The Exchanges

Q Now, I want to start a little bit before Tezzy, and I want to start With, you've been at some of the most incredible organizations from your Instacart to your thumbtacks. If we start breaking them apart, actually, with Instacart, when we think about big lessons learned, what was your biggest product lesson learned from the days at Instacart?

A Biggest product lesson learned from Instacart was as much as A-B testing and like high velocity optimizations feel like progress, and to some extent they will be, When you're at a business of that scale, uh, where we're talking like 30 to forty billion in gross transaction volume, the only thing that really moves the needle is, like, new capabilities. So we would try and test a lot throughout the customer experience to optimize conversion, make things easier, make search better, the results more relevant, and it would deliver wins, but the sleeper, huge win, Was when we launched Snap EBT, which is like food stamps, um, to be able to pay with those on Instacart. It became a billion dollar business line in less than a year and then grew up to be four billion. So it changed how we plan and how we think about placing bets from what are the quick win optimizations on the current experience to What do we need to do to unlock the next billion to five billion?

AI assessment note: “Biggest product lesson learned from Instacart was as much as A-B testing”

Q What's the biggest non-consensus product decision you made that didn't work, and what did you learn from it?

A I think the biggest non-consensus product decision that I made that didn't work is how we built TASY, like, the V-one. We took a breath-first approach. The consensus view is, you know, start narrow, Point solution, land and expand. It's the wedge, that sort of thing. What we set out to build is the equivalent of a full life cycle recruiter. And so that necessitated a ton of surface area across the entire recruiting journey from opening an offer all the way up and opening a role all the way up to the point where you're ready to extend an offer. Um, that was the like product scope. So very broad. What was right about that is it's a very differentiated positioning and has helped us get a ton of meetings, um, and customer interest because companies are tired of point solutions and they don't want to integrate another one. What we've struggled with is it's a lot of surface area and it's hard to get, you know, 12 things to a sufficient quality level. The thing that I learned from it is You can't do an MVP for a breadth first approach. And I think Ben, the CEO at equals, uh, has a post on this where it's basically MVP versus full build. And so they have a spreadsheet product and you had to take like a full build approach. We took an MVP approach to a full build breadth product when really we should have Delayed going to market by another six months to get it all to like a higher quali…

AI assessment note: “I think the biggest non-consensus product decision that I made that didn't work is”

Q In terms of everyone being in every meeting, it seems like we've seen this kind of intensification of work culture over the last six months to a year with the rise of AI, bluntly. Have you seen work cultures change when it comes to intensity over the last five years? If so, how?

A I haven't found that work cultures have changed In my observation, based on AI in terms of their intensity, I think of Instacart as effectively pre AI and super intense. A good week was. 60, 70 hours. A bad week is a hundred to a 110. And maybe this is just that like VP level or something like that. But no, at 11 PM, 12 PM, I could ping anyone on Slack and they were already online. Like, and they're immediately responsive. And so it was like an around the clock culture. And I think we're still pretty intense at Daisy. We're not that extreme, but maybe what's changed is I feel like the glorification of like nine, nine, six by certain companies has been like coincidentally timed with also the AI wave. And maybe that's what makes it feel More intense.

AI assessment note: “I haven't found that work cultures have changed In my observation, based on AI”

Q Now, I want to start a little bit before Tezzy, and I want to start With, you've been at some of the most incredible organizations from your Instacart to your thumbtacks. If we start breaking them apart, actually, with Instacart, when we think about big lessons learned, what was your biggest product lesson learned from the days at Instacart?

A Biggest product lesson learned from Instacart was as much as A-B testing and like high velocity optimizations feel like progress, and to some extent they will be, When you're at a business of that scale, uh, where we're talking like 30 to forty billion in gross transaction volume, the only thing that really moves the needle is, like, new capabilities. So we would try and test a lot throughout the customer experience to optimize conversion, make things easier, make search better, the results more relevant, and it would deliver wins, but the sleeper, huge win, Was when we launched Snap EBT, which is like food stamps, um, to be able to pay with those on Instacart. It became a billion dollar business line in less than a year and then grew up to be four billion. So it changed how we plan and how we think about placing bets from what are the quick win optimizations on the current experience to What do we need to do to unlock the next billion to five billion?

AI assessment note: “Biggest product lesson learned from Instacart was as much as A-B testing”

Q You said something brilliant before, which is about founders tend to fire the same three roles over and over. It was a bit of a cliffhanger. What are the roles that they tend to fire over and over? And can you unpack that statement for me?

A Yeah. I've seen this movie at, you know, many companies over the last 15 years. The three roles that I see that get fired over and over are CPO, CMO and head of talent. They typically have like one and a half to two year stints, and then they like move on to their next thing, you know, sort of like quietly, um, and discreetly. So most folks don't understand that they were fired, but when you're in the senior leadership team, you do. Why does that happen? I think it's very similar to the same deal of like any mishire. It's muddy thinking and it's like founders vaguely know what they want. They want the product to work. They want the business to grow. They want to make good hires, but the level of clarity and vision of like, okay, here's the product. I have a crisp product vision for where we're headed. I have a crisp like workplace vision for like Who we are and what our culture is. I have a crisp vision for growth. That, I believe, belongs with the founder, and until they arrive at a place of, like, clarity for themselves, they're not gonna get it through somebody else.

AI assessment note: “The three roles that I see that get fired over and over are CPO, CMO”

Q you not benefit from the combined engagement? And what I mean by that is by having a set weekly time that one time where 10 people come, if you have three come on Monday and three come on Tuesday and three come on Wednesday, the lack of mindshare combined ideas is spread across three days, not one, and you don't get that benefit output. How do you think about that?

A We've solved it both at Instacart and TZ with basically required versus optional attendees. So project team, uh, opts in, they invite themselves. Required for us right now is me and my co-founder. And then the rest of our R and D team is optional. They will opt in if they have availability or interest in the topic. And so we usually end up with, you know, eight or nine people. The way I worked at Instacart, Was product reviews for the consumer app, uh, was me, my eng counterpart, my design counterpart, my data science counterpart, and then my senior directors across the various pillars of the product and then the project team. And so they always have a good quorum to provide feedback, but doesn't necessitate everyone all the time. You know, being in every meeting.

AI assessment note: “We've solved it both at Instacart and TZ with basically required versus optional attendees.”

Q Can I ask you, we said there about kind of Instacart and the importance of building new capabilities over iterations and optimizations. If we discuss Thumbtack, another incredible company, When we think about lessons learned there, what was the biggest product lesson from Thumbtack?

A For me, the biggest product lesson from Thumbtack is you should just do the common sense things and whether you have the data to support them or not. And chasing the data to support why we should do something can actually just be a distraction and a delay. And so to unpack that, what I mean is I recall so much time we spent hemming and hawing in 2016 through 20 18 about Whether we'd be native first, like how important app usage was versus web usage. And we knew it was correlated to retention, but we didn't know if it was causal to retention and to do sort of analysis and study to determine the causality of something takes quite a long time. When at the end of the day, we could have just said, listen, the long-term secular trend is everything's moving towards mobile. Um, More usage, whether they desktop and mobile is better. So yeah, let's just go all in and we'll see what happens. So I guess it helped me see the value of even in like late stage companies of conviction based best that are based on like sensible, secular, long-term trends or, you know, otherwise.

AI assessment note: “the biggest product lesson from Thumbtack is you should just do the common sense things”

Q What did you not know about fundraising that you would most advise someone starting a company who's been an OG like you were in big companies before?

A What I didn't understand about fundraising when we first set out is the game that we're playing is as a company, we're looking for a yes. As an investor, it's in their interest to preserve optionality. For as long as possible. Super easy to get meetings. Meetings mean nothing. And I thought, holy shit, we have all of this interest. This means that like it'll turn into funding any day now. No, it doesn't actually get real till you get like that first term sheet in and then all of it clicks. Everything turns a corner and it's very important to time box it. It's very important to sort of drive towards that first like term sheet in and then run a quick process after That first term sheet is in, but everything else is, I guess, kind of theatrical, um, until that point.

AI assessment note: “What I didn't understand about fundraising when we first set out is the game”

Q Don't laugh. What's the best way to know what you want? I know that sounds super meta, but say we're hiring an engineer. What is the right way to know what we want? Should we write out all the daily tasks? Should we write out the values, the skills? How do I know?

A For me, at least, how I figure it out is I start from the outcomes. That I'm trying to achieve. So what would actual like success be for this person? If they did the job for 30, 60, 90 days, like how would I actually be happy with them? What's true about our culture that they just need to have a, like a level of grit, drive, ownership, that sort of thing. And so I try to articulate that and make chat GPT or, you know, our max Daisy agent output it as a JD. Post it. And I just really need to start talking to people. For me, I learned through talking to the first 20, 30 candidates to get a feel for at least some pattern matching across them of what is an excellent answer sound like versus, you know, not so good. But initially I can't tell the difference. Um, so I have to just talk to a ton of people. Um, and I also look For help from other people that have hired for the role. And so, hey, who do you know who's good at this? And at least I can talk to that person, even if they're not on the market, about what I should be looking for.

AI assessment note: “how I figure it out is I start from the outcomes.”

Q people and you're just picking across, you lose all the, this is a marriage and I'm not suggesting that it's a literal. Do you not worry that you lose personality? Ability to build relationship and we're running a two week process. I'll open up data up on Monday the seventh. We'll do second meetings on the 10th and we'll do third on the 14th and we'll decide on the 21st.

A Yeah, it's a, it's a good distinction. We try to build relationships between rounds and with people that have potential to be high quality partners. That's effectively what you're bringing on is capital and a partner on your journey. And so It's worth building the relationship and getting to know folks outside of the fund funding cycle. But then once you're in it, you're in it by getting to a tight process. It's really like getting to a moment where someone's at least like verbally expressed and committed that they're interested. They're going to go deep on diligence and stuff like that. Then, okay, we'll be flexible. We'll figure out like how to close out the next couple of weeks and Not be a pain in the ass, because, like, we want to be good partners and collaborative about it, but otherwise you can just get strung along and not really know where you stand, because it's in their interest to, like, just kind of keep you in the mix.

AI assessment note: “We try to build relationships between rounds and with people that have potential”

Q What do you mean push, not pull, when it comes to product reviews? Just so I understand that.

A Yeah, so For Taisy, and then eventually, um, at Instacart, when I, like, worked through this, we just have recurring time slots that are, like, always available to book every single week. So we have six time slots for reviews. Anybody can grab one at any time, and it's to get more eyes and feedback on a particular thing that they need feedback on. It could be really early in the life cycle. It could be, like, right before they ship it. It could be design. It could be eng. It could be, Just a requirement, but anybody can grab the time when they need it. With that, since they're selecting into it, it's at the right time for them. And they have a particular question when it goes the other way around, they show up and it's more about optics and managing upward when I'm saying, Hey, bring this project on this date. I want an update and I want to like, make sure, you know, it's on track.

AI assessment note: “when it goes the other way around, they show up and it's more about optics”

Q So they don't know what they want is number one. Any other ways where people seriously make mistakes on hiring in that core first process?

A Yeah. And I've gotten this wrong. So I get sort of why it happens. Um, but a lot of people hire for resume and I found some of the worst hires that I've made. Some of the worst hires I've witnessed are based on incredible resumes. And I think why people get that wrong is you mistake that this person was in a seat at a certain time. And in many ways it was an elevator going up and they were on the elevator. But what we failed to understand was like, what did they do? Like, what did they actually do to like change the trajectory? Did they contribute or just were they at the right place at the right time? And I find that right place, right time can get you really far in terms of getting, you know, big roles and stuff like that. But you get found out after like a year, year and a half. And so Uh, those are, you know, often execs that turn over quickly and it's unfortunate.

AI assessment note: “a lot of people hire for resume”

Q focus on an initial core segment or wedge. When we spoke before, you said something that I really liked. You said most hiring managers hire poorly, and I, I totally agree, dude. I've interviewed, you know, a thousand of the best founders of the last decade, literally. And 50% is the strike rate on, like, good hiring. Ok, so what are the most common ways that people screw it up?

A I think the most common way that I see hiring managers screw up hiring is, like, over-delegation, which then is compounded by the fact that they have, like, muddy thinking. They don't know what they want and they haven't taken the time to like really suss that out. So then you think you're hiring a recruiter and it's the recruiter's job to like hire for you, but the recruiter doesn't really get what you want because you don't know what you want. And so I think the best hires will come from like super active involvement by the hiring manager that they are in it. And very hands-on. But most hiring managers think that they've outsourced it and they are hands-off and it doesn't go well.

AI assessment note: “most common way that I see hiring managers screw up hiring is, like, over-delegation”

Q What do you mean by that? Like, they've done it to get a better fundraising story, to be different? Unpack that one for me.

A Oh, like why do, why would a company glorify nine nine six? I think it's incredible marketing. It's a way of differentiating against sort of the market. Everybody's going to be intense, but we're super intense. And so like, we're going to be like Olympians and, you know, in terms of work ethic, it's like a coded way of saying, Hey, we're self-selecting to like 20 to 26 year olds. So you'll end up with a lot of like junior and mid-level people that are working very hard But also socializing very hard, um, tends to be true in these types of cultures. And at the end of the day, like, 72 hours a week actually isn't that much. Like, for me, at least, we're, like, I've metabolized that over time to be, like, a normal week. But it does help with their sort of recruiter brand, their employer brand, to be, like, We're for the really intense people that really want to win. And so, like, a D-one athlete feels like, oh, this might be for me.

AI assessment note: “I think it's incredible marketing. It's a way of differentiating against sort of the market.”

Q Does that mean that you don't agree with the kind of test it, ship it quick and dirty, get customer feedback, and sticky tape from there?

A Sort of. So I guess, how do I Feel about like MVPs and test and learn. It's when we're building, we'll have varying degrees of clarity as to what the customer needs and wants based on everything else that we've learned from customers over the last, you know, year, year and a half at Taisy. And when we have clarity as to what they want, let's just build the thing. Now where there's ambiguity and we're unsure as to Does this really meet their needs? There's so many steps along the product development journey that we can bring in a customer and like whether that's at the design stage, whether that's with a prototype alpha that was vibe coded to right before launch, we released to just a few customers. And so I think of it as kind of like one swing and I want to finish the swing, but in between we will need to involve customers if we lack certainty.

AI assessment note: “Sort of. So I guess, how do I Feel about like MVPs”

Q see, then, the continued segmentation of roles between design, as you said there, who maybe use magic patterns, but still a segmented role, and then Eng, who still a segmented role, but use cursor to be more efficient. Do you think we have the same, or do we see the bundling where all of them come into one product engine design, and we have this kind of uni superhuman being?

A So the way I see roles evolving is really towards what I think has been like more natural and organic all along. When I think about early days of tech, it's like engineers, and then they were building and all of a sudden, like UX actually mattered because we had like graphical user interfaces and some of them could design and some of them like couldn't design, you know, they couldn't imagine it. And like make it usable. And so a design as a field UX design as a field emerged. Then what happens is product emerges because someone needs to understand the business requirements, why we're doing what we're doing, um, and articulate and prioritize that. But the thing is some engineers could have always done it and some designers could have always done it. And so that's how I transition from design into product management is It's not really different. Like design is a set of decisions about a product. And so those decisions can be made by an engineer. They can be made by a designer. They can be made by a product manager. So I don't see the roles as all that distinct. It's just sometimes you're lacking skills within the team. So then you need somebody to bring those skills with emerging technologies, perhaps You can have more skills basically. So an engineer can get to something like solid design wise when it might've been challenging for them in the past. That's great. And same with a …

AI assessment note: “with emerging technologies, perhaps You can have more skills basically.”

Q focus on an initial core segment or wedge. When we spoke before, you said something that I really liked. You said most hiring managers hire poorly, and I, I totally agree, dude. I've interviewed, you know, a thousand of the best founders of the last decade, literally. And 50% is the strike rate on, like, good hiring. Ok, so what are the most common ways that people screw it up?

A I think the most common way that I see hiring managers screw up hiring is, like, over-delegation, which then is compounded by the fact that they have, like, muddy thinking. They don't know what they want and they haven't taken the time to like really suss that out. So then you think you're hiring a recruiter and it's the recruiter's job to like hire for you, but the recruiter doesn't really get what you want because you don't know what you want. And so I think the best hires will come from like super active involvement by the hiring manager that they are in it. And very hands-on. But most hiring managers think that they've outsourced it and they are hands-off and it doesn't go well.

AI assessment note: “most common way that I see hiring managers screw up hiring is, like, over-delegation”

Q So they don't know what they want is number one. Any other ways where people seriously make mistakes on hiring in that core first process?

A Yeah. And I've gotten this wrong. So I get sort of why it happens. Um, but a lot of people hire for resume and I found some of the worst hires that I've made. Some of the worst hires I've witnessed are based on incredible resumes. And I think why people get that wrong is you mistake that this person was in a seat at a certain time. And in many ways it was an elevator going up and they were on the elevator. But what we failed to understand was like, what did they do? Like, what did they actually do to like change the trajectory? Did they contribute or just were they at the right place at the right time? And I find that right place, right time can get you really far in terms of getting, you know, big roles and stuff like that. But you get found out after like a year, year and a half. And so Uh, those are, you know, often execs that turn over quickly and it's unfortunate.

AI assessment note: “a lot of people hire for resume and I found some of the worst hires”

Q Do you think fundamentally it is so competitive today that if you're an early stage company, you will never have a great person who has contributed with a great resume want to join your company? Like, do you have to find diamonds in the rough? Because it's so hard so early that no one who's been there and done it wants to go back.

A If I think about the founders and experience sort of veteran, like myself and my co-founder, where we're like, we're 40 plus, we've been in the industry for one or two decades. Yeah. Like we can pull those folks. But when I think about, okay, new grad or like dropout, I don't think they actually need it. I think you just need people that are going to be like very competent and very driven and very ambitious, like, and like hardworking and stuff like that. And that'll get you Plenty far to where you've earned the right, I guess, to be able to get some remarkable people to come work for you, but I don't think it's essential to get off the ground.

AI assessment note: “we can pull those folks. But when I think about, okay, new grad”

Q I often find the weak CPOs oscillate a lot around process implementation, and it makes it sound really scientific and structured in a way that they know what they're doing, but in a way just adds a level of bureaucracy. Would you agree with me in that reflection?

A I think that that could be a CPO's job in a certain company. Like, what do I imagine the CPO of, like, ADP Does, I would imagine it's, you know, along those lines because it's a big org. It's hard to wrangle a lot of people. And so most of the job is wrangling and aligning. It's not a very fun job, but like, I'm not here to like necessarily disrespect it. I think it's probably a necessary job, you know, at that scale of company, but it's the thing why I get into product at least. And what I value is The creative, like, aspect of it. It's building something new. Process management is a means to that end, um, and I want to spend as little time on it as is, like, necessary to achieve the end, which is, you know, shipping great product.

AI assessment note: “I think that that could be a CPO's job in a certain company.”

Q Don't laugh. What's the best way to know what you want? I know that sounds super meta, but say we're hiring an engineer. What is the right way to know what we want? Should we write out all the daily tasks? Should we write out the values, the skills? How do I know?

A For me, at least, how I figure it out is I start from the outcomes. That I'm trying to achieve. So what would actual like success be for this person? If they did the job for 30, 60, 90 days, like how would I actually be happy with them? What's true about our culture that they just need to have a, like a level of grit, drive, ownership, that sort of thing. And so I try to articulate that and make chat GPT or, you know, our max Daisy agent output it as a JD. Post it. And I just really need to start talking to people. For me, I learned through talking to the first 20, 30 candidates to get a feel for at least some pattern matching across them of what is an excellent answer sound like versus, you know, not so good. But initially I can't tell the difference. Um, so I have to just talk to a ton of people. Um, and I also look For help from other people that have hired for the role. And so, hey, who do you know who's good at this? And at least I can talk to that person, even if they're not on the market, about what I should be looking for.

AI assessment note: “how I figure it out is I start from the outcomes.”

Q Why? What makes it so successful in a way that founders can learn from?

A I put a real premium on getting to the right answer quickly, and it's hard to find those people, but when you find them that they can take in the context, they understand the problem, and they're going to solve it in a way that is simple, clear, and quick. I will hang on to those folks for like the rest of my career, and I have. So I don't know how to always get there. Like I don't know it on the front end. I only know it through like doing, but I do like working with part timers or fractional people because I get to try it before I buy it. And then sometimes it's like, oh, this really, really works. And so we found that like three or four times already in the last year with some part timers. So I would advise founders try it a lot. If you can get the people to do some work like embedded team members, but You just don't have to have them full time, which also then creates more flexibility about, you know, how we manage cost and burn and stuff like that.

AI assessment note: “I do like working with part timers or fractional people because I get to try”

Q What do you mean push, not pull, when it comes to product reviews? Just so I understand that.

A Yeah, so For Taisy, and then eventually, um, at Instacart, when I, like, worked through this, we just have recurring time slots that are, like, always available to book every single week. So we have six time slots for reviews. Anybody can grab one at any time, and it's to get more eyes and feedback on a particular thing that they need feedback on. It could be really early in the life cycle. It could be, like, right before they ship it. It could be design. It could be eng. It could be, Just a requirement, but anybody can grab the time when they need it. With that, since they're selecting into it, it's at the right time for them. And they have a particular question when it goes the other way around, they show up and it's more about optics and managing upward when I'm saying, Hey, bring this project on this date. I want an update and I want to like, make sure, you know, it's on track.

AI assessment note: “anybody can grab the time when they need it... when it goes the other way around”

Q see, then, the continued segmentation of roles between design, as you said there, who maybe use magic patterns, but still a segmented role, and then Eng, who still a segmented role, but use cursor to be more efficient. Do you think we have the same, or do we see the bundling where all of them come into one product engine design, and we have this kind of uni superhuman being?

A So the way I see roles evolving is really towards what I think has been like more natural and organic all along. When I think about early days of tech, it's like engineers, and then they were building and all of a sudden, like UX actually mattered because we had like graphical user interfaces and some of them could design and some of them like couldn't design, you know, they couldn't imagine it. And like make it usable. And so a design as a field UX design as a field emerged. Then what happens is product emerges because someone needs to understand the business requirements, why we're doing what we're doing, um, and articulate and prioritize that. But the thing is some engineers could have always done it and some designers could have always done it. And so that's how I transition from design into product management is It's not really different. Like design is a set of decisions about a product. And so those decisions can be made by an engineer. They can be made by a designer. They can be made by a product manager. So I don't see the roles as all that distinct. It's just sometimes you're lacking skills within the team. So then you need somebody to bring those skills with emerging technologies, perhaps You can have more skills basically. So an engineer can get to something like solid design wise when it might've been challenging for them in the past. That's great. And same with a …

AI assessment note: “I don't see the roles as all that distinct.”

Q Does the role of a PM change in a new world of AI where vibe coding is so prominent, where speed of development is so changed? Does the PM role change and does the skills required to be a good PM change?

A I think where PMs spend their time will evolve. And by that, I mean, instead of long PRDs and like set of requirements, or like even just the idea of like CEO of the product, I think is really obnoxious and not helpful. So a good PM will be filling gaps. That a team has. And I think that there will be continual gaps. It is hard when you're in it in the thick of building things to be able to zoom out and say, okay, of the 50 things we might do, what are the like three that we should do like next cycle? And how do we prioritize against the customers we have versus the customers we want versus fixing what we've already built? Versus new capabilities. That requires a building a lot of context, and I think that that continues to be where PMs spend a lot of time, but articulating it in some doc, like, it becomes a one pager instead of a 20 pager.

AI assessment note: “I think where PMs spend their time will evolve.”

Q Does that mean that you don't agree with the kind of test it, ship it quick and dirty, get customer feedback, and sticky tape from there?

A Sort of. So I guess, how do I Feel about like MVPs and test and learn. It's when we're building, we'll have varying degrees of clarity as to what the customer needs and wants based on everything else that we've learned from customers over the last, you know, year, year and a half at Taisy. And when we have clarity as to what they want, let's just build the thing. Now where there's ambiguity and we're unsure as to Does this really meet their needs? There's so many steps along the product development journey that we can bring in a customer and like whether that's at the design stage, whether that's with a prototype alpha that was vibe coded to right before launch, we released to just a few customers. And so I think of it as kind of like one swing and I want to finish the swing, but in between we will need to involve customers if we lack certainty.

AI assessment note: “we will need to involve customers if we lack certainty.”

Q I often find the weak CPOs oscillate a lot around process implementation, and it makes it sound really scientific and structured in a way that they know what they're doing, but in a way just adds a level of bureaucracy. Would you agree with me in that reflection?

A I think that that could be a CPO's job in a certain company. Like, what do I imagine the CPO of, like, ADP Does, I would imagine it's, you know, along those lines because it's a big org. It's hard to wrangle a lot of people. And so most of the job is wrangling and aligning. It's not a very fun job, but like, I'm not here to like necessarily disrespect it. I think it's probably a necessary job, you know, at that scale of company, but it's the thing why I get into product at least. And what I value is The creative, like, aspect of it. It's building something new. Process management is a means to that end, um, and I want to spend as little time on it as is, like, necessary to achieve the end, which is, you know, shipping great product.

AI assessment note: “I think that that could be a CPO's job in a certain company.”

Q Does the role of a PM change in a new world of AI where vibe coding is so prominent, where speed of development is so changed? Does the PM role change and does the skills required to be a good PM change?

A I think where PMs spend their time will evolve. And by that, I mean, instead of long PRDs and like set of requirements, or like even just the idea of like CEO of the product, I think is really obnoxious and not helpful. So a good PM will be filling gaps. That a team has. And I think that there will be continual gaps. It is hard when you're in it in the thick of building things to be able to zoom out and say, okay, of the 50 things we might do, what are the like three that we should do like next cycle? And how do we prioritize against the customers we have versus the customers we want versus fixing what we've already built? Versus new capabilities. That requires a building a lot of context, and I think that that continues to be where PMs spend a lot of time, but articulating it in some doc, like, it becomes a one pager instead of a 20 pager.

AI assessment note: “I think where PMs spend their time will evolve.”

Q Why? What makes it so successful in a way that founders can learn from?

A I put a real premium on getting to the right answer quickly, and it's hard to find those people, but when you find them that they can take in the context, they understand the problem, and they're going to solve it in a way that is simple, clear, and quick. I will hang on to those folks for like the rest of my career, and I have. So I don't know how to always get there. Like I don't know it on the front end. I only know it through like doing, but I do like working with part timers or fractional people because I get to try it before I buy it. And then sometimes it's like, oh, this really, really works. And so we found that like three or four times already in the last year with some part timers. So I would advise founders try it a lot. If you can get the people to do some work like embedded team members, but You just don't have to have them full time, which also then creates more flexibility about, you know, how we manage cost and burn and stuff like that.

AI assessment note: “I would advise founders try it a lot. If you can get the people”

Q You mentioned the 15 hours a week that the person from Slack partakes in with you. How do you advise founders on part-time advisors in that respect, which you see more and more of?

A As a founder, I find that everyone wants to give me advice and get paid for it, and for the most part, I don't need it. There's a few key areas where advice has been useful as it pertains to fundraising or go to market, because that's like, it hasn't been a strength, uh, or part of my background or my co-founder's background. But by and large, I find advice to be like cheap. I can get it for free if I want. So like, I don't know why I would pay for what is useful is people that'll put in the work. And so this guy, Colin, um, has been advising companies since leaving Slack. I got connected to him through my wife, who was an engineer at Slack and worked with him. And my pitch to him was, Hey, wouldn't you like to get in the game a little bit? Um, and instead of just, you know, give advice, like, can we do some work together? And it's a nice change of pace. It's hard to find like a way to contribute when you don't want to sign on full time. To, you know, do a 15, 20 hour a week, like, PM job. And it's been remarkable, actually. So I, I've worked with, like, a hundred PMs probably over the course of my career. I like his 15 to 20 hours more than most of their, like, 50 to 60 hours.

AI assessment note: “what is useful is people that'll put in the work”