Ignite Product: How Continuous Discovery Builds Better Products with Teresa Torres | Ep215

I'm really into using AI to create deliberate practice. Deliberate practice is how do we practice the sub-elements of a skill with expert feedback. So we can create practice opportunities by interacting with it. Like interviewing, for example. We now have all the technology we need for you to practice your interviewing skills by interviewing an LLM. We can even do it over voice. We can even do it over what looks like a Zoom call with a real human. And then what's great is I already have an interview coach that can evaluate how well you did in that interview. So we're getting to the point where we can just provide infinite practice. Like this is something we could never do in a live talk course, right? Because we've got one or two instructors and a number of students. But like now you can practice as much as you want and always have extra feedback. And so I'm really excited about that. Hey, everyone. Welcome back to the Ignite Podcast. Today, we're thrilled to have Teresa Torres on the mic. She is an internationally acclaimed author, product discovery coach, and founder of Product Talk Academy, best known for her book, Continuous Discovery Habits, and for helping teams at companies like Spotify, Capital One, and others build discovery into their weekly rhythm. She teaches thousands of product people how to align business outcomes with real customer insights. Test assumptions early and make discovery a habit rather than a checkbox. Thanks for coming on. Thanks for having me. Telling you this before we started recording, but I'm a longtime fan. I mean, I remember reading all your product blogs. It must have been 15 years ago at this point, back in the early. So it's like meet your hero day here on the Ignite podcast. Thanks. It has been a long run, but I'm still having fun. So that's all that counts, right? That's amazing. Yeah. well i'd love to get your kind of your backstory what's your origin story it's sort of funny i got introduced to human-centered design as an undergrad in college and i naively thought that's how business worked and so i went into my first job thinking like we're going to be talking to our customers and we're going to be testing our designs and everything's going to be great and then i very quickly learned that's not how business works at all i worked at a number of really early stage startups early in my career and that was really fortunate I know a lot of people early in their career, they chase the big name logos and that's valuable too. But I think what I really liked was I got to play a lot of hats early. I got sort of firsthand exposure to business early. And so I've always sort of blurred the boundaries between even front-end engineering, design, product management, really influenced the way that I think about the world. and probably after i think i worked at four different companies and after the seeing the same pattern everywhere i was like okay maybe i should just focus on trying to get product teams to talk to their customers a little more often and so i think let's see we're in 2025 14 years ago i started working as a product discovery coach yeah and it was uh not too long after that probably a year or two that i ran across your online musings you know i was entering product around 2012. And it was very much like you said, no one was talking to customers. I worked at Rocket Fuel, which was an ad tech unicorn. And I got there and it was just complete waterfall, just complete waterfall. And no product managers were talking to customers. And I used a lot of your teachings and others to say, hey, no, let's actually get out of the building and really talk to people and figure out what their pain points are. And it was just so surprising that it was very much in hindsight, like a technical YAM role. You know, you just kind of take the requirements from the sales and product marketing teams and you just go build what they tell you to build. Yeah, unfortunately, that's still very common today. I mean, I remember being blown away when I worked at a startup where our roadmap was our VP of product literally just captured everybody's product ideas. I mean, everybody's product ideas into a spreadsheet. And then he just reorganized the rows based on the flavor of the week. And I remember just being blown away that like, really, this is your strategic decision-making? Like, I don't really get it. And so what's good though, is I do think the product craft, the craft of building products, how we decide what to build is evolving. We've made a ton of progress. It still feels glacially slow, but if we take the long arc. we have amazing tooling we've made many many more teams are talking to customers on a regular basis we're starting to test our ideas so it is nice to see a little bit of evolution in the industry yeah how do you advise teams how often to talk to customers because a lot of times you go you figure out the problem you're like okay now we got to go execute on this and it's not going to be a two-week sprint to do this thing it's going to be a quarter you know or two what is the right cadence does it kind of depend on what you're building and what the kind of the cycle time of what you're building entails or yeah how do you kind of advise teams to think about that they get even to your comment about it's not going to take two weeks it's going to take a quarter i think a lot of this happens because it's so easy to think in big projects you know all of business works on it from a project mindset but i think they're really the real power and product comes from embracing what's often called a product mindset i like to think about it more as a continuous mindset it's this idea that like our product is never done there's never going to be With a project, there's an ending. We complete the project, we're done. With product, especially digital products, like it's always evolving, like we're never, ever done. And so if we instead adopt this continuous mindset, what we have to do is we have to learn how to take these big, giant, monolithic projects and start to think about them more continuously. So how do we break that work up into bite-sized chunks? And of course, we've always been doing this, right? Like engineers work in bite-sized chunks. The difference with a continuous mindset is we're working in bite-sized chunks where each chunk is valuable to the customer. And that's the hard part, right? So it's easy to say, like, it's going to take it three months to build this thing. And at the end of the three months, we'll deliver this really valuable thing to our customer. It's a whole nother thing to say, okay, well, let's take this three months of work and think about how do we carve it up in a way that each little piece creates some value to customers. And the benefit of this is it means that we don't have to wait till the end of three months to create some value for customers. They're getting value all along. So I think to learn how to do that really well, we actually have to be talking to our customers continuously. We have to understand instead of thinking about it as like a big three-month project, we're kind of thinking about it as like, let's say it's 10 nine-day projects or even better, 24 and a half-day projects. I know there's weekends in there. I'm just trying to make my math easy in my brain. So like maybe it's 23 day projects. And then I think that if we think about it as like, we always want to get feedback from customers on our projects. Well, now we have teeny tiny projects. So we need to get continuous feedback. And that also is what helps to like, make sure each chunk is actually providing some value to customers. Yeah, I love that. How do you advise, you know, this is very much a startup podcast, you know, we invest in early stage startups, you know, precedence seed, you know, which are basically founder led product teams, right? know it's the founder sometimes just a business co-founder and technical co-founder as you're handing off the reins it seems to me as the organization grows in complexity it slows it down right it slows this continuous discovery process down how do you advise you know founding teams that are starting to grow the product and engineering and design orgs to kind of maintain what they had when it was just a two pizza team as amazon would say yeah i think what's hard about startups and this is where all of my full-time employee experience was is like a startup founder is a unicorn in and of it themselves right like they're crazy enough to think that their vision in their head is the right thing is the right view of the future and they're willing to bet a lot of their own oftentimes money definitely sweat equity into making that vision come into like the near future and that's amazing and i feel like good founders have a ton of conviction and that's a big part of their strengths. The challenge is our strengths are often our weaknesses. And when it comes to discovery, that strength of like, I have this really strong view of the future and I'm willing to bet everything to go after it, makes it really hard for you to see how the how you're going to get there might be wrong. And so I encourage founders to like have a strong conviction about the future, but be really flexible in how you're going to get there. And this is where like discovery, Actually, let's think of a really famous example of this. Dropbox released their launch video where it was just this picture, this vision of you have access to your files everywhere. That sounds so silly now because we've all had access to our files from everywhere for so long, but when it came out, it was revolutionary. I remember before that, I grew up in the Napster generation. We already had this idea of peer-to-peer and files everywhere, but it was hard. And we weren't quite sure if it was safe, let alone legal. Dropbox comes along and they look like this, like they're going to build this thing that is clearly legal, clearly safe, and you can still have all those benefits. And I think it just really resonated. Now, Dropbox had to make a million decisions along the way to make that feature come into reality. And if the founders of Dropbox had been really stubborn in all of those decisions, they probably wouldn't have been successful. To get to a successful product, we have to be really flexible in the how. And we probably have to try 400 different ways to get there on a million different decisions. And I think what's hard for startup founders is that sometimes that conviction we have about the future, we misapply to the how. And then we build the wrong product because we're stubborn about our view of how it should work instead of our customer's view of how they need it to work. And so I think the first thing is we have to get through that awkward, how do we separate? our conviction conviction is good so we talk about founder mode but getting micromanaging in the how can be really problematic okay i think maybe what was behind your original question was this idea of like eventually a startup is going to outgrow the founder being the product manager and how do we like scale ourselves and how do we make sure that good decisions keep happening I think, again, it's really important not to confuse those things. Strong conviction about the view of the future while being flexible on the details. And I think we have to really communicate to our teams what is non-negotiable about that future vision and then let go of literally everything else. Now, that doesn't mean we can't hold our teams to a high standard of quality from a craft standpoint and from a design standpoint and from a taste is the word. of the day lately right like i do think that it is a founder's job to maintain that standard of quality but i think it like the way that we scale ourselves in a startup or not is that we're really clear about the non-negotiables that view of the future and we're really flexible on everything else because we're going to get the best work out of our teams if we give them if we empower them if we give them a lot of space to go and explore and you know what's funny is right now we're going through this period where the industry is like swinging back a little bit like there's this push to move back to kind of command and control i mean there is even some of that in that founder mode essay by paul graham and like brian chesky is even going this way a little bit maybe not as much as people think the spotify guy is definitely going back in this direction that's interesting because spotify historically was very agile they had yeah they had distributed teams they all made their own decisions about their kind of part of the stack and they had this beautiful org structure that enabled the decision making to be pushed down to the team level everything's easy in a zero interest rate environment right like this cash is flowing we can give everybody all the empowerment they want and then things get scary and our human nature is to tighten up but i think again this distinction between conviction about the future vision versus conviction about the details is a good distinction because i think in lean economic times it's actually you should be doubling down on that vision and you should be even more flexible on the details because the details are going to change like it is impossible for any human or even any set of humans to get it right 100 the first time so like we have to have flexibility on the details but i also think like the stronger your vision the more likely you're going to find the right details So I do think we are making some mistakes right now. Like we are overreacting a little bit, which is fair. Like the world is in a state right now where we should all overreact a little bit. But I would love to see this stronger distinction between founder mode should be about the vision, the future vision of the product and the quality of standard, the craft. And I think we can still give our product teams the flexibility on that. How do you make a good? That's a hard question. You know, there's one word that comes to mind. It used to be curiosity. And I actually think curiosity is the root of the second word. But the term that people are throwing around a lot, which I think is really important, especially in an LLM world, is agency. And I think that like... agency is a word we use a lot i'm not sure everybody understands what it means but i think steve jobs really captured agency in his quote about i'm going to butcher the quote but something along the lines of like everything that you see in the world was created by a person and you could be that person right like love that agency is this idea of like we do impact our future and we do impact the world around us and the more that we internalize that and believe that the like more we're able to do and so i think for product managers the reason why this is so critical is it's so easy to think like well my boss isn't asking me for this they're asking me for this other thing so i'm just gonna do what i'm told well no like if you have conviction about this other thing and you think what you're being told is wrong you need to like communicate that and again not stubbornly right like I think one of the hardest things to balance is like we should have conviction and we should be really flexible. So how do we balance both of those things? I will share I didn't always get it right in my career. I definitely learned to like I should either be a founder or I should run my own business because I have a lot of conviction and I didn't always have a lot of flexibility when I disagreed with the founders that I worked with that also had a lot of conviction. And I think that's an important thing to recognize. But I think all of us, even those of us that like don't want to be a founder, we're going to be individual contributors. And for a lot of our career, this balance of conviction plus flexibility i think is really powerful so i think a good product manager has high agency they just find ways to get things done um they they don't the things i talk about is like when you want to look at change it starts with yourself so it's easy to complain about how your organization works but literally change starts with you so how do you look at what's in your span of control what you're contributing to the system that's at play and then i think related to agency is curiosity So I think it takes a lot of curiosity to start to envision what the world could look like rather than how it works today. You've seen a lot over your career, and I'd love to get your thoughts on how AI, large language models specifically, are changing product, if at all, because you're on the ground. I haven't been a PM since LLMS really came out a few years ago. What are you kind of seeing on the ground in product organizations? So I'm going to take this all the way back to the 90s. Like I started my career in, like I was in college in the mid 90s. I think I built my first internet product. I think we're about the same age. Yeah. Like I started college late 90s. Yep. Okay. I built my first internet product in the fall of 96. I actually built an Amazon site scraper that then led to me being one of their very first Amazon affiliates, which was super fun. Nice. Anyway, at the time, so like I was a freshman in college in 95, 96, and I remember we all, I was at Stanford, they had computer clusters, like not everybody had their own computer. We were the first year that like everybody checked email. And so we would bring down the mail servers, like right after lunch, there was like these points in the day where everybody would try to read their email at the same time. All of this is gonna sound crazy to folks younger than us, but there was a time when even email was new. We were all learning how to build web pages for the first time. And I remember we tracked when HTML 3.2 came out and then when HTML 4 came out. I remember when tables were new and then frames were new. And this was super exciting because every release we were very curious. We were like, well, what else can we do with this? And it felt like a toy. In those first few months, it felt like a toy. But a lot of us very quickly turned them into businesses. I paid for college by building websites for businesses. And I ran a web hosting company where I basically paid to host to get a dedicated server and then I resold it as virtual hosting. And so what felt like a toy quickly turned into a business, and the reason why I'm going into this in way too much detail, is LLMs feel like exactly the same thing. It feels like that exact same vibe. Everybody is playing with this as a toy. We're still trying to figure out what does this do? What can we build? Everybody gets excited about every single release. Like Anthropic puts something out and people go nuts. OpenAI had their dev day and everybody followed it like the old Mac developer days, right? Like there's all this hype and excitement around it, but a lot of it still feels like a toy. But some of us are starting to push into what can we really do with this? And I feel like that is so fun. Like I know so many people and I put myself in this category too. We're just reconnecting with what it means to be a builder because it's just fun again. And I think- Yeah, SaaS was feeling pretty stale in the late 2010s. Yeah, I feel like- Getting really developed and not a lot was changing. Tech got a little yucky. I mean, we've seen it a lot, right? Like all companies are following the same patterns, the political side of tech and like- especially on the national politics scene is just weird i don't care what your politics are what's going on is just weird all the layoffs and layoffs when companies are doing really well like that's just insane to me and so i feel like for those of us that have worked in tech for a long time like it's not what drew us to us to this right like what drew us to tech was this vision of the future of like what tech could enable and what tech could change i think we lost that like i think there was so much money in tech that it started to bring in the folks that were just in it for the money and not the idealists and the innovators i mean i'm starting to think of that cheesy apple ad but like it resonated with us yeah the rise of the tech bro basically yeah instead of going to wall street i went to silicon valley yeah a bunch of money yeah right i mean i get it like i made a good living and having money is very comfortable and it can be life-changing and everybody should have that like i'm not saying making a good living isn't a worthwhile thing but it shouldn't be your primary driver like and especially after you have enough it definitely shouldn't be your primary driver and i feel like those early days this is why i equate it back to the 90s like those early days of the internet were so fun and i think llms and generative ai is bringing that back i think it's reawakening a lot of makers it's also that there's still that much greed part like I go to LinkedIn and I cannot believe how much snake oil there is. Like there is so much like very surface level people writing about AI like they're experts. And I'm like, I always want to like repost. I'm not this much of a jerk, but I really want to repost and like just put a comment of like, how do you tell me that you've never used AI without telling me you've never used AI? It's like, this is not real. Like you don't just come on. It drives me nuts. I think what's coming with it is it is reawakening a lot of people for why they got into tech in the first place. And I don't want to be all Pollyanna-ish either because I actually have some huge concerns about the amount of power that's being consolidated in a number of small companies. I have some huge concerns about the crazy investment in data centers and the impact on the environment. We also had those concerns. I mean, not those exact concerns, but the internet didn't grow up without... its own share of concerns. But I just think it's super fun to be riding this next wave. And I feel really fortunate that like I get to experience this shift, big of a shift in technology more than once in my life. You know, people equate it to mobile. I call baloney. Like mobile was a shift, but it wasn't a huge shift. Yeah, it was like more like a new interface, but it wasn't like a big platform shift. Maybe cloud a little bit. I think cloud changed startups probably the most in tech besides the internet and AI, I'd say. but i think in terms of like impact on the average human ai feels more like the internet than mobile or even cloud cloud on business i agree with you entirely but if i'm thinking about like my neighbor across the street who's a optometrist i think ai is going to impact him as much as the internet did whereas i know he's got a phone doesn't care yeah so looking back at some of your formative moments are there any stories that kind of shape how you see good product management, good product discovery? You know, it's funny. I learned a lot by what not to do. I can't say that I ever worked at a company where from the top down, it was a good product culture. In fact, a lot of my career was advocating for good product culture. A lot of this is because I worked at early stage startups. And I think it's really hard for founders to create a good product culture. I also probably didn't work at some of the best companies, you know, and Some people... Like the intuits of the world were really famous for a good product culture, stuff like that. And actually, some people criticize me for that. They're like, how does she know anything about products? She works at all these terrible companies. Well, you know, first of all, I don't feel the need to defend myself. But second of all, I'll say that you can learn a lot by learning what not to do. Yeah, you and me both. I mean, most of my product experience is at early stage startups. And when I say early, I'm like, you know, it's like series A. You know, I think my first product role was at a Series A company. And then, wow, that's the Wild West of product. It is the Wild West. I mean, I was the... I was second product hire. The guy before me was hired a month or two before me. You know, they got a guy managing us who's never done product before. At my first company, I was the 50th employee. At my second, I was the 10th employee. At my third, I was the 23rd employee. And at my fourth, I can't quite remember, but we were around 20 people. I think people grossly underestimate how many startups fail. It's not that I picked crappy companies, it's I picked high-risk companies, right? And most high-risk companies fail. And also not all of those companies failed, but that's a different story. Thing is, I could have worked at Google. I actually got a job offer from Google at one point. I had plenty of friends that worked there in the late days. Like when they were less than 50 people, I played roller hockey there. One of my really best friends, like closest friends at the time was the sixth employee at LinkedIn. Like I had plenty of opportunity to go work at these places. The thing is, that's hindsight bias. Like I, at those times, I had very good reasons not to work at those places. Like I was a designer and Google was very adamant that search is just 10 links on a page. It was like the least interesting thing from a design standpoint. Feed is the design feature, right? Yeah. Now I was wrong because they got into all kinds of stuff, but at the time it was a good decision. So like, I can't fret about that. Here's the thing. And especially now that I've worked with a ton of companies, including some very successful companies, we see the same dysfunction literally everywhere. So like you mentioned Spotify, they're known really well for the Spotify model on agile teams and cross-functional squads. And there are pockets of Spotify that do that extremely well. And there are pockets of Spotify that really struggle with that. Just like at Google, they're known for OKRs and there are pockets of Google that are amazing OKRs and outcome focused. And there are pockets of Google that are very output driven, engineering driven, no understanding of the business context, right? Like we don't have good companies and bad companies. There's too much luck and uncertainty in the business world to be like, oh, you're amazing and you're terrible. we have successful companies and we have companies that struggle to succeed we have companies that survive and get by and then we of course have companies that fail and i think like what leads to a good product culture so many things have to be in place that you have to have a ceo and this is the thing that is most often missing you have to have a ceo who gets it and not just like oh i read marty kagan's book and i want to have a good product culture but has felt the pain of not having a good product culture and has learned we're a product company first because any digital company is a product company first and is willing to prioritize that and invest in that. And I think historically, our CEOs have come from sales and come from market. They're not coming from product. That thankfully has changed. But historically, that's not where they came from. It's not what they cared about. And so it was really hard to get these companies that would be good. product companies but we're seeing in the last 10 years like the stripes of the world product um figma we're talking about founders that have product mindsets and i think that started so it's starting to change but it does start at the very top like with the ceo do they care about a product culture and is that what they're prioritizing and then i think it the harder part like if that's in place what's really hard is all your middle managers have to also be bought in so we tend to think about it as like we got to train our teams that you need to learn some new skills yes that is true but if you train your teams and they learn some new skills and all your middle managers are still asking for 12-month roadmaps and telling them what to build nothing is going to change and so it's actually really hard to build a good product culture unless you're a product founder and you grow your company to be a good product culture and literally every hire is coming in with a product mindset which is why we see companies like stripe And it's funny that I mentioned Stripe because they didn't have product managers for a long, long time, but they had product minded employees across the board. Right. And so thankfully, like the internet has been around long enough that we are starting to see this. We're starting to see like product native companies, which is very exciting. But I think it's extremely hard for a non-product native company to make that jump. It kind of reminds me of the good product manager, bad product manager. Yeah. I think it was Ben Horowitz. Yeah. He wrote that, right? yeah so maybe we could we could on the fly brainstorm good product culture bad product culture i think at the root of all of this so why it's so hard is like for all humans all of us 100 of us like zero exceptions it's so easy in our heads to feel like we're right right like we're right i have this idea it's based on my experience and then we try to rationalize like our what we've already made up our mind about yeah and so what's hard and thankfully the internet makes it possible to see this we're actually wrong more often than we're right even at the best companies the best product-minded founders are wrong more often than they're right and the really good ones admit it they talk about like 80 of our test fail right like we are like full stop i think 80 of the products i ever built like hardly anybody used and generated no revenue like easily i remember spending months on on the oh this is going to be this is going to change the company and then you release it and like just dead yeah you know nobody used it So we have to acknowledge that as part of the process. We are fundamentally going to be wrong more often than we're right. So then I think the key to a good product culture is how do we learn we're wrong as early as possible so we don't over invest in the wrong idea. And so that's where the earliest widespread adoption of what some people call discovery, I don't call this discovery, was the adoption of large scale A-B tests. And I think I'm not poo-pooing A-B tests. I think they're an amazing- Step in the right direction. Yeah. Tool for measurement. Did what we build have the impact we expected it to have? They're a terrible tool for should we build this? Because you have to do all the work before you can even get an answer to that question. Whereas with discovery, I like to avoid building the wrong things. Like that's my goal with discovery is how do I learn as early as possible that my idea sucks? And if we assume most of our ideas are bad, now how do we do as little work as possible? to learn whether we should invest in this or not so that's sort of my framing of discovery and even that like we could just stop here because that's what's hard about product culture it's really hard for humans to adopt this mindset of i'm probably wrong and that we see that up and down the organization and all it takes is one person in that like organizational hierarchy to have the view of i'm right let's just build this and it all falls apart so i actually think it's a little miracle when we get companies with good product cultures because it's just really hard and you know it's it's there's this mythology around steve jobs right as like kind of the best product manager ever and he you know he had a vision and he was willing his vision into the world but really he was doing a lot of customer discovery constant focus groups and constantly interviewing customers and asking them and showing them like prototypes and iterating and i really like walter isaacson's biography of steve jobs isaacson is clearly obsessed with this idea of like the human that has an outsized impact on the world and i love his biographies because of it but he doesn't fall prey to the like single hero myth i mean he does a little bit like that's how he picks his people but he shows the whole person and their warts and all and you know steve jobs was not He was a terrible manager his first round at Apple. So bad he got fired. And he's criticized a lot for those early days of Apple would not have existed without Wozniak. And a lot of the hardcore engineering nerds are like, what did Jobs do? And that's not fair. Jobs was a very, very good salesperson from day one. But he clearly learned a lot in his next Pixar years. And when he came back to Apple, he was clearly a much better manager. But we also know he still wasn't perfect. And if we talk about founder mode... I'm sure there's plenty of details that he micromanaged. I mean, that's kind of what he's known for. Firing people on the elevator and things like that. Yeah, but he was exceptionally good at the two things we talked about. He was exceptionally good about being stubborn about his view for the future, and he was exceptionally good at holding a high standard of quality. Now, he may not have always been great at being flexible on the how. But it turns out no human is perfect. That's amazing. So walk us through the core habits and continuous discovery habits. What are the hardest to make practical for teams? Yeah. So the first is starting with an outcome. This is not a new idea. I mean, Stephen Covey wrote in the seven habits of highly effective people begin with the end in mind. That's all this is, right? What's the outcome you're looking for? What's the business impact or whatever it is. Yeah. It's this simple. What does success look like? What are we trying to do? And for a long time, product teams defined success as we built this thing. And so with outcomes, what we're trying to do is shift from just saying we built this thing, which is the output, to getting more into, okay, we have this thing. What impact did that have on the customer? What impact did it have on the business? So one, like the tagline in my book is how do you discover products that create business value and customer value? And I would argue it's not that hard to do one or the other. And neither is sustainable if you only do one of the two. What's hard is how do we do both? And so the first step is we have to get really clear on what does success look like. And this is, I always define the outcome as success for the business. And this really ruffles my UX folks' feathers because they think we should start with the customer. I have a reason for this. Customers have infinite needs, like infinite needs. We could spend the rest of our lives satisfying customer needs and we would never be done. And so when we start with the business need, it acts as a filter on the customer needs on that realm. So we're still customer focused. We're still making sure we're building for our customer and creating value for them. But we're using the business need as a filter so we don't waste our time building the wrong stuff. And actually, I'm going to pick on Google because Google is really great at creating a ton of customer value and doesn't always create business value, which is why Google has a graveyard of very popular products that were then killed. Google Reader. I couldn't believe I can cancel that. I use that every day. Google Reader is the one, right? It's devastating for those of us that used it regularly. I love that. I love that product. And not only did that product not, for people who aren't familiar, Google Reader was an RSS reader. If you don't even know what RSS is, just go learn some internet history. It was a feed of blogs, basically. I mean, I still use an RSS reader. It created a ton of customer value. It was free. It created almost no business value. And not only did it not create business value, It actually competed with the business value Google needed to create, which was Google doesn't want you to have a feed of all your blogs where you don't have to go to Google and search for them. Advertising in that product was always challenging. RSS is just technical enough. It was never going to cross over to hundreds of millions of people, right? Like I can see why they killed the product. But what I would argue is like, why from a product culture did they not start with okay, this is a cool idea and it's somebody's 20% hobby project in the company, but like, how did it get so big before they realized this is not a good product for our business? So that's, I think where the outcome comes in mind. And it's funny that I used a Google example because they're like, okay, our advocates right now. But then I think the next one is, I think about this as like, we have to learn about our customers so that we can build the right things for them. And actually, I forget which Stripe founder said this, but I absolutely loved it. He tweeted something like, when we're interviewing our customers, our goal is not to learn about our solution, it's to learn about their mental model so that our product matches their mental model. And I remember when I read that, I was like, that is a founder that 100% gets it. So many product teams, when they interview customers, they walk in with a prototype and say, what do you think? And I think we're missing the point of interviewing. When I interview you, my goal is to learn how you think, what you're trying to do, what your goals are, the environment in which you're trying to achieve those goals, how you think about achieving those goals so that when I build a product for you, it matches your mental model exactly. Because even a little bit of deviation means there's going to be friction and you're not going to use my product. So that's the second habit is just interviewing. And I like to see teams interview every week. Just continuously invest in your understanding of the customer, what their mental model is. making sure what you're building matches exactly how they think what are some key questions for the founders listening out there that are like okay i'm talking to customers am i doing it the right way what should i be asking them it kind of depends on the stage you're at so one of the things that i teach that's in the book is what's called story-based interviewing so it's this idea of collect a story about past behavior so if i work at netflix i don't like my intuition is to ask what do you like what do you like to watch how do you decide what to watch who do you watch with What device do you watch on? The challenge with these questions is humans, again, 100% of us, are really bad at answering direct questions out of context. And by bad answering questions, I don't mean you won't be able to answer the question. Your brain will give you a fast answer, but the answer your brain gives you won't necessarily reflect what you actually do. So you might say, well, I like action movies. I typically watch with my partner. I watch in my living room. What you're forgetting is like, yeah, but you also watched Ted Lasso and you watched a documentary and you watched a whatever. And sometimes you watch in bed on your phone, but you don't want to tell me that because you find it embarrassing, right? Like our brain. takes these fast shortcuts we know them as cognitive biases and they interfere with our ability to give a reliable response and the way that we're going to overcome that is we're going to keep the person grounded in actual behavior so i'm not going to ask you what you like about netflix that's a purely speculative question your brain is going to give you an unreliable answer i'm going to ask you to tell me about the last time you watched netflix and then i can customize this if i'm on the search team i can say tell me about the last time you had to find something to watch If I'm on the mobile team, I can say, tell me about the last time you watched streaming entertainment on the go. But I'm still asking for a specific story. And then my job as the interviewer is to keep the participant grounded in that story. And that's great if you have a product or you have a product realm or you have a theory of your product. What's hard for founders, like if you're pre-product, a lot of your research should be at the business model canvas level and like understanding customer segments. and understanding value propositions that can be a little bit different you might do a little bit broader more exploratory interviews but i'd still try to keep it really grounded in real behavior so i might do something like tell me about your full day tell me about the recent like the last week so it might be a much broader scope but it's still grounded in actual behavior so let's take this from the the vc angle right so i'm evaluating founders all the time uh put your you know your I don't know, your LP hat on now. Yeah. And how should VCs be assessing founders' ability to do customer discovery? I mean, if you want to have deals, I don't know that this should be high on your list. Most people are bad at discovery. Like I wish I could say it was different. I mean, my book has been out for four years now, over four years. People long before me have been writing about this. Like these are not. well they're not new ideas they're fairly old ideas but starting probably with like steve blank four steps even way before that right like a lot of this work is really grounded even in like human factors from like physical products like that dates back to like the 50s and the 60s so like this isn't these aren't new ideas again what's hard is that we feel right in our head and like if you're investing like especially if you're doing like seed investing and even series a investing like I would argue the most important thing is how much conviction does this person have? And are they going to run through walls to make this happen? Right? Like that's the most important thing. Like, I don't know that I would look for that flexibility we talked about in that early stage, which is hard because, I mean, you tell me, I would guess more startups fail because the founders don't have enough conviction. Yeah. That's probably the number one reason. Right. Either one or both, they're all co-founders decide to quit. in my ideal world like my little idealist view we would look for founders that had that conviction and had really good discovery habits but like i don't know that that human exists now if i'm a series b investor series c investor and we're talking about companies that are starting to scale and they're turning into real businesses i absolutely want to be looking at like okay you founder may not have this skill have you brought in a head of product that has this skill so i think it really varies by stage I mean, I know so many early, early stage investors talk about you're making a bet on the team. And I think that's true. Like the product, even what the company does might change 17 times. But I think the question is, is will they run through walls is way more important. Yeah. And why are they going to run through walls? Yeah. That's the interesting part is trying to figure out why. Let's talk about Product Talk Academy. What is that? How did it start? How has it evolved? And what are you kind of learning by running that community? Yeah. So when I went out. on my own. I started like almost everybody did. I just started as a consultant. And one of the things actually didn't make a conscious choice to go out on my own. I had been a startup CEO. I was not the founder. That's a weird situation. I don't always recommend that. I became the CEO in the 08 recession. We sold recruiting software. I don't recommend any of this path. It was really hard, like extraordinarily hard. And I ended up leaving that job right around the time where we closed an asset sale. So that company kind of got bought as an asset sale. And I left And I was really burnt out and me leaving was even political. Like I had some ethical disagreements with my board. So like I actually left about six weeks before the sale closed and it was just yucky. And I was not ready to get a new job, but I also was not financially independent. So I needed some income and I had a board member. I'm pretty sure this is how it played out. Like I had a board member that had my back because I resigned on a Friday and on Monday, I started getting phone calls about jobs and I was not a somebody at that point, right? Nobody had heard of me. I'd worked at four startups you've never heard of. Like it was just, I remember waking up on that Monday and getting those phone calls and being like, oh, this is how Silicon Valley works. Okay, somebody is helping me out here. I get it. And so I started taking phone calls and I actually just was so burnt out. I had this like pit in my stomach of I can't, I can't go do this again. And I, but I didn't know what the answer was. And so what I did was when companies started calling me, I asked them, I said, can I just do some research projects for you? And I actually started doing like user research as a consultant and i didn't love it because i really believe that product is continuous and i was like i'm just handing you some research and walking away and you're no better off as an organization and so i slowly started asking those companies like i will do this research for you but you need to send your product teams along with me so that i can teach them how to do this so that at the end of this engagement your teams can do it on their own and that's what led to my coaching practice and so for probably about 10 years No, not even that long. Probably first six or seven years. I coached teams and I started by coaching product managers. By the way, they were not product coaches at the time. I know now there's like a thousand of them. I think 7,000 of them. I'm positive. Like I searched for product coach on LinkedIn and there were zero when I started. And I wanted that framing. I was like really into this, like create your own category. I wanted that framing. because I didn't want to do the work for companies. I wanted to enable companies to do the work themselves. I mean, they were coaches. They were executive coaches. It's like coaching was a thing, but not at the product team level. And so I started by coaching product managers. Some of my earliest companies I just found through like personal connections. I got, frankly, I got pretty lucky. And then I realized like it's not enough because a lot of the challenges product managers were having was with their cross-functional peers. So then I moved into team coaching. I did extremely well as a product team coach. A lot of that is because I met Marty Kagan pretty early on. I get a ton of referral business from SVPG. I'm eternally grateful to Marty for that. But I also started to see, like I got to the point where like I had a six month waiting list, a year waiting list. And I was like, okay, this is great for me personally, but it's not the best way to scale my impact. And so I started to think about like, how do I make this more scalable? And because I'm a product person, I just started to think about it from a product mindset. And I started to experiment with online courses. It took a long time to get online courses right because it's very different to have someone on a phone call who can help you versus having a curriculum that you kind of have to go through. And there's a teacher, but you're not in the context of your own work. So through the Product Talk Academy, we have trained thousands of people. I think we're almost, I think we are over 18,000 students now. We've worked over people from over a hundred countries, over 3000 companies. Like it's just, it's mind boggling. And what's fun is, especially now with AI, like I'm starting to build AI teaching tools and what AI teaching tools unlock is I get really clear visibility in the student outcomes. So now like 14 years in, I can now scale the impact I'm having and actually really measure the impact. So I can see in week one, students came in and their interviewing skill was here. And by week five, it's moved. the needle so much amount and so today we have our flagship program which is called product discovery fundamentals it's our introductory course you get introduced to all the discovery habits historically we've offered what are called deep dive courses these are like they're about a single habit i just decided to sunset all of those courses and it's not because they weren't doing well it's because i think ai is going to completely disrupt the way that we train and so i kind of burned the ships to force myself to reinvent what the future is I have a portfolio company that helps professors using AI make, you know, coursework more engaging. Yeah. It's a company called Advisor. I'm really into using AI to create deliberate practice. So how do we, like deliberate practice is like, how do we practice the like sub elements of a skill with expert feedback? So we can create practice opportunities by interacting with, like interviewing, for example, we now have all the technology we need. for you to practice your interviewing skills by interviewing an LLM. We can even do it over voice. We can even do it over like what looks like a Zoom call with a real human. And then what's great is I already have an interview coach that can evaluate how well you did in that interview. So we're getting to the point where we can just provide infinite practice. Like this is something we could never do in a live talk course, right? Because we've got one or two instructors and a number of students. But like now you can practice as much as you want and always have expert feedback. And so I'm really excited about that. That's amazing. Yeah. So we're keeping our fundamentals course as is because a lot of big companies use that to like get people to level one. Our deep dives used to be like our 201 series, but I'm going to replace all of that with AI driven. Yeah. Well, you have all the content and now you just have to kind of structure the delivery using AI and voice and interactivity. And yeah, that's amazing. What are you excited about besides that in the coming years? You know, to be honest, it's enough. remarkable to me like i've shared a little bit of story publicly but like before march of this year so before march of 2025 like i was a consumer of ai just like all of us i've been using chat gpt i think since december of like the month after it came out busy i'm running a company of one and yeah i just didn't have time to really dive deep but in march i broke my ankle i had to have ankle surgery i literally was on the couch for three months And I decided to use that time to like dive deep on AI. And I'm doing it on two paths. I'm building AI products now to support my courses. But I'm also forcing myself to like change everything I do in my business. So from like a day-to-day operations standpoint. Cloud Code is just such a remarkable product, even if you're not an engineer. The way I'm starting to think about it in my head is I am basically building my personal operating system. And in the month of November, all of my blog posts on Product Talk are going to be Cloud Code recipes because I feel like this product unlocks so much productivity for me that I really want to start showing the world what is just now possible. And then it's funny that I use that phrase because I do have a new podcast where I interview teams building AI products. Great. And where this came out is I started building my first AI product and I was like, I don't know how to do this. Who am I going to learn from? and there wasn't a lot out there there's a lot of snake oil like there's lots of people writing about how to build with ai but if like you dig a little deeper they're not actually building with ai and so the goal of the podcast is i interview cross-functional product teams so the builders themselves about their ai products and it's super fun so i'm just that's really cool i'm all in i'm all in on the ai even if it's a bubble i'm just riding the wave well you know when when companies are adding billions of dollars of revenue like anthropic is right now in a matter of months I don't know if it's a bubble. It's not PetSmart, I can tell you that. Not PetSmart, Pets.com. Sorry, PetSmart. You're a real business. Webvan. It's not Webvan and Pets.com. There is money to be made here. I actually worry more about OpenAI, just the amount of money they've raised and the amount of money they're spending. Yeah. Anthropic feels like they're a little more. We could look back and they're like the CompuServe or the AOL of AI, right? So they have a big target on their backs. They got to stay ahead. I mean, the new software model is amazing. Right? Like the Yahoo. Yeah, they could end up being like the Yahoo. Yeah. They're killing it. But is it sustainable? Right. Yeah. And, you know, like if you're an investor in the late 90s and, you know, Larry and Sergey come along and you're like, well, isn't there like 16 other search engines? Yeah. You know? Yeah, there are. They're just all bad. And maybe there's another foundational model company none of us have heard of yet that's going to blow everybody out of the water. Actually, what I would be worried about if I worked at any of the foundational labs is open source models are starting to get really good. Yeah, the Chinese, the DeepSeq stuff, yeah. Yeah, they're just getting really good. And then if that's your business, and we're starting to see this, look at both. I mean, Gemini has always been this way, but Anthropic and OpenAI are pushing hard into the consumer and productivity apps. Like they're getting more into the like use case stuff. And I think it's because the writings on the wall, the model itself will start to become a commodity. Right. It's a go, it fades into the background, like the cloud. It's just kind of there. Right. And it's undifferentiated. And I got to actually build things that people that, you know, find valuable. Let's let's wrap up with some quick rapid fire. Sure. One question that comes to mind is, you know, the changing nature of product given, given AI, like, how are you like, and I asked this earlier, but I'll kind of say like, how do you like in your crystal ball looking forward in the next three years as ai becomes part and parcel to how we're doing everything how do you see it changing product and maybe blending the roles of product managers and designers and engineers and so forth i think part of the reason why i embrace such a like continuous mindset is because i've learned you can't predict the future i can share what i but i also believe in the william gibson quote which is the future is already here it's just unevenly distributed and so if i look at like the people right on the edge right now I think that's a great way to think about what's coming. I do think your title is going to matter a lot less. I think your ability to have agency and be a builder, no matter what your background is, is going to be really critical. I think there's this skill. This is a lot of what my blog posts are going to focus on in November. There's this skill of learning how to use LLMs well. And what's going to be required to use LLMs will evolve as the technology evolves. But there's some fundamental skills. And I don't just mean prompt engineering. I mean this idea of understanding context windows, understanding how to get the right context in at the right time, understanding how to break complex tasks down in ways that an LLM can excel at the individual pieces when they struggle at the big complex task. There's these fundamental kind of like... understanding the technology part and then learning how to use it really well, that I think will probably be evergreen. And I think it's really easy to think those sound like engineering skills, but I see lots of non-technical people learning those skills. Like some of the best marketers right now, people with no coding skills are like automating everything they do so they can do like 400 campaigns instead of one campaign. That is awesome. So I do think that like agency and curiosity become even more important because this technology when you learn to use it well when we talk about 10x engineers and i hate that term but i have a feeling it's a little bit too close to the like hero man the the like single man hero myth for me and don't get me wrong i have worked with engineers that have had outsized impact and i do think the best engineers are qualitatively different from most engineers so like i get what's behind that term I also hate that everybody thinks they're a 10X engineer and some people have the gall to like label themselves a 10X engineer. Does a 10X product manager exist? There are product managers that have had 10X impact, right? So like, I guess I struggle with applying it to the person. It's like you can have 10X impact, but I don't know that it makes you 10X better. If you have the right tools and the right training, you know, like for product growth, right? And I think what's really key, work in the right environment. Like I think we grossly underestimate. like matching the person to the right environment. So someone that struggles in one environment could really excel in another environment. And so I think that's the part we tend to leave out when we talk about 10X engineers. Like you have a 10X impact in this environment, but you may not in this other environment. Well, I could talk to you for another hour, but I want to be cognizant of your time. Where can folks find you online? Yeah. So I blog at producttalk.org every Wednesday, long form article, usually about discovery. These days I've been writing a lot about AI, but always with that discovery flavor. So how do we keep it customer centric? And then I have a new podcast. You can find that at just now possible.com. Awesome. Well, thanks so much for coming on. Thanks for having me.

Teresa Torres is a product discovery coach, bestselling author of Continuous Discovery Habits, and founder of Product Talk Academy. After years helping teams at companies like Spotify and Capital One rethink how they make product decisions, she’s become one of the clearest voices on modern product discovery.

With decades of hands-on experience across scrappy startups and global product orgs, Teresa brings sharp insights on customer interviews, continuous learning, decision-making, and the messy psychology behind why teams still ship the wrong things. In this episode, she breaks down why conviction makes founders great—and also gets them into trouble—how AI is reshaping product work, and why the best PMs operate with agency, curiosity, and a builder’s mindset.

In this episode, she discusses how she stumbled into human-centered design, why product cultures go sideways, what great discovery feels like in practice, and how AI is quietly rewriting the rules for PMs, designers, and engineers.

In Today’s Episode We Discuss:
00:01 Meet Teresa Torres
00:59 Origin Story
02:13 Early Startup Chaos
03:52 Continuous Customer Conversations
06:33 Scaling & Slowing Discovery
07:13 Founder Conviction vs. Flexibility
09:55 Scaling Product Beyond the Founder
11:21 Return to Command-and-Control
13:06 What Makes a Great PM
15:23 AI’s Impact on Product
18:19 Tech Losing Its Spark
21:39 Why Product Culture Fails
23:36 Lessons From Early Startups
26:47 Ingredients of Great Product Culture
29:39 Most Ideas Are Wrong
31:34 The Steve Jobs Misconception
32:10 Core Continuous Discovery Habits
34:18 Customer vs. Business Value
36:30 Better Customer Interviewing
39:20 Evaluating Founders on Discovery
41:30 Product Talk Academy Origin
45:01 Scaling Discovery With AI
48:45 Rebuilding With AI After Injury
51:00 The Future of PM Roles
52:20 Closing

Subscribe on Spotify: https://open.spotify.com/show/6Ga6v0YUsHotLhjap67uu5

Subscribe on Apple Podcasts: https://podcasts.apple.com/us/podcast/ignite-conversations-on-startups-venture-capital-tech/id1709248824

Follow Teresa Torres on LinkedIn: https://www.linkedin.com/in/teresatorres/

Follow Teresa Torres  on X: https://x.com/ttorres/

Follow Brian on X: https://x.com/brianrbell

Follow Brian on Linkedin: https://www.linkedin.com/in/bblinkedin/

Visit Our Website: https://www.teamignite.ventures

Subscribe to Our Newsletter: https://insights.teamignite.ventures/

👂🎧 Watch, listen, and follow on your favorite platform: https://tr.ee/S2ayrbx_fL     

🙏 Join the conversation on your favorite social network: https://linktr.ee/theignitepodcast