Martin Eriksson | How I Tested The Decision Stack
Martin Eriksson - Strategic Advisor, Author and Leadership Coach
In this episode, Martin talks about the difference between speed and velocity, why empowered teams still need strategic context, and how assumptions exist at every layer of an organization.
Have a Listen
Summary
In this episode, I’m joined by Martin Eriksson, founder of Mind the Product and author of The Decision Stack, to talk about how organizations can make better decisions from strategy all the way down to everyday execution.
We talk about the difference between speed and velocity, why empowered teams still need strategic context, and how assumptions exist at every layer of an organization, not just inside individual products.
We also explore what AI changes and, maybe even more importantly, what it doesn’t. Building has become drastically faster and cheaper, but that doesn’t eliminate the need to understand the customer, the business, and whether or not you should be building something in the first place.
If you want to learn more about how strategic choices connect to everyday decisions and how assumptions can be tested throughout your organization you’ll love this episode.
Takeaways
Speed without direction is not enough. Teams can build, ship, and test faster than ever, but without strategic alignment they may simply move in different directions more quickly. Martin frames this as the difference between speed, velocity, and organizational momentum.
Empowered teams still need strategic context. Autonomy works best when teams understand where the business is going, how it plans to get there, what matters now, and how their work connects to those choices.
Assumptions exist at every layer of the organization. They are not limited to products or experiments. Vision, strategy, objectives, opportunities, and execution all rest on assumptions that can become dangerous when they remain unspoken.
Strategy is a series of connected choices. The value of a decision stack is not just having vision, strategy, objectives, opportunities, and principles—it is making sure those decisions connect from the top down and can be traced back from everyday work.
Use testing and discovery on strategic decisions, not only product decisions. Teams have become good at de-risking what is directly in front of them, but the same tools can be applied to bigger bets around strategy, markets, and the future of the business.
Not every assumption can be eliminated, but it should be named. At the strategy and vision level, some choices will always remain bets on the future. Making those assumptions explicit allows leaders and boards to monitor when conditions change and decide whether the risk is still acceptable.
AI changes the economics of building, not the fundamentals of product. Faster and cheaper delivery reduces feasibility risk, but teams still need to understand customer value, business value, usability, distribution, pricing, and whether something should be built at all.
Guest Links
Martin’s Website: https://martineriksson.com/
Martin’s LinkedIn: https://www.linkedin.com/in/martineriksson/
Coherent & Wrong: https://www.thedecisionstack.com/coherent-and-wrong/
Transcript
David J Bland (00:02.862)
Welcome to the podcast, Martin.
Martin Eriksson (00:05.646)
Thanks so much, David. Happy to be here.
David J Bland (00:08.161)
I'm so glad we get to hang out. It's just more of an excuse to hang out and chat with you, and then our listeners get to hear what we chat about. But I'm a huge fan. For those of the don't know, I'm a huge fan of Martin's work. I think I I think I'm trying to think when I first met you. Probably in person was at Mind the Product in San Francisco, because I was a huge fan of Mind the Product, and I had been around product tanks because that is just a really natural spot.
Martin Eriksson (00:15.672)
Absolutely.
Martin Eriksson (00:19.436)
Likewise, likewise.
David J Bland (00:37.324)
You may not know this, but like a product tank was sort of my main channel for promoting the book when everything was locked down during the pandemic. And all the product tanks were like, we're really struggling how to do virtual events. And I was like, I will do it. I was raising my hand, you know, I'll do I'll do them all. And I just did this like world virtual tour. It was so awesome. But I think we first met in San Francisco when you were doing wow, was that twenty nineteen maybe? What was the product tank at the opera house?
Martin Eriksson (01:02.913)
Something like that. Yeah, I think I think we I think we probably met online as as w as one did back in the day when when Twitter was still Twitter and useful. I think that's how we probably met, but yeah, in person Mind the Products at the Davies Symphony Hall in San Francisco.
David J Bland (01:20.654)
That was an amazing venue. It was a vibe. it it was great. I mean, that was it kinda I mean, I speak in front of large crowds, but that setup kind of gave it an air of I don't know, like significance to to all the talks. It was it was really cool.
Martin Eriksson (01:25.249)
Yeah we
Martin Eriksson (01:37.398)
Something pretty amazing about being on those stages. I mean, we do used to do the but well, it still happens at the Barbican in London. It's al also an amazing stage. And I think the best part is when you're in the green room bac backstage and you see all the faces of all the like amazing artists and composers and musicians that have got up on that stage before you. It kind of puts it all into perspective a little bit when you're gonna go out there and talk about product. But it's it it's an amazing experience for sure.
David J Bland (02:03.916)
Yeah, it is. And I really appreciated it. And and a story that maybe people don't know publicly about us, so I was actually suffering from vertigo at the time. And I remember coming up to you and I was like, Hey, I don't really want to fall off the stage mid keynote. So can you put a stool like somewhere? And you're like, yeah, yeah, we'll put a stool like behind this podium for you. You could drag it out if you need it. You were just like so gracious and so, you know, just welcoming and just making sure you're taking care of your speakers.
And I I wasn't super steady when I walked out. And if you watch the video recording, I pull the the stool out and then I I I sit down on it and just kind of hang out with the audience. And all the feedback afterwards I got was, that was so cool how you sat down and just hung out with us. And I was like, Yeah, I was worried about falling off the stage at the time. Luckily I no longer suffer from that condition. But it was just really I just really appreciated that kind of like thoughtful touch to the way you approach work and
Martin Eriksson (02:49.066)
Yeah.
David J Bland (03:00.834)
the the way you take care of your peers.
Martin Eriksson (03:04.211)
Thank you. Yeah. I mean the for us the speakers always came first because we were trying to put on an amazing event and the speakers were the content and so we wanted to make it as an ex as an amazing experience for them as possible. And like little things like that, they're so easy to do. Like why wouldn't you support that and help out with that? Like I wouldn't want you falling off stage. That wouldn't be good for anybody, right? So of course we're gonna help out with that.
David J Bland (03:27.084)
Yeah, looking back, it's just it was just interesting how it landed. But well since then, so a lot a lot has happened since twenty nineteen or so. And I remember I think you had invited me at the time to talk about how to influence the business and how product influences business. And you've since come out with your your book, The Decision Stack, which I love. I'm wondering, like, what have you seen change what what have you seen with regards to
product influencing business or has that improved or not? Or what's what's your take on this? Because you're pretty plugged into the community.
Martin Eriksson (04:00.348)
Yeah, I think it's it's kind of frustrating how long a conversation we've been having about this, because as I look back at like my first writings about product and the kind of infamous Venn diagram blog post that came out in sort of twenty ten, twenty eleven, I say it's this is a business role, right? Like we are ultimately here to drive business impact. Like of course we want to do that by building an amazing customer experience and putting the user first and doing the discovery and and all the things that we talk about. But if we're not delivering impact, then none of those things matter.
And so we've been talking about it for a long time, but I think we finally had a reckoning as an industry in the kind of pandemic era, where I mean two things happen, right? One, I think during the pandemic a lot of people overhired and suddenly we have all these product people and you're starting to wonder like, okay, what's what's the ROI of all this investment that we just made? And then suddenly zero interest went out the window as well. So suddenly CapEx and Cash is not free and we have to pay for that. And again, where's the ROI on all this investment we're making?
And so that kind of came full circle and just required us to come back to impact and be like, yeah, wait, there's a thing called unit economics. Maybe we should focus on that. There's a thing called, you know, bottom line, maybe we should figure out how to deliver some of that. And so I think a lot of people have been leaving that conversation, right? Rich Miranov wrote an amazing book on money stories again, how to influence with that. Mattler May has written an amazing book on impact for his product teams again, how to kind of focus on that.
And then I think the evolution of that for me is really just then how do we do that in a way that's really about moving the business, not just moving, you know, a business number forward. How do we think about product skills, product tools, product capabilities on a on a company stage, right? And not just in tech startups, but in in any business that is touched by technology or software, right? I think there's so much opportunity for product people, designers, engineers to step up and be like, hey, I think there's a different way we can do this.
Not just look at like how do we build the best possible feature, but really think about how do we build something that has impact on our market and therefore on our business.
David J Bland (06:00.952)
Yeah, I feel like we're still having this conversation. I'm still reading my LinkedIn feed's a bit mess at the moment, but I it's as as we're recording this, I remember a few days ago reading something from I don't I don't know, I was it seemed like a pretty influential product person saying viability is just not part of what we're about. And it just I don't know, maybe it's how we define viability, because I define it as it's usually cost and revenue.
Martin Eriksson (06:21.028)
Yeah.
David J Bland (06:28.11)
You know, I'm looking at a higher level or it's it's like how do you move the needle on some metric that's business metric, right? A business outcome. And I'm wondering, even even though you you drew that well, let's see, like sixteen years ago at this point, fifteen years ago, it it feels like doing the drawing is harder than the drawing.
Martin Eriksson (06:48.654)
For sure. And and as with anything, I mean, the Venn diagram, the decision stack, any, you know, two dimensional simplification of an N-dimensional reality is always gonna get something slightly wrong, right? or oversimplify at least. I think it's I think it comes down to semantics sometimes, right? I think viability maybe just isn't the best label on this. I actually like talking about value a lot more just because then we can talk about customer value, we can talk about business value.
I think viability starts making people maybe think about like the PL and then it's like, well, I don't have responsibility with the PL. I don't know how much this kind of costs the sales and like all of this other stuff that goes into actually delivering and selling, marketing the product. And I think there's a whole other conversation. I mean, there are some business startups and scale-ups out there that go after that GM model more and actually put PL responsibility on their product teams. And I I do think that's kind of interesting. I think Shopify is one of the more famous ones for kind of really leaning on that GM model.
But obviously most businesses don't, right? And so product people, product teams, engineering teams are often at the end of a a long line of you know, goals where they can't necessarily move that big number, right? They can't move revenue, they can't move our margin necessarily on its own. So we have to break down like how do we how do we deliver value? And maybe that's where viability trips some people up, but I fundamentally believe it's at the heart of the job. Like the the main reason we exist is to drive that business value, right?
David J Bland (08:11.97)
Yeah, yeah. And I I think there's been some confusion over the years too, and I I get some flack for this sometimes when I write about it, but I'm not afraid to call out a lot of the communities I've supported over the years when we're doing stuff that doesn't seem to move the conversation forward. And in doing so, you know, there there are times when I feel like the the product owner product manager debate has kind of lost the it its thread in a bit. I feel as if now with the
Introdu introduction of AI. Now we have this, am I just managing or generating stories? Like how does that work? And and I I feel like it's it's this like tactical versus strategic level of responsibility. But I wanna try and draw that back to your work because I I was gracious enough to see you or lucky enough to see you speak in San Francisco recently at product tank about your book, The Decision Stack. And I really loved how you were kind of layering things
and showing it and visualizing it and trying to put words to something that I think a lot of companies struggle with. And I'm wondering, can you give us maybe just a high level of what your message is in that book? And then we then we can talk about maybe the testing angle on it. But for those of you who haven't read it yet, can you give them sort of high level of how you approached it or why why even you wrote it?
Martin Eriksson (09:30.729)
Yeah, so I think it came out of a couple of things. So one, I co-authored a book called Product Leadership in twenty seventeen with my friends Richard Banfield and Nate Walkingshaw. And that was really an excuse to go interview a hundred plus product leaders from around the world of like what do good teams look like? You know, how do I learn? Okay, great. I can go interview some people because I'm gonna put it in a book, they'll they'll they'll answer my call. and I think the the very short answer version of that book, right, is another
Done amazingly well, but it's really comes down to empowered teams, right? Because that was kind of the thread we were unpicking back then. And and you Mart Marty Kagan obviously started this, other people have talked about it, but this book really cemented that like the best teams out there are the ones that are empowered because they're closer to the customer. I think the thread that was missing in that conversation and and has been building since is that that only works if you also have the strategic context within which that team should be empowered, right? Because you can't have a hundred empowered teams running in a hundred different directions.
And I think we're seeing that even more so now with AI, right? With as things are moving faster, we're able to build faster, we're able to ship faster, we're just going in those hundred directions faster. And so I think over the last couple of years I've been advising a bunch of startups and scale-ups and it's kept seeing this thread of like, yeah, we're empowering our teams, we give them autonomy, we do this flat structure thing, but it's still kind of going wrong. And that's because that alignment piece just isn't there and that strategic clarity isn't there. And so I started to think about like what that actually should look like.
I gave this talk at a at a leadership event in Hamburg where I talked about the decision stack the first time. and it looked a little bit different, but it the structure was largely there of like how do you connect the dots from where we're trying to go as a business, how we think we're gonna get there, and then what it matters right now. And I think as practitioners, we tend to get stuck on that like layer of that's right in front of us, right? Like what matters right now, what am I gonna go build? Like what is the action we're going to take? And so we get stuck into all these great tools that we have around, you know, how do we
How do we think about opportunity solutions tree? Like Teresa really opened my mind up for that and I love that tool set. how do we use assumptions mapping, like all the tooling that you've been developing and working? Like, I love all these tools, but they tend to get us stuck at this kind of opportunity layer of like what's right in front of us, how do we solve this one specific customer problem? And then there's a gap then to the execs above us, and we're kinda waiting for them to tell us like who's the customer and where are we going and why does this matter? And if we don't get that, we kind of just run off on our own.
Martin Eriksson (11:53.51)
And if that con disconnect is there, that's when we start running off in lots of different directions. And so the stack really asks fundamentally five basic questions, which is where where are we going? How are we going to get there? What matters right now on that journey and how do we measure progress? What are the actions we're going to take to actually move forward? And then finally, how do we choose between those actions?
Now they sound deceptively simple, right? But as you go into organizations and you start asking these questions, and I do this all the time, it's kind of my my party trick of like just going in and talk talking to execs, you might not be entirely surprised to hear that you get very different answers from just about everybody in the building. And so the way you answer those, again, I'm I'm very ag agnostic too. Like I'm not a really dogmatic person about any tool or any anything, but I like to use vision obviously to answer like where are we going.
strategy for how we're going to get there, objectives and key results for what matters right now and how we're gonna measure it, opportunities for the actions we're gonna take, and then principles for how we choose between those actions. And the point is you can use different tools, right? Some people don't like OCARES, that's fine. But you have to have an answer to that question. And even more importantly, those answers have to connect. So the other connector that seems to unlock this for people is if you think about it from the top down, you're ask asking the question how?
So you have a vision, okay, how are we going to get there? Here's our strategy. Okay, how are we going to get there? Well, here's, you know, what matters right now, and so on. But even more importantly, from the bottom up, as those individual contributors, as those practitioners, we can look at a piece of work and just ask why. And then we can hopefully connect it to some goals. And we can ask why are those important? Well, hopefully we can connect to some strategy and so on. And what you find is most organizations have a decision stack, right? We have all these pieces somewhere.
We have a vision from a couple of years ago, some you know, it's up on a wall on a poster, maybe in the head office that nobody goes to anymore. We have a strategy from an off site last year in Lisbon that kind of, you know, sits in a PowerPoint deck somewhere in SharePoint. And then we have some objectives that somebody wrote down in a spreadsheet and we haven't looked at since in this quarter. And of course these things then start getting broken down and people start heading off in different directions, or or even worse, just relitigating a lot of these decisions over and over and over again. And so it's kind of become my mission to just help people connect those dots and
Martin Eriksson (14:08.974)
figure out how to make those more aligned decisions and just clarify that we have some of these decisions in the organization, right? So that our teams can do their best work and and pull together towards the same goal.
David J Bland (14:22.263)
I like that explanation. It's very clear. I love how you're putting a language to each and also calling out some of the challenges we have and how we work inside companies that seem like obvious way forward, but then we don't if we don't have that information at the time where we need it or that direction, we can we can become easily misaligned. Barry and Barry and I were talking about this a bit, and you and you and Barry are friends too. We we were talking about this idea of decisions and why it's becoming a
more important topic or more front and center I would say than I've seen it in the last maybe five to ten years. Why why do you think that is? Like Bar Barry gave his take on it, but why why do you think like what's a drawing you to to write and talk and help people about decision making at the moment?
Martin Eriksson (15:09.207)
I think there's for me it kind of comes back to the conversation about speed. And so I think physics or science has a really useful definition here, which is speed that is basically just a constant, right? It has no direction, has nothing. And and that is a what we often are chasing as individual teams. And if we are misaligned, everyone's going fast. Everything's going faster. We're pouring A on top of stuff. We can build stuff. We can ship stuff. We can test like everything's just going so fast.
But velocity is when you add a direction to the speed, right? So speed plus direction is velocity. So that's much more interesting to me, because then suddenly we're heading the same direction. We're starting to all, you know, run in that speed. And then even more interestingly is when you talk about momentum, which is velocity times mass, which is actually bringing your whole organization along with you. And I think that's where it gets really interesting, because then you can move at speed and bring your organization with you.
But I think that takes that clarity of decision making, right? So that's why I think the decision conversations come up so much more. and I think there's an interesting analogy, right, to to cause a lot of the work that you've been doing around assumption stuff, like, and again, this this conversation we we had just before we got on around like the practitioner level here, where we are so good at focusing all these tools, but we're focusing them on the the easy decisions almost, right? The ones that are right in front of us. and not that we shouldn't do that.
But we kind of forget that we have this amazing toolkit that we can also apply to strategy, we can apply it to vision, we can go test these things, we can go do discovery around these things as well. And the more that we could do that, the more we might actually understand that there's a completely different path that would actually be even better for not just us, but our customers and for the market. But that decisiveness is really, really hard if you're not using those tools. And so I think that's that's the that's the kind of dichotomy, I guess, right now of like everything's moving so fast that people are just requiring better decision making.
But then we have this like lack of alignment and lack of clarity on how to do it. We're probably using the tools that's a slightly the wrong layer. And so we're not as good at it as we should be. I don't I don't love everything about Jeff Bezos, but he has this great line, right, about one way doors and two way door decisions. Where two way doors are are the ones that you should you know, you can go through, you can test something. It's not a high cost. If you fail, you can go back the other way and undo it, right?
Martin Eriksson (17:24.915)
A one way door is where you go through it, you can't come back, right? So that's a really high cost decision. And I think what we find in most organizations is that most decisions are two way doors, right? You can't actually just go test these things. You can go, you know, whether you build it, whether you do discovery, whether you whether you do some research first, or actually just build it and test it. You can kind of do it as treat it as a two way door. And we're using all these tools against those two way doors. What we're not doing is doing it against the one way doors enough.
We're not thinking, especially now when you know, the rug's been pulled out from under so many organizations and we're re having to rethink what we do and what we mean to people, we're not really thinking about how can we use these tools to maybe answer that question in a better way, about our strategy, about our vision and about what we could do as a business.
David J Bland (18:08.721)
I I like that framing too. And yeah, I'm also conflict conflicted about there's some things I like about Amazon. Also also, you know, he had a quote I liked which was, you know, a test should only get expensive after it's been successful. I like I like that framing too, like trying to keep it small and everything. but I I wanna come back to I wanna I wanna talk about doors for a few moments here. 'Cause I I do like that framing. I get it, framing door. Never mind. Okay, so the idea here is two way, one way. I get the sense when I come into orgs
It's almost like they treat everything like a one way door. But it doesn't make any sense because then you would want to test the one way door. And so do you see that? I mean, I just I just I just I pick up on that time and time again where it's I it's obvious it looks like a reversible decision to me, but it's not obvious to the team that is in front of that decision at the time.
Martin Eriksson (18:45.565)
Yeah.
Martin Eriksson (18:57.491)
That's fair. I think especially maybe organizations that are less mature in their kind of experimentation and kind of assumption mapping and testing, right? They do treat every decision as a one way door and then but that doesn't mean that they go test it. That means that they just spend six months debating it before they, you know, just don't make a decision and just keep doing what they've been doing. And so I think that's a really fair call out, right? Like if if you're looking at every decision as a one way door, they're they're terrifying, right? It's like this is consequential. Like this is a big decision. This
could c you know, this could cost the business. but that doesn't mean we should shy away from them, right? And we we can do a better job of actually tackling them.
David J Bland (19:36.587)
Yeah, and I I do see that permeate also in the testing process as you hinted at there, because I do see the well when I do my assumptions map, I have to get it right. These stickies have to be in the exact right placement. And it's often followed by a conversation of we have to quantify the entire map because we need to put a you know, a number to each exact location. And I see that with experiment selection. You know, I want to run a test. Well, I have to pick the right test. And the kind of questions I get is like, how do I pick the right test? How do I know it's the right test?
What happens if I pick the wrong test? And that anxiety is just pulled through everything from okay, this seems like a one way decision when it's probably not, to I have to get it right on the map, I have to get the assumptions right, and then I have to get the experiment right. And then that experiment can't fail. It just feels like it just when when you start with that moment, and maybe it's the culture and the environment we're creating inside these companies, unfortunately.
But it feels like it just it permeates the entire process. It's not just one step of the process where people act this way. It feels like it blends into different steps.
Martin Eriksson (20:40.581)
Yeah, and I think that's safety, I g it comes back to I think you were talking about psychological safety a lot as well, right? I think it comes back to that point of like if if you don't have the psychological safety in the room when you're doing these things, then yeah, you're gonna default to the safe option, right? You're gonna look at the safe assumption, you're gonna look at the safe test, you're gonna go look at the safe information. And that might not be the right way to actually resolve that. I think I find often that teams also, you know, if you're defaulting to that safe thing, like I'm I'm
literally going through a a sprint at the moment with a client where we're using assumption t mapping and testing. Thank you for that. and we're working with them on it and they're looking at a lot of existing data, right? Which is amazing. Like they're pulling up gong calls or pulling up like all these amazing troves of information they already have in the toolkit. but then there's the risk that you're just looking at backward behavior, right? You're kind of looking at what are your customers doing today. You're not really looking at like well what would they maybe do in the future, right? What what what what would the world look like differently?
And maybe that's the challenge as well of moving away from this kind of opportunity layer, practitioner-led thing, where we're looking at something right in front of us, because it's come it's usually about something we're going to ship, you know, in the next couple of weeks or months at the most. We we don't have to be too forward looking, but when you think about strategy, you probably have to be a lot more forward looking. You have to be thinking and and making a bet on the future, right? You have to be making this hypothesis like I think the world's going in this way. I think there's a wave behind that.
And that information is probably not gonna be in your gone calls, right? That's gonna require a different set of tests. That's gonna require a different bit of research to go and figure that out or de risk that before you commit the business to it. but I think that's the other challenge to some of that safety that you're talking about.
David J Bland (22:19.114)
Yeah, and I and I think some of the conflicting market research, we w I was doing something with an AI company recently and we were talking about data centers. And, you know, there's some research to say, well, here data centers need this type of cooling, and there's another paper that says, well, no, they did need this other type of cooling, and they're at odds with each other. And it's like, well, how do you reconcile that? How do you look at the assumptions in that? And and it kind of led me to this next question, which is are there assumptions at different layers or how how are you taking
that into consideration in decision making at different layers beyond just what's right in front of us.
Martin Eriksson (22:53.314)
Yeah, I'm I'm working on a piece at the moment actually on essay, so maybe we'll time it to go at the same time as a podcast because I think it'll be very complimentary, which is this idea that I started to touch on in the book, I think, because I talk a lot about assumptions, again, crediting your amazing work and others before that. as the core of like how I think people should tackle strategy.
But actually what I've realized more and more is that assumptions exist at every single layer of your stack, right? So there's almost this like shadow structure or scaffolding that sits behind your decisions. And even if you have this like amazing, clearly aligned, you know, strategically you know thought through strategy and decision stack, if you don't name those assumptions, if you don't call them out, it's very easy to start drifting as well, right? Or to
you know, miss a change in the market that comes around quickly because you haven't named that assumption. I lived this very viscerally at a startup that I was an interim CPO at a few years ago called Kazoo. I can I can talk about them a bit openly at these point. I was a used car seller online and I mean fundamentally grew like crazy. I was a sixth person in the door within two years they had I think
five, six hundred people and they had SPAC'd for eight billion dollars in an IPO. And then within two years after that, they had crashed and basically gone bankrupt. So it's like real like boom and bust times, right? I was only there for the first six months, kind of helping set up the product org. But I think looking back at it, like one of the core assumptions, like almost at the vision layer, if not the strategy layer, was basically that we can keep raising capital at this cost going forward, right? Because it was definitely one of those strategies that was like it's you know
Go for broke on growth. We'll figure out the unit economics later. We kinda know that, you know, we've done some modeling. We kinda know we can get there, but we're not gonna focus on that right now. So, you know, there was an amazing founder who'd done a bunch of startups before that. So he had a you know, he had the track record, he could go raise money at a ridiculous rates, and he just kept raising money and he kept acquiring companies, he kept going for growth, you know, he kept sponsoring football teams to b build that growth, and it like the growth was there. Cars were selling, like everything was working until cash.
Martin Eriksson (25:04.051)
wasn't free anymore. And the whole thing kind of fell apart. And because they hadn't called it out, my hypothesis is that in hindsight, that took them too long to react to that change, right? Because they hadn't called out that like this is such a fundamental underpinning of the choices that we have made that if this condition changes, we have to come back and change that. And so I absolutely think assumptions live at every layer of the stack. And I think actually the higher you go, the more unspoken they are and the more untested they are.
especially as conditions change and and markets change around us.
David J Bland (25:36.749)
That makes a lot of sense. I I was talking to Mariana who who who founded Mural and and he and I worked together for a bit and I advised them and we he and I kept having these conversations about the board level, not just at their company, but just across, you know, different companies. We're not having these conversations at the board level or at the C suites suite level enough, you know. And when you think about why things fail in a big way, not a not a small way, if I trace it back, and again, maybe I'm
slightly biased because how I work, but it normally comes back to, well, there were assumptions there that that weren't we didn't talk about them. Like we didn't we didn't structure them. We didn't share them. and we didn't go test them. And they were there and they were unaddressed. And then we found out when it was almost too late, you know? And I feel like a lot of my work, if you t if you sh if you strip all the frameworks away and everything, if you look at the principles of everything, it's more about
Can we openly talk about the assumptions and spend time on that? Can we align on those and share them and make sure we have people in the room that can get aligned on them? And then can we just go test them? I mean, yeah, you could use my book, different book, whatever, whatever your own test. I I really don't have a strong opinion there. It's more of can we go find out quickly? And I I think I've just seen this play over and over and over again inside companies. And it it's somewhat tragic if the teams figure this out, you know, and the teams
like you said earlier, can apply it to what's right in front of them. But then when you ladder that up and you look at, no, we're we're at this like really micro level de-risking something, but in the big picture, you know, we haven't actually addressed some of the other giant, you know, looming assumptions that we didn't, you know, take in consideration.
Martin Eriksson (27:24.635)
Yeah. No, and I think I that's a conversation I'm I'm trying to help change, but kind of board by board, which which is tough. But I think boards are asking the question. They just don't know how to ask the question and therefore the answer we give them is wrong. So I think a lot of boards ask for this, like, okay, like where are we going? They're asking those kind of questions, they're asking like, you know, they're asking for strategy. And then we tend to come back as product leaders, especially, we're like, here's my roadmap for the next three quarters and like
That's not necessarily the answer to the question. And I do think, again, as practitioners, and and this is one of my biggest things with like the book and the work I do, is like, I think product people are awesome. Like I would not have started product anchor and mined a product if I didn't. I think product managers, designers, engineers, data analysts, like we have so much amazing talent and skill set. We just have to be empowered and be trust, you know, trust ourselves, I think, to use these things at different layers.
And so a lot of my work is just convincing product people, like, hey, you have this whole toolkit that you have been using, right? Like you're a successful VP or CPO. You got to this point by doing discovery work, by, you know, mapping out your assumptions, by testing them, by figuring out how to, you know, build a hypothesis-driven team. Why are you not having that conversation as soon as you step in the exec room, right? And you're just immediately going, like, well, here's my roadmap, here's the deck, let's move on, right? And so I think, if nothing else, I would love more product people to step up.
And have that conversation with the exec team and actually bring that to the board. And a lot of my coaching, I'm I'm te telling people and and working with people to do that, to kind of again just bring into the board room, like, I know you asked for my roadmap or I you asked for the strategy. Here's the choices we have in front of us, the assumptions that underpin them, here's the testing I've done so far. Do you guys have any information that I've missed? Right. Like again, just open the conversation. It's it's not necessarily even about coming with certainty, it's just coming with some of those choices. And then suddenly you can see a board come alive of like, hey, this is a different conversation.
And a lot of board people are, you know, their former founders, their former senior level operators, like they get this stuff if they're allowed to have that conversation. but I think it's only on on on us and and the CEOs out there to make sure that that is the conversation happening, right? Rather than just a rote kind of update of what's happened the last month and what's coming the next month.
David J Bland (29:35.86)
Yeah, I like that too. I think my personal, maybe like passion project recently has been this idea of challenging business artifacts about the assumptions and the risk that we're presenting to people. And we have all these almost un it's almost like this unspoken, of course we would use a PRD, or of course we'd use a roadmap, of course we use a one pager. And we have all these documents. And what I'm seeing
I don't want to feel it's like the matrix moment or something like that, where you like see see the code. But it's I'm like, do you understand this is sequenced risk here? Like if this doesn't happen in the first part, the second and third parts don't matter. And then, you know, the way you even described showing something to the board, like, or even the exec team, if you show an exec one pager of like, here's what the opportunity is, here's what we're gonna do, here's how much funding we need, and this is why we need to go forward. There is so much compressed risk on that, just one page document.
Martin Eriksson (30:09.486)
Yeah.
David J Bland (30:29.788)
And without I mean not that you have to go through every assumption you've tested, but I do think I don't know, I feel like it's like assumptions all the way down, it's like turtles all the way down here or something. It it's it's like can we at least acknowledge there are assumptions being made in the documents we're using to say this is what we should do? Because we treat those things as facts as if it's a certainty, and quite often it it's not.
Martin Eriksson (30:38.594)
Yeah.
Martin Eriksson (30:54.796)
And that's that's actually a fantastic way of phrasing it as well. I think the the risk framing is such a thing that boards are primed for, right? And that's also how your CFO is suddenly gonna be side, like, hey, wait, he's talking or they're talking about risk, right? They're talking about how do we manage risk and how do we trade those off and where are we comfortable taking risk? I think so often when we talk about assumptions as well, it's it's about testing them. But I think sometimes it's worth just calling them out, like especially at the strategy and vision layer, like just call it out. Like this is a massive assumption that we have.
We have validated it as much as we can. It is still an assumption. It's still a bet on the future. Are we okay taking that risk? And at least again, naming it calls it out so that when you can track it and you can see the conditions have changed, you can go, we're either in the right direction or hey, hang on a minute, we should change this. But I think even just naming it and talking about risks, like that's the kind of stuff boards are prying for, right? That's governance. That is all about the kind of leadership that they want to be involved in.
David J Bland (31:50.252)
Yeah, I think it's language too. Like we need to use language that people it doesn't make people defensive or it's something they are familiar with. I think governance, risk reduction, I'm constantly talking about de-risking things, or how do we make good investment decisions? You know, it's it's the language I think is is very, very key to this as far as speaking a language at the layer you're interacting with. And I also just feel as if
Martin Eriksson (32:16.225)
Hundred percent.
David J Bland (32:19.189)
You know, you take all of this and then you throw AI into the mix. And I you you touched a little bit on this in your talk in San Francisco. And I think all I mean, I I listened to the room. I mean, I'm not gonna out people, but like I listened to the kind of questions you were being asked. And I think the QA could have gone for hours, by the way. Like we had to cut it off at some point. But there was clear anxiety in that crowd of very seasoned and talented product people.
going, how do I reconcile my role with AI? And for those of of you all that weren't in the room at the time and was was able to hear the I mean, what is your message on this at the moment as far as, you know, AI coming into this equation that we're already kind of struggling to figure out?
Martin Eriksson (33:07.177)
I think it's it's probably two parts, right? I think one is as product people, we are so focused on the building side of AI, right? We're like, my God, we can build this stuff and like we should be product builders and I can just use lovable now. Like I should put all my PRDs in a you know, in a vibe code instead. And I think we're like knee-jerking into the wrong direction. I think as product people we we love the artifacts, right? Whether it's the document or the thing we build. And it's so easy to focus on that.
That we forget that fundamentally our job is about all those risks, right? Going back to the risk conversation, right? The the classic Marty Kagan, you know, feasible, viable, valuable, you can add Teresa Torres's risk about you know, how does she phrase it exactly?
David J Bland (33:50.677)
She has usable, there's there's all the bulls. There's adaptable, sustainable, there's a lot of them at the moment, yeah.
Martin Eriksson (33:53.796)
Yeah. Yeah. There's other it's it's the it's the kind of the ethics side of it that I think Teresa adds that I love, right? Like how do we think about this ethically? But you know, we forget all those other risks. Like AI has you know, when we use AI in our work, it has absolutely fundamentally changed what we can do and how quickly we can build stuff. But that's just the feasibility risk, right? We still have viability, we still have value, we still have usability, like we have all these other things that is supposed to be our job. And so I'm less worried about the product role.
in in the long term. I think what we're seeing right now, I I don't want to take away from the pain a lot of people are seeing right now. A lot of people are getting laid off. The market's messy. I I don't want to pretend otherwise. Like I've been there. I've been laid off during the dot com boom. I got laid off in the you know financial crisis. Like I've been there. It is not fun. So I don't want to minimize that pain. But I don't think that is actually because of AI. I think it's because of market corrections. I think it is because of again when CapEx goes up, suddenly people don't want to invest as much anymore.
Our books don't look we overhired, our numbers don't look good, the SAS pocalypse, again, a whole other conversation. But the real market impact of that is, you know, shares prices have gone down. Great way to prop those up is to actually have an AI story, which is like, hey, we're letting people off because of AI, not because we were suck at management. And so I think there's a lot of things that are driving that that ac actually don't have to do with AI. So I think long term, like or medium to long term as opposed to short term, I'm I'm still bullish on products and and design work.
So I think that's that's the kind of craft side of it. I then think the other problem is we tend to then get these like mandates. Again, boards, CEOs are like, what are we doing about AI? And then we suddenly go into roadmap mode and we're like, well, we have this feature coming up, or we're gonna build this, or we're gonna go test this. And it again, it just becomes all about that opportunity layer. And instead of kind of having a conversation about, hey, hang on a minute, again, the assumptions that underpin all this, right? It's assumptions all the way down, David.
if we look at our strategy, like, hey, we made some assumptions about like how expensive this is gonna be to do, or how, you know, we couldn't personalize this too much, or we couldn't do XYZ, AI's changed those assumptions, right? A lot of those assumptions have suddenly been changed, but we're not going back to that layer and and rethinking it. We're not going back to strategy and rethinking like, hey, maybe we could do something slightly different. We're not going back to vision and going, Hey, there's a whole new world out here where suddenly we can profitably deliver to this market that everyone's been ignoring because it wasn't pro be able to do profitably before.
Martin Eriksson (36:15.91)
Like that's the conversation I would be hoping product people would be having more. And again, the one I'm trying to empower people through through the book and through the talk.
David J Bland (36:25.001)
Yeah, AI is interesting dynamic here. I I feel I remember early days agile. also dating myself a little bit because I also went through the dot com boom. Although our startup survived somehow, I think because we pivoted and we just hired all our friends that their their dot comploded and we were like, come work with us. So we ended up with this amazing team, and I was there for almost eight years. but coming back to like even early days agile.
Martin Eriksson (36:30.98)
Hmm.
David J Bland (36:53.215)
It was all about delivery and about iterating through possible solutions and being very nimble and agile, right? And being able to course crutch really fast. And but I think the the expectation there was like, okay, yeah, we have this ability. Like if this is a car, we can steer it back and forth a bit better than we could before. but this this ability to deliver is going to make things better. And then I think if you go back and look at like the car being driven.
It's like, okay, now we're just yanking the wheel this way and then yanking the wheel that way. And I I think it was it was really interesting because I'm not I'm not anti-agile or anything like that. I I do think that whole movement was needed in order to make delivery trusted again. You know, like like we can actually deliver something really quickly and iterate on it. But I don't think
I I don't think it necessarily solved the issue of the driving part. And it maybe even exposed that a bit more because now it's like, yes, we could deliver anything and iterate through it. And then you realize, wait, where are we going? And that's what initially pulled me into all the work I'm doing now. Is because I was on the tail end of that, you know, these teams would be like really agile and we we'd like we used XP, we did T D D, we did all the things. And yet we just like really efficiently delivered stuff that nobody cared about and then threw it all away and hit a different direction and
Martin Eriksson (37:48.996)
Yeah. Yeah.
David J Bland (38:12.617)
I think AI reminds it's not the exactly the same, but AI reminds me a little bit about that. It's just here's another wildcard thrown in and yes, now we have this capability. and yet if you can't fix the steering, it it's still problematic.
Martin Eriksson (38:28.619)
Yeah. And I think, you know, I think both Melissa Perry and Teresa Torres have called this out early on of like Agile's great for the delivery side, but it kind of forgot about the discovery side. And if you go back to, you know, I I love h I love going through the history of the stuff and trying to figure out where all the stuff actually came from, right? As you know, and most of the readers pr our listeners probably know, right? It came out of the fundamentally lean and then fundamentally the the Ken of Toyota production system, right? And what we forget about is like
Yes, we use all these tools on the production line when we already know the car we're trying to build. The design studio does not work this way. They are not in a factory. They are not using these tools in ex that the way. And so I think they rightly call this out in Software Agile. And I think we've seen that massive correction. Obviously, amazing work from Teresa on the discovery side and just making that just such a core part of the conversation.
And I agree. I think that's what we're seeing in Agile as well, that just like over correction to like what's right in front of us, 'cause cause the world changed, right? And that's that's scary, right? Everything changed around us. It's happened really fast. We don't quite know where it's going 'cause it's changing so quickly. And so it's comfortable. Again, it goes back to that risk and safety thing, right? It's like just comfortable to like focus on the thing right in front of us rather than go back and you know completely reassess the thing that we spent two days in an off site six months ago, and have therefore written in stone since, right? Like
That is the thing we should go be challenging. I think also a lot of this just comes back to that risk conversation, because fundamentally lean and agile came out of that risk, right? Like the risky thing is building stuff because building's so freaking expensive, right? Software engineering was really expensive. And so we built all this tooling to minimize that waste. That's what lean was about. That's what then agile and scrum and everything was about. that's what discovery is about really fundamentally, right? It's like, how do we minimize that waste?
And suddenly when that waste conversation changes, I I wouldn't say it's gone away, because we can now just build a thousand prototypes and might be wrong in the same time. but it's still the same fundamental conversation. Like how do we de risk the thing that we're trying to de risk, right? and that's where coming back to those first principles to me is always more important than it's following any framework, really.
David J Bland (40:38.891)
So where do you think this is headed? I mean, you've been in product for quite some time and you've helped push the product community forward in in my opinion with your work and your thinking and your leadership. Wh where do you see this headed over the next five years or so? Where where do you I you can't predict the future, but where where do you think it might go?
Martin Eriksson (40:57.17)
Yeah. I think obviously all bets are off if we actually ever get AGI. I'm I'm less convinced that that is as imminent as many people think. Like that is a fundamental game changer if it happens. But in the world that we're in now, I do you know, even as the models are are changing and e becoming ever more capable, I think fundamentally it just comes back to that same truth, right? Like we have to figure out what do our do we want to build in the first place? What is the right thing to do? How do we drive value for our customers and therefore our business?
And if we can bring the conversation back to that as product leaders, there's gonna be a role for us. It's gonna be valuable, it's gonna be successful. And I think the more we can bring the conversation back to that over the next couple of years, the better. I also think like this is the the macro layer is that the fact that it is getting cheaper to build stuff and that, you know, we can have a different conversation about the quality of people building on cloud code and others, but you can build a lot cheaper than you than you could in the past. There's gonna be a lot more competition.
We're already seeing people who are like, well, I built this app over the weekend, and then I built another app over the weekend, and I'm not getting any users. It's like, yep, that's that's the product problem right there. Like, A, you probably haven't tested it properly enough. You haven't thought about distribution, you haven't thought about pricing, you haven't thought about how are you actually gonna market it. Like that's part of product as well, right? And I think that's actually a benefit to incumbents, right? Where they might already have distribution, they have a customer base, they have a brand, they have trust, they have all these tools in place.
That they can build on and and keep a market. Because if you're looking at a thousand ways to track your expenses, you're gonna be like, I'm gonna go with Expensive because I heard of them, I know them, I've been using them for two years, it's fine. Like it I the the switching cost is not quite worth it. On the flip side, I also think there's a ton of opportunity and spaces that we haven't really penetrated, right? I mean, again, I I can't decide how much I I want to quote this guy, but Andreeson's famous quote about software eating the world, right? I don't think it's done eating the world. There are so many corners of the world that
don't have software to help them out. I was on this amazing trip to Mongolia this year with my wife and we were sitting in a gare with a a literal nomad, like he has his family, he has a hundred horses, a hundred goats, like that's his life. He's just got a GPS tracker for his horses so he can figure out where they are without having to ride around looking for them. he was pulled up his phone to look at the BBC weather app to check what the weather was gonna be like tomorrow. This is not a person who is ever going to go vibecode anything.
Martin Eriksson (43:22.287)
But this is a person who would definitely buy some software that helps him track his horses, figure out what the weather's gonna be tomorrow, when do I need to think about bringing them in or you know taking them to a different pasture or whatever? And that's such a niche that five years ago no one would have built for that. It's not profitable, would never have been profitable. But now a couple of people in Mongolia could start that startup, they could build that software, they could build a great business, they could build a good life for themselves by providing that service as a software.
So I think there's two sides to this, right? There's gonna be a lot of noise in crowded spaces, but there's gonna be so many new opportunities that we can go build for as well.
David J Bland (43:59.049)
Yeah, I've seen that too with a lot of the early stage founders that I I work with some VCs in the Valley where we we take like waves of founders from different parts on grants, basically, different parts of the world and and kinda I teach a lot of this stuff. And it's so fascinating, you know, to hear the stories about it's almost like they jumped
Or skip to the industrial era and now they're and and the and the issues and the and the technology and how they're using the technology and everything. It's super fascinating. And also conflicted about Dreeson a bit. I mean, I've definitely he's influenced a lot of my thinking. He also blocked me on Twitter. I don't know. I must have made him mad at some point. So I didn't survive the Twitter apocalypse thing that he did when he blocked everybody that he liked or had wronged him in some way. But I do think I like how you're thinking and you're optimistic.
about the next five years. It's not a no, everything's unraveling. It's more of a No, I actually think this stuff still matters and here's why. I I like that point of view.
Martin Eriksson (44:59.513)
Thank you. Yeah. I think it's I mean, every technology innovation in history has followed that pattern, right? Of like short d short term doom and gloom. everyone kind of doom scrolling or whatever the eighteen hundred version of doom scrolling was about like, my god, the printing press is gonna end the world. or you know, cars are gonna end the world 'cause horses and you know, all these impacts have all had massive changes. Eight you know
This happened when tell to tellers when ATMs came along. This happened to radiologists when AI first came along 10, 15 years ago. This is the story I tell in the talk. but what happens fun ultimately is after that short-term pain is actually long-term abundance, because we tend to do more, right? and this is the Jevons paradox, which I think a lot of people are talking about these days with AI in mind, that like when something becomes cheaper and faster, we don't just do the same amount of it, we do vastly more of it. And so there's more demand, there's more need for it.
And I think that's true with software as well. I think there's gonna be a boom. I think there's an opportunity to build a a new renaissance for software. I think there's so much more exciting th stuff we could do. I'm also frustrated still. I love AI, I use it every day, but I'm so frustrated I'm still using it in a freaking terminal window. I feel like I'm back in MS DOS. Like this is not what the future was supposed to look like. And so like new UXs, right? We we have people talk about this and think about this, but there's so much exciting stuff that's still to come.
And so yeah, I I I get optimistic about this. I think there's a there's a lot of amazing stuff we can build. I do think that doesn't mean we shouldn't shy away from the risk, right? Again, going back to risk and how do we manage risk. It's all about risk and assumptions, David. I think we do need to be mindful of that, right? And like wherever you live in whatever country you're a citizen of, like, think about how you're voting for this, because we are gonna need to be a lot more intentional about how we manage this stuff. And if AGI ever comes along, obviously these bets are kind of all off.
And those companies will very happily gobble up every job out there. And that is a very different, much darker future, and one that can probably only be managed by policy. But I do think that's further off, if ever, basically. And I think in the meantime there's more opportunity than there is risk. I think we just have to look at it differently.
David J Bland (47:14.992)
I I love that. I love that. It certainly does make some mistakes. Daniels and I were going back and forth laughing the other day because
He asked it if I had spoken at the specific conference. It's like, yeah, David spoke at that conference in June. And I messaged him back and I was like, dude, I've never spoken at this conference. And then it ended up being a inside joke about, well, maybe it was like a multiverse, David, and it was like lizard David speaking at a different conference. So i it is important to check your work as well and and go through it. But I do think it comes back to to risk. And I just appreciate your
Martin Eriksson (47:35.509)
Yeah.
For sure.
David J Bland (47:45.886)
take on things. I've always it's it's just really refreshing, you know, you're you're super smart and you're also optimistic and I think that's why I'm drawn to you. And I love hearing your stories. I love hearing you share your wisdom with our listeners. If people want to reach out to you and they don't know how, what would be the best way for them to get in touch?
Martin Eriksson (48:08.382)
Well obviously the book and all my current writing and I'm writing new stuff and blogs and essays all the time is at thedecisionstack dot com. I'm personally at martin eriksson dot com and you can reach out to me at Martin at martin eriksson dot come Very long but very simple email. and I'd love to hear what people are working on and what's not working for them out there.
David J Bland (48:26.63)
Amazing. So we'll put all those links in the description as well as in the detail page. Thanks so much, Martin, for hanging out with me. It's just always great to to hear how you're thinking about things and how you're drawing it back to risk. I am in total agreement with that. So thanks a lot.
Martin Eriksson (48:42.74)
Thanks so much for having me. I'm I'm a massive fan of your work as well. I gotta get that in there 'cause otherwise I'm gonna start blushing at all the nice words you're saying about me. So love hanging out with you as always and look forward to doing it in person soon.
David J Bland (48:54.602)
Thanks, Martin.
