, amazing. And therefore, I will never have to sell. Wrong. You actually will have to sell like your first few users and you're going to sell. You're going to be the number one salesperson, the number one kind of customer support person, the number one you're going to be learning and tuning along the way and you're going to put your own human self into the mix to understand and develop empathy for those customers. So, you can't skip that part. You can't just build it and then wait for the customers to show up. Even if you have great mechanics, you actually have to get into the conversation. What's the difference between a PLG motion and a PLG company? Can you have one without being the other? You can try. You can try. I mean, the thing about PLG is it's a mindset. It's a mindset. I, you know, it's a mindset that um embraces weird things like like empathy and generosity. So yes, you can, you know, mechanically you can build a PLG motion inside of a nonPLG company, but if you do it right, if you're really developing empathy for your customers and you're really understanding what their job to be done is, and you're really kind of meeting them where they are with generosity, like, hey, that's okay, that's the problem you're trying

[3:35] to solve. Cool. Take this. All right, what do I owe you? Nothing. When do I pay you? Never. That's the PLG ethos. And once it once it takes hold in your company, it will also make the rest of your company better. Let's say you have a second go to market motion that's salesled. That can also be empathetic and gener and generous. Just because there's humans involved doesn't mean we have to extract our humanity. We still want to be empathetic and generous whether the product is leading, whether AI is leading or whether humans are leading. When does it make sense to layer sales onto PLG? And what does it kill momentum? When does it kill momentum? Yeah, I mean that you know that that would be a great problem to have. Rick, I've got momentum on PLG. Amazing. That's hard to do. That takes takes months and years of iteration to get kind of momentum on PLG. So, you definitely do not want to kill that momentum once you get it. Um cuz you worked really hard for it. You iterated your way into it. Two things. One is, you know, there will be a point where your growth rate starts to attenuate and you're going to want to prepare for that and you're going to want to add some sales assist in. Um, but I don't like going too early. Like if if everything is up and to the right and I've got kind of I've got acquisition, activation, and

[4:48] monetization all working. I've got good renewal characteristics and I've got momentum and my ARR stack is going. I'm only going to add a sales assist motion kind of when I start to see the growth attenuate or when I or when I'm like that is really stable. I'm up to $10 million in ARR and I'm going and I can afford to then go put my creative energy on putting a sales assist motion on top of it without messing up what I'm already doing. The main thing I see is going too early and and usually you go too early because you're getting pressure from all sorts of outside forces like your VCs are like hey if we can do you know we if we can sell the name brand customers why wouldn't we? If we could sell million-dollar contracts why wouldn't we? One of my other companies over here did this. I would resist that until you feel like you really have a solid foundation under you and kind of 10 million or more of a arr and then you can afford to go iterate your way into a next GTM motion. What are the critical early indicators that a PLG motion is working or failing even before revenue shows up? That is a really really critical question and we get impatient, right? Like we want the revenue but like that's our scorecard.

[5:54] That's our y-axis. Hey, where is where's ARR on my y axis? No, no, no. The first thing that's going to show up is usage retention. The very first thing that's Yeah, that that's what I like to optimize for first. So, have I developed a solution to a problem that humans have that is so compelling that they'll keep coming back and coming back and coming back. If I can retain their usage, retain their attention, now I got something. Dollars will follow, but they're not going to come back and log into something that doesn't deliver any impact. they are going to come back and log into something that is delivering impact. So once I have that right and I can and I can measure usage retention on a daily, weekly, monthly basis. I don't have to wait for a year and I certainly don't have to wait for a monetization event. That's what I track first. You've launched dozens of products. How has your GTM playbook changed in the AI plus PLG era? Well, it's faster, Rick. It is faster. You can prototype faster. You can launch MVPs faster. All of the kind of initial stack for putting a product together and measure and monitoring patterns like usage patterns, they're all pre-integrated and a lot of that is AI facilitated. Embrace all of the AI

[7:06] native GTM tactics like all of them. And think in days and weeks, not months and quarters. Almost anything can be made self. So that's that's number one, faster. Number two, almost anything can be made self-service. Like you say, hey, my product's too complicated. Really? Is it could it be explained through some sort of a generative AI agent or bot chat agent or voice agent or help agent? Probably could I could probably smooth over some of the complicated pieces of onboarding with uh AI. So almost anything can be made self-service these days. Faster, more self-service. And then the third thing is like don't don't pay a human to do something that a that a robot could do. Like in the age of AI, we can task robots with doing almost any piece of the GTM motion. And why would I pay a human to do something that a robot could do? Humans are amazing and they can uniquely handle certain things. So, let's have the humans do those things and let's have the robots do the things that are scalable, repeatable, and mechanical. Your all-in-one AI assisted

[8:12] GTM, what's one workflow every SAS team should be automating today, but still isn't. I love the way you phrase that. Every SAS team, every SAS team, Rick, every SAS team has a renewal. Every SAS team. And and that renewal assumes I already have a contract. I already have contract terms. I already have a configured application with features. I've already deployed because I already did the work to get the initial kind of year worth of um of usage in place. And now all I'm trying to do is extend that into the next year. Automate the renewal. Automate the renewal. It's simple. You don't have to do anything other than just allow the customer to accept re-accept the terms. I don't have to renegotiate the terms. I don't have to reestablish anything. I don't have to reconfigure anything. Automate the renewal. As soon as you get automated renewals, then you're going to get hooked on automation. You can go to the next uh pieces of GTM automation. And this applies to enterprise as well, right? Not just low-end SAS. So that I can automate SAP renewal like the most complicated software in the world. I I went through all the pain of getting it established. Now all I got to do is automate it. Absolutely. Your definition of a GTM engineer. What is it? And why

[9:24] do you think they're replacing traditional revops? Okay. So for listeners, Rick Priest sent me some of these questions. So I thought about this a little bit. I'm skeptical. I am skeptical of how we talk about GTM engineers. Okay. I mean I I I love the concept as far as it goes, but I don't love how far it goes. I almost think sometimes in the public discourse we kind of say, "Remember a growth hacker? Well, now it's called a GTM engineer. We're going to go find like we're going to go find the clever ways to, you know, to spool up play on some piece of our GTM motion and we're going to call ourselves GTM engineers and we're going to lean into these arbitrage opportunities." I think that's all great, but against what playbook, against what architecture, against what kind of unifying theory of the case. I want a GTM architect and then a GTM engineer to go execute the pieces of that. I want to understand like what is the customer journey beginning to end. Where does growth come from? How does growth compound? And and I've hired, you know, I've hired um GTM engineers and growth hackers before who don't have any theory of the case and

[10:32] they're just running fast after like whatever the next shiny object is. That thing is not that's not a stable growth architecture. What I want is someone who can do both. I want to lay down a really stable kind of measurement framework bedrock and then I want to go run experiments on top of it. So I would think about GTM architecture as much as I would think about GTM engineering. Yeah. Now that's that's a really good point because everything uh boils down to the architecture and the foundation, right? If you're using a certain suite of tools, bringing in bringing in some GTM engineer to kind of come in and change things is not it's going to mess up the alignment amongst the teams to begin with. Maybe taking a few steps back and starting off with like the architecture and yeah um and Rick, you know, when I say architecture, I'm not necessarily saying technical architecture. I mean that factors in but it's really like my information architecture of like you know how how do I track a customer from awareness to a to activate um acquisition activation monetization renewal expansion referral like how do I track all of that what is it like how and then how does growth and now once I'm and how do I measure it and now I can run experiments once I know how to measure if I'm running experiments

[11:48] without measurement like they're useless but if I can run experiments against measurement, not only in their own silo, but against the whole growth architecture. Now, now I can really get now to your point, like I'm not going to mess that up by running experiments with a with a um GTM engineer if I if if I've laid down that bedrock. And if you were building a modern GTM stack from scratch today, what three AI native tools would be foundational? Boy, there's a lot of this depends on the type of company you're uh you're building. I mean, I would certainly not. Anthropic just announced last week that they're up to 90% of their code being written by their own agents. 90%. 90%.

[12:30] That is like crazy. Like I really want to get in there and see that. But, you know, I certainly don't want to be at 10% of writing my own code. So, I I would I definitely would be leveraging AI and engineering. I definitely would be for integrations for you know for for raw code generation for documentation QA test like all of that I would do with with AI in terms of GTM stack which is I think the question that you asked I would uh kind of depends on what motion I'm running but almost all the companies that I work with are on the early stages are using HubSpot and HubSpot just did a native integration with OpenAI which is amazing so now I can if I'm running a GTM stack no matter what my motion is, but especially if I'm having conversations with customers. Well, I'll get to that in a second. Anyway, I can use Open AI to analyze that that entire kind of customer journey like we were talking about, which is great. Then I want a noteaker. So, this is not necessarily for PLG. If it's PLG, my signal is going to come from somewhere else like heap or amplitude or a mix panel. I definitely want to track kind of engagement and and I want that to be all interrogable against that customer journey architecture. If I'm having conversations, I want a notetaker. I want a notetaker listening in on every

[13:43] conversation, transcribing it, normalizing it to some sort of a framework like spiced. And then I want that to be interrogable, too, which open AAI makes possible if you're on HubSpot. So that kind of signals at the bottom interrogability with kind of open AAI or some other tool. There's other note takers have their own interrogation tools. and then kind of HubSpot to make it all fit into a endto-end um customer journey would would be a place to start. There's lots of other stuff. It depends on the motion you're running. What are the risks of overmating early stage GTM? And how do you balance speed with signal? Uh no risk, but like like like we said before, I don't ever want a human to do something that a robot could do. That being said, I think the key word you said is speed. Like if I'm just heads down trying to automate every single thing and I never ship, um, that's a problem. And I never talk to anyone, that's a problem. I never get a feedback loop on learning, that's a problem. Eventually, I want to automate everything. And I don't think there's any risk to that. But if it means that I'm staying in my cave and not shipping because I don't have it fully automated and fully perfect, that's a problem. And if it means that I'm not talking to

[14:54] customers, that's a problem. like if I'm excusing myself from actually developing humanto human empathy with my end users that's a problem. So really every there is almost no moat left Rick other than speed. So I just want to go fast. Um, and in my wake, let's say I discover that formula ABC is working on a GTM basis, like this kind of content with this kind of ICP with this kind of call to action and this kind of onboarding flow that's working. Cool. As soon as I know it's working, boom, I want to automate that and move on. Then I want to go find the next thing that's working. Boom, I want to automate that and move on. I don't want to keep people powering things forever. And there's no risk of overmating as long as I'm only automating the stuff that works. How do you see AI transforming customer success and onboarding in a PLG world? So a lot of onboarding is super critical. That time to first impact that initial experience that you have, you know, you you you all your hopes and dreams from reading the marketing copy. Cool. You know, try it now. You click the button from that moment until you get first impact is a really really critical part

[16:04] of the journey. and we see most of the seeds of a successful kind of long-term retention and expansion cycle planted right then. Same thing with the seeds of of defection or loss of interest or I can't get first impact before I lose their attention. AI can really help there especially with complex product. I've seen some instances Rick where in the old world where I' my my product is complicated enough that I need to give the customer a vision of what could happen. So, I'll in I'll literally insert a video before they start launching the product. And that video will get in their head like, "Oh, that looks easy enough. Now I can start." But AI can do a lot of that work for me, too. It can walk me through an interactive demo. It can help me do steps 1 2 3 of a configuration. It can ask me a few natural language questions and then boom, show me a draft version of the website I was trying to build or the report I was trying to build. So AI can get me to that kind of place where I'm envisioning myself as successfully using the product or even actually successfully using the product faster.

[17:05] You know, I had a customer recently. I was doing GTM alignment for them and one of the first kind of areas that I identified as needs work was only one of the personas was really engaged with using the platform. Even though they had glided upstream, they had multiple stakeholders involved with the platform, but usage it was just so low across the board except for one of the personas. So, and then I realized the onboarding was just standard for all of the different personas. Someone with this job title is not going to want to use this platform. And if they had a personalized onboarding process and every time they opened the platform, it prioritized the features that they most were going to use the platform for, then it increased the usage. And my question to you is this. Yeah. Do you think it's possible nowadays to prioritize this considering it's such a crucial component of the whole acquisition journey, right? onboarding is is even done until the aha moment or some value is derived. Right? So, do you think it's possible to start incorporating that

[18:17] from day one opposed to, you know, we finally figured out our low-end SAS where we've started to glide upstream now let's throw this extra experiment into the bits to to increase usage of other stakeholders within the platform. Yeah, it's possible. Well, if I understand the question right, I do think it's important to get it right for that first persona first. Like being on being on average having an average set of value propositions and an average, you know, collection of features for the average customer with the average job to be done. That's not real. None of that exists. There is no such thing as an average customer who's going to use average features. there's a specific persona and you found one in that company who had a specific job to be done and they were actually really succeeding at it. So I want to find that and then I want to once I found that then I want to go find the next one and optimize for her and then I want to go find the next one and optimize for her.

[19:13] And the thing I got to be careful about is I want you know I do want to go for the center of the bullseye in terms of my target market. Like if mostly what I'm going after is creative you know like graphic designers great. Let me figure out where they work. let me figure out what their job to be done is and let me meet their needs. If the secondary person I'm going for is a um marketer who just has to get up simp kind of simple designs in place, but that's a really big kind of market. Maybe I have a different onboarding, you know, flow for her. And if my tertiary one is the is, you know, somebody sitting at home who wants to, you know, do a poster for their kids' school or whatever, okay, great. That's probably a different onboarding flow, but I want to take them one at a time. You've helped scale companies at 50% growth rates and also had to restructure teams. What's the signal to pull back risk double down? Where were you when I needed you to ask this question?

[20:08] Um any anyone in our seats Rick is an optimist. That's why why we're doing what we're doing. We're leaning into the future. We're like curious. We're intellectually curious. We're we're also willing to deploy energy against hard problems more so than your average person. So often when we run into obstacles, we just treat them as something that we got to overcome. And that's a great thing, but once in a while those obstacles that we're really trying to overcome are intractable or they, you know, or it's something that, you know, I had a mentor um who once told me like, "Dave, I've been working on this company for like a year. it if I can just tell from the way you're talking to me. It feels like you feel like you're pushing a rock up a hill. I was like, "Yeah, I feel like I'm pushing a rock up a hill." He's like, "That's not how it should feel." Like, and the answer when you're pushing a rock up a hill is not push harder. You got to look for water flowing downhill. You'll know it when you see it. So, you know, at some point, Rick, um, and it when it feels like you're pushing a rock up a hill and just you've tried everything, you've deployed all your creative energy and like, you know what? Let me work on a different surface or a different problem or with a different let me see

[21:17] if I can get to where water's flowing downhill. Because as soon as I got water flowing downhill, now all of the challenges are about growth and about acceleration and about extending kind of that success versus just continuing to struggle against that problem that's just, you know, for whatever reason, it's intractable. And you can feel it. I don't have like a a metric, but you can feel it. You know, when you're fighting against the tide and it's time to try a different tag. What separates a great GTM leader from an average one in today's world of lean teams and AI leverage? I just came back from two weeks of working with private equity CEOs and CRO and CPOS and you know related to your questions at the top around AI you know the number one thing that we all agree with is that what got us here is not going to get us there.

[22:09] Let's say I'm a, you know, a successful kind of, you know, second time, third time executive or founder. Whatever got me to my this level of success, the game has changed going forward. You know, we're watching we're watching tools like lovable and and replet and cursor grow to 10, 20, 50, 100. I think we just heard last week 500 million in revenue with teams of like 30 or 50 people. That is a completely different playbook than the one that I learned as I was coming up. Completely different. So if I want to be a successful GTM leader today, I got to get that in my head. Like what got me here is not going to get me there. Which means the job is not management. The job is not running the playbooks that I learned in my current you know my past company running them in this company. That is not the job. The job is architect. like I got to figure out what is what is the appropriate GTM motion that's going to work. And that's going to take some experimentation, which means as a GTM leader, even though I've already learned how to do this, even though I've already become successful, even though I've already proven myself, quote unquote, I got to go back to curiosity, roll up my sleeves, start running experiments, and

[23:20] figure out how I'm going to compete with the AI native companies that are redefining the game. as a board member at Forester, what trends are you seeing in how enterprise buyers approach PLG products? So, this will not surprise you, you know, and Forers's published research on this, you know, that people that consumer the the information asymmetry has flipped. It used to be that sellers knew more about their product than their buyers. Now buyers know more about the product than even the sellers who are selling it because we can get because we as buyers can get information way more reliably. We can go to our we can go right into any of the generalized models with claude or open AAI. We can ask for a comparison table of the product I'm considering plus its three top competitors. I can find things like reviews and pricing and timed implementation and use case appropriateness. I can find all that out as a buyer probably more efficiently than in the conversation with the seller. So, um the so I' got that's a trend and I just got to embrace it. um that the buyer is going to have more information than me.

[24:25] The buyer is going to get that information from lots of places that are not me and the buyer is going to control the purchase process, not me. So, as soon as as soon as I can embrace that as someone bringing something to market, what I really want to do is pave those paths like those are paths worn in the grass because the buyer's going to her network to get information. The buyer's going to to um Genai tools to get her information. The buyer's going to review sites to get her information. So rather than try to stop her from doing that and bring her back into my controlled process, that ain't going to happen. Let me just go pave the paths that she's already traversing and make it easier for get her to get information from her network. And so I'm going to create communities. I'm going to create information. I'm going to build heroes. I'm going to build stories. I'm going to build an information network. I'm going to lean into LLMs. I'm going to make sure that all of those sources are pro have full access to the information about my product that I want out there.

[25:19] So, I'm actually going to lift lift the gates and get my information out there rather than trying to close the gates and control information. And when you work with clients, Dave, how do you kind of incorporate the product team there to come in and build in say viral loops into the product to um to help with that growth? Yeah. So, for a for a larger company, the biggest biggest unlock Rick is the creation of a growth team. So the first and and this is a bit it sounds it sounds a little crazy because as startups we you know a startup is a growth team. It's got an engineer and a marketer and a um and an analyst and a and a UX designer like all on the same team all working on the same thing. But when you get big those people end up getting siloed. The engineers are over here the product managers are over there. Designers are over there. The marketers are over there. Um and so and they're all kind of working on their little piece of the puzzle. As soon as you can kind of peel off kind of a a small team of like five to eight people that includes a marketer and an engineer on the same team, that's a growth team that can now iterate really fast. That can iterate fast against growth objectives. And now instead of that team having the objective that the engineer

[26:30] is used to, which is essentially I got to ship I got to ship, you know, one story this week against my kind of, you know, story points in Jira. Like no, no, no. The objective is I got to go get activation rate from 7.5% to 8.5%. Create a growth team that has marketing and product and engineering all in the same team. And then I start giving them GTM objectives like actual business objectives. That unlocks a whole bunch. And um it's hard to do cuz not every because big companies you know even kind of 20 million and above companies you know may not be operating that way but as soon as you do it unlocks a lot. You teach the first MBA level course on PLG. What case study or concept gets the strongest reaction from your students?

[27:16] Here's the thing, Rick. Like, so these students are in their 20s and early 30s, mostly mostly late 20s, and they're just humans just like you and me, and they experience products just like you and I, and they buy products just like you and I. So, you know, they all have Slack, they all have Zoom, they all have Canva, they all have they may or may not have Docu Sign, but they they've probably used Lovable or they they have experience with the real world and and then they have, you know, they have Instagram and and, you know, all sorts of consumer kind of apps as well. Um, they know that self-service is a thing and they know that premium is a thing and they know that free trials is a thing. What they haven't done is kind of translate that into the business world. But what's happening is those people are going to have jobs. They're going to work for GM or Ford or Dell or Microsoft, you know, in one year and then a year later they're going to have be evaluating tools and purchasing tools and bringing tools into their company.

[28:12] They're not going to forget that they're humans. What um and and sometimes when we go into an MBA program, I did an MBA, we're like, "Ooh, in my personal life I'm a human. In this life, you know, I'm a robotic kind of, you know, steely business professional." No, like that's not right. So, as soon as I get them connecting the dots and I show them case studies like Dropbox or a Docu Sign or even HubSpot, a lot of them have experience with HubSpot. How did HubSpot introduce um PLG into their go to market motions? They're like, "Oh, that makes a lot of sense. That's how I buy software. That's how I experience the world." Now, we're just paving that over and make it easier for customers to be humans and solve their problems with the self-service kind of, oh, that makes a ton of sense. And then we and then we tie it into that measurement layer so that they can kind of measure it and monitor it and run experiments against against an information architecture that's pretty robust like whoa this is all like making sense to me. And then we run and then we we give them case studies with kind of progressive disclosure data for a certain company that was iterating their way into the future. And these students just start being like I can see the patterns like this is great. I'm going to optimize this thing. We're going to unlock something. Ooh, it worked. That's great.

[29:21] Now let's go. Oo, it didn't work. let's try something different. So I put them into the shoes of the of the uh kind of the growth engineer as we were talking about like a growth architect and I just let them let them go and as soon as they have that fundamental understanding they they take to it like fish to water. You've sold five companies and advise many more. What's one GTM pattern you consistently see in successful exits? Scalability, Rick. Repeatable processes that scale. So, I have been involved in companies that have gotten away with not being scalable. Like, every deal is a snowflake. Every every set of contract terms are custom. Every deliverable has a lot of customization. But that doesn't happen anymore. I don't see that happening anymore. Like, the companies that are successfully exiting are building a scalable, repeatable machine against a large TAM where I can just kind of, you know, turn the crank on something that is going to continue.

[30:18] um the scale. If I run out of customers, that's a problem. If too much kind of customization is required for every single customer acquisition, that's a problem. If too much kind of time and attention is is required to keep renewals flowing or to keep expansions going, that's a problem. So, I just want I want something that just works every time. Center of the bullseye or not every time. I want something that work I can do the same thing every time and it will work 70% of the time. If I can find that formula, it will scale. Hey Dave, so you've published a book called Fremium on PLG productled growth. Tell me more. Yes Rick, it's been a lot of work like just trying to figure out those patterns of success, interview successful entrepreneurs, you know, talk to people who added PLG later. You know, they started in one way and then they added PLG later start talk to people who are PLG from the beginning and even it even leans into some of these trends with AI unlocking more and more self-service within our go to market motion. So yeah, the book's called Fremium. It's coming out from Stanford University Press in August, but it's available for pre-order right now anywhere that you buy books. Amazing.

[31:25] So, for everyone out there, is there a website we could go to, Dave, to pre-order? Well, if you go to my Substack, productledg.com, I have all the links, but you know, from Water St., Barnes & Noble, Amazon, Target, like anywhere you buy books, you can just search premium and it'll come up. That was a master class in modern goto market. Dave doesn't just talk PLG and AI. He's built it, scaled it, and teaches it to the next generation of software leaders. If you're a founder or GTM operator navigating lean growth, this is one of those episodes to revisit and study. Full episode playbooks and more at gtm.wiki. And if you want early access to exclusive frameworks and tools from guests like Dave, join GTM Vault Pro. See you in the next one. Thanks, Dave.

[32:13] Thank you so much, Rick. It was fun. Tech founders and VCs careers lessons. GTM