Stop Chasing OKRs: Solve the Real Business Puzzle

00:00:07 Mark Smith
Welcome to AI Unfiltered, the show that cuts through the hype and brings you the authentic side of artificial intelligence. I'm your host, Mark Smith, and in each episode, I sit down one-on-one with AI innovators and industry leaders from around the world. Together, we explore real-world AI applications, share practical insights, and discuss how businesses are implementing responsible, ethical, and trustworthy AI. Let's dive into the conversation and see how AI can transform your business today. Hello and welcome to the AI Unfiltered Show. Today's guest is from Lexington, Massachusetts in the US. She's the author of Radical Product Thinking, a visionary voice helping teams stop chasing goals and start building what truly matters. Radhika, welcome to the show.

00:00:59 Radhika Dutt
Thanks for having me here, Mark. I'm excited to be here.

00:01:01 Mark Smith
Good to have you on. Tell me about food, family, and fun as we start.

00:01:06 Radhika Dutt
Let's see. I'm definitely a foodie, so that is a passion, though cooking is not a passion. I like eating a lot more than cooking. Family, well, we're all a little bit from all over the place. It's always fun to travel somewhere and people ask us, where are you from? And it takes a moment to say, okay, here's the whole story. Also, I'm from India, I lived in South Africa and then went to university in the US. My husband's from Italy. My kids are half Italian, half Indian, but they were born and brought up in the US. So we're a little bit from all over is usually the answer. So that gives you food, family, and fun. Lots of travel involved.

00:01:45 Mark Smith
Nice. Tell me the best place you've traveled to.

00:01:47 Radhika Dutt
Oh, I'll tell you the place.

00:01:50 Mark Smith
Where would you go back tomorrow if you had the opportunity is probably my mind.

00:01:53 Radhika Dutt
Easy answer. Singapore. I'm always nostalgic for Singapore. Wow.

00:01:58 Mark Smith
That's so interesting. That's so interesting. I've been there lots of times.

00:02:02 Radhika Dutt
We lived there for 2 1/2 years.

00:02:05 Mark Smith
And you'd still just be back there in a heartbeat.

00:02:08 Radhika Dutt
In a heartbeat.

00:02:10 Mark Smith
Okay. We'll see, I would choose and my wife would choose Italy any day of the week.

00:02:16 Radhika Dutt
Really.

00:02:16 Mark Smith
Yes. I mean, I've gone to Italy every year for the last two years and we went a lot when we lived in the UK as well, but it's our number one favorite country in the world easily. We're almost hit 50 countries we've been to, so yeah. Love it.

00:02:34 Radhika Dutt
That's wonderful. See, Italy for me is a lot about visiting family and things like that. So it's very different in terms of vacation. Vacation for me is Singapore, I think, because, it just, I don't know, it's a place that makes me happy. I actually wrote about Singapore in my book, Radical Product Thinking. The country is built like a product.

00:02:56 Mark Smith
Yeah. Interesting that you say that, right? And it is. I don't know, I forget the person's name, that led the country into what it became. Yeah, but how fascinating. the first time I went there, I was reading, I was there because of Microsoft delivering training years and years ago. And then my career has always, work-wise, taken me back to Singapore quite regularly. So yeah, it is a beautiful country, and just, yeah, so much choice.

00:03:30 Radhika Dutt
More than anything, it really reminds me of the power of product thinking, which is kind of what I talk about in the book and how you systematically engineer the change you want to bring about by having this clarity of vision and translating it into a strategy, into priorities, into everyday action. And it is amazing to me that Singapore was built exactly with the sort of product thinking. Like if you listen to Lee Kuan Yew's first speech as a prime minister after Singapore split from Malaysia. Nobody knew that the country could survive as an independent country. And he talked about Singapore like a product. He said, you know, we want to produce an oasis for companies to come explore the Asian region. And that is exactly what they built in a very systematic way. It was so vision driven. And how many countries do you know which came out of colonization and then were that successful, no corruption? It's just a brilliant case study in product thinking.

00:04:34 Mark Smith
Yeah. I mean, I think they're a textbook example of a country operating without corruption and how strict he was on corruption and controlling that and preventing it and paying their ministers well. So they didn't need to be corrupted, compared to what we see in so many governments, including mine and New Zealand, the level of corruption's through the roof, you know, from a citizen's observation perspective. But before we, one thing I'm interested in, how did your book come about?

00:05:05 Radhika Dutt
So the way it happened was because of a lot of hard lessons. So, just my background is that I studied electrical engineering at MIT, and my first startup was right out of our dorm rooms. And it all sounds very glamorous, right? But it was a ton of mistakes that we made. And here's one example of a mistake. So our vision for that first startup was revolutionizing wireless. And you asked me 25 years later, like, what does that even mean? You know, I couldn't tell you. But it was all about being big. It's about scaling. We measured success by the logos of companies on our website, by the amount that we had fundraised. And you see the same sort of, what I call now, product diseases. They're just as ubiquitous, right? So this product disease that we had caught in that first startup is what I call hero syndrome, where you're so focused on being big, but you don't think about what's the problem you're setting out to solve. And since then, right, I've come across other such product diseases, either caught them or as I grew in my career, even though I knew how to avoid those product diseases, I was then watching others make those mistakes, catching those same product diseases. And so the burning question that led me to the book was, Are we all just doomed to learn from trial and error? Or is there a systematic way of creating change that anyone can adopt? Can we learn to create change through a framework so that we can avoid these product diseases and build world-changing products? And so that's what led me to radical product thinking.

00:06:45 Mark Smith
Nice. I just think it's so relevant to right now. You know, the work my wife and I are doing in the area of AI and I've probably been three years now in the AI space. And one of the things that companies want is the innovation that AI can bring to their organization. And one of the phrases that we use a lot is this humble and curious approach to things and life. And the last big corporate I worked for was IBM. And in that type of structure, and my wife's ex-Google, It is OKRs and follow the rules, tick the boxes, do your job is what gets you to move up through the organization. But we don't want you to think, we don't want you to come up with anything new. We don't want you to innovate. Just follow the prescribed pathway. And now we go into this area of AI innovation and we're saying, you can become more productive with AI. You can do all these, but I don't think it's around productivity and things like that. It's about how do you redefine business in a world where AI is central to that world and can influence so much, but none of the people in the organization have been trained or know how to innovate because they've been told not to for their entire careers. How do you address that?

00:08:11 Radhika Dutt
My gosh, so much to unpack in what you said, right? So let's start with the problem at hand. So what happens is, you mentioned people are told, just here's what you need to work on. Here, you focus on the tactics. We've already thought about the strategy. And OKRs is very much along the lines of this mindset. we already know the answer. Here's how we define progress. All you have to do is execute on this. It's based on the assumption that we already know what the problem is, what the solution is. You just execute and achieve these targets, right? And the critical thinking that is missing or that we don't leave space for is the puzzle. It's not just about achieving those targets. So let's take a very concrete example, and then I'll talk about how AI gets in the way. So here's an example. In the dating industry, one of the most important metrics is user engagement, right? And so it would make sense to set OKRs for user engagement, you know, achieve X percent increase in user engagement. So what happened in the dating industry? They did pursue user engagement as a key indicator of progress, right? In fact, when Tinder launched in 2013, they introduced swipe left, swipe right. And user engagement went through the roof. All the other dating apps looked at this and said, oh, now we know what to do. We can also copy this feature. So everyone copied swipe left, swipe right, or some version thereof, right? And so everyone's user engagement skyrocketed. Great. It all worked out in the short term, all of this optimization for metrics. There was someone in boardrooms showing wonderful numbers. But how did that work out in the long run? The more intimacy was gamified, the more it created a toxic dating environment, And then people started to feel this dating fatigue, started deleting apps. The entire dating industry fell into a slump. Bumble last year laid off 30% of all of its stuff. So the point of this is what the corporate world sometimes drives us to do is just focus on these metrics. Here are OKRs, you just have to pursue these. And What it drives is the incentive in employees that, yes, we're going to show you, ta-da, we've achieved those numbers. And so you achieve those numbers, and yet it doesn't mean that you're actually building for the long term. And this is what happened in the dating industry. Now let's talk about where AI comes in. All of what I just described about the dating industry was numbers that were optimized without even AI. Now you add AI into the mix. everything that we humans can do in terms of optimizing numbers, AI can do it even better, right? And so with AI, we're going to be able to optimize for whatever numbers you decide. And therefore, it calls into question this very mindset that we have, that we already know the answers. You just have to optimize for these numbers, right? And that is what we fundamentally need to shift away from. What the dating industry needed to ask was, what is the puzzle? And the real puzzle at hand was, how do we improve the dating experience while also monetizing our apps? That's the puzzle we should have really been thinking about in the dating industry. And missing that puzzle, focusing on numbers, which is increasingly going to be the case with AI, that is what we need to change. We need to think about the puzzle. That human element is still important.

00:11:55 Mark Smith
Yeah, that's so interesting. And the use of metrics and particularly vanity metrics that people, tune into, I've been podcasting for eight years now. And about a year and a half ago, I decided, hey, I'll bring in an external agency to promote my podcast and increase what I thought was views, listens, downloads, that type of thing. Within a week, I realized what had happened is that the agency was back-ended out of Bangladesh, and they were using bot farms, and my traffic just went through the roof stupidly, right? But there was no engagement. Of course, I quickly worked out, there's no people here. I've driven a number, and there's no people that are behind it. So what's the point of paying for that? Now, I understand, like through all social media platforms, people will inflate their numbers because that potentially can get sponsorship deals and things like that. But it is just short-term treadmill, of showing kind of fake success. And I think what you said about tying to revenue and I even feel that so much of the POC failure that we've seen in the last two years in the AI space is that nobody could monetize what they were talking about. And I'm not talking about the LLMs and stuff and Mark, I'm talking about a business doing a POC internally, loving the shiny new toy, but ultimately couldn't financially create something out of it because that wasn't their original metrics. They just wanted to play with the new AI toy. How do you help people in your research and stuff go from, one, probably identify, well, two things. How do you create an innovative environment, right, inside an organization? And how do you make sure you're measuring the right things as part of that innovation?

00:13:46 Radhika Dutt
And so this is what I'm working on in terms of my next book. And it's about why goals and OKRs don't work and what actually works instead. So basically how we transition from this mindset of goal setting to one of puzzle setting and puzzle solving. So let's talk about a concrete example, right? So there's a company that I've been working with and it's based in Greece. It's in the maritime industry. And what they do very simply is it's a data platform. to help people in the maritime industry figure out what is the best shipping vessels to match to which cargoes so that you can maximize profits. And so when I joined this company as a consultant in 2023, sales had stalled basically on this new product. And what I needed to do was to come in and figure out, you know, how do we think differently about this? How do we go from the stalled product where we just aren't really getting an expanded user base. And so how do we go from there to fixing this puzzle so that we get a much larger audience for this platform? Because the technology was brilliant. And so it had to start with defining the puzzle. And so in defining the puzzle, I talk about 3 O's. So the first O is the observation. So that's the problem that might be at hand. So here's the problem we were seeing. What we were seeing was, only tech-savvy brokers in the maritime industry were using our platform. So that's the first observation, that others who are tech-averse and in other roles, they're not using it, right? So then comes the second O, which is the open questions. So that's the, you know, we saw this observation. The open questions is where we ask stupid questions. We ask very basic questions, say, what don't I understand? Like, why is this happening? And so some of the basic questions we had to ask was, how do the people who are not brokers, so maybe the vessel, shipping vessel owners, how do cargo owners, how do they think? What are their mental models? How do they work? How is that different from the brokers? Without knowing that, we can't answer, why aren't they using it? And then we had to ask, like, how are these tech-savvy users different from the tech-averse users? Like, how do those people think differently? And we didn't know the answers to these questions. And we had to do some customer interviews. And it's so interesting that, staying with the discomfort is so important because our salespeople were saying, guys, you should already know the answers to these questions. Why are you asking these stupid questions and wasting people's time, right? But you kind of need to stay with the discomfort and really engage before you can understand the nature of this Rubik's Cube, right? You have to look at it to understand how it moves before you can attempt to solve it. And then the last O is the objective. So this is where you summarize the puzzle, which was, how do we solve this puzzle to grow beyond? just the tech savvy brokers and expand our user base so that we can get back to the growth trajectory. So I'll pause there because I talked about the puzzle setting, then we can talk about puzzle solving. But I want to address, you know, in terms of innovation, this is the starting point where instead of just setting numbers saying, you know, achieve X percent user engagement or increase sales by X million, right? You want to set the puzzle so that your team knows what are we working towards? What's the problem? Align on the problem. So I'll pause there before we get into puzzle solving. Thoughts? Anything you want to add?

00:17:19 Mark Smith
Yeah, so very interesting. When you said about, you know, staying in the discomfort, the first thing that came to mind was, you know, stay on the problem. Like, and I've been in so many workshops and sessions in my business career, where the customer starts to uncover their challenge. And it's like, I've got the answer, let's go. And it's like, well, hang on a second, have we really unpacked? And I like the Toyota 5 Ys and really, do we really understand what the problem is? so critically important before running off to solution land or what we think is a solution. that often isn't necessary, the right path. In anything you're doing there, in my mind, I'm thinking design thinking. Are you using any of these type of practices and, yeah.

00:18:07 Radhika Dutt
You know, so when we're solving the puzzle in observation, right, like to be able to interview customers and really understand the problem space, et cetera, it's all design thinking, right? And what's interesting is even design thinking says the same thing that I'm saying, we're just stay in the problem space. first expand on the problem, then you converge before you expand on the solutions, right? But here's the thing. I was talking to someone on a podcast and this host was a user experience expert, a user experience design expert. And she was saying, what you're saying isn't all that different from design thinking, but here's the problem. Every company that I've worked at, you never really get to stay in the design space.

00:18:50 Mark Smith
Yes.

00:18:51 Radhika Dutt
Or sorry, you never get to stay in the problem space. That's what I wanted to say. You never get to stay in the problem space because everyone wants to quickly jump to the solution. And even in this company, right, when we were talking about the puzzle and we first started wanting to observe, There was this temptation, like, wait, but we have some data already from our customers. We know what they're complaining about. They're complaining about this one feature where you have to filter all these lists. They're complaining that filtering is so hard. We know that that's what we need to fix. Let's go fix that. And by the way, even while you're fixing the design for filtering, don't rethink it completely. Try to make small changes so that we can deliver quick improvements. right?That's the temptation, because we have all these goals, targets. We're all hunting for those quick fixes that's going to address those goals and targets. We want to show that quick progress. And so it makes it really hard to stay in that problem space.

00:19:53 Mark Smith
Particularly, how do you encourage leaders to do that? You know, where they're motivated to get to outcomes quickly, and and they want to keep the, I suppose, the structure of their business intact, and they don't want the disruption so much that this potentially would create. How do you get them on board?

00:20:15 Radhika Dutt 
Yeah, a key question here is, what is it that we want to achieve? Because if sales have stalled and all you want are really tiny improvements, fine, keep doing what you're doing and look for those tiny iterations, but they're not going to really move the needle, right? If you're finding that there is something that you're not addressing in your business, like if you want to create a fundamental shift, you have to be able to look at the basics. You have to look at the puzzle as opposed to just keep tweaking one or two things in your Rubik's Cube.Like you just don't It's like, if you're trying to do the Rubik's Cube and solving just for one color without thinking about layer by layer, if you're solving every corner, right? It's kind of like that, you really have to be holistic in your thinking. And in terms of, you know, how do I convince leaders? Usually my message has been, look, if you're finding that things haven't really worked in the past. Like you haven't seen dramatic results with what you have been doing. Give this a chance. Let's try setting a puzzle. And you know what? It's not like we're going crazy and doing something, spending huge amounts of money and coming back to you with, you know, there's no huge investment in this for you. All we're doing is let's set the puzzle and we're going to solve this puzzle in small chunks. No huge investment needed. So what's to lose?

00:21:51 Mark Smith
Yeah, brilliant. So continue your story around the shipping industry and tell me then how you moved across. Just it's interesting to me. I work for a shipping company, Mitsui, in work I was doing in Hong Kong around their global logistics of moving their ships around the world. So it's interesting to me.

00:22:13 Radhika Dutt
Very cool. Okay, so how did we get to the puzzle solving, right? So when we observed kind of how these people in different roles were working and their workflows, we understood sort of what they need, how they think differently from brokers, right? And what we realized was our entire product and our mental models, because we had worked with brokers, we were initially mostly focusing on their workflows. So now we learned a lot more about these other roles. And so in terms of the puzzle solving, there were two things we thought we would try. One, new workflows that were targeted at these different roles. And 2, we wanted to address the tech-averse user, right? So what we thought was, you know, for the tech-averse user, instead of making you do a lot of these manual steps, what we'll do is we'll automate it so that we give youThe final answer, much faster with a click of a button.Here comes puzzle solving. These were our two attempts, right? And so puzzle solving means when you take an attempt, you ask three questions. The first question is, how well did it work? So notice, by the way, how well did it work invites the good and the bad. I'm not asking like a binary question, have you or haven't you achieved this goal? Good and the bad, tell me all. So how well did it work? So the new workflows, people loved the prototypes we built. The different roles were loving it. What didn't work? The people who were tech averse hated the automation we had created. So that's interesting. Wait, why did they hate this automation? So therefore comes the second question, which is, what have you learned? And this is where I say to teams, don't just spit out numbers at me. Don't just tell me user engagement, time spent on site, blah, blah. Great, look at all of those numbers, but tell me, what is the story? What's really happening? Tell me the narrative.And that triggers the right brain. So what was the narrative here? What we were seeing was the tech averse users, they hated magic and they did not trust technology. And when we automated stuff, they felt like it was magic. It felt like, wait a minute, how did you get those numbers? How can I trust it? How do I know where you pulled those numbers from? And because I was, as a tech averse user, I don't know these details. I'm not going to dig in like that's too much fear and worry. They did not like it. Now comes the third question, which is based on how well it worked and what you've learned, what will you try next? And now we had enough information to know what our next attempt at this puzzle was going to be. One,the workflows were a goal, we're going to build that. But 2, instead of automation, we were going to guide this user step by step through every step so that they feel like they're in control. None of this, you know, feeling of this is magic, how did you get here? And so we tried this. But this is more of a mindset, right? Like we started thinking in this way for every feature, every attempt at various parts ofthe whole platform, every part of the product, right? And the more we did this, we were able to holistically grow our user base. And so the results, we went from stalled sales in 2023 to doubling sales in 2024, again doubling in 2025, and we reduced customer churn from 26% to 4%. Wow.

00:25:47 Mark Smith
So obviously you found your product market fit?

00:25:54 Radhika Dutt
We did, right? And what was really exciting is what the CEO had to say about it, because he had been using OKRs.He had been the one who had introduced OKRs.And when he saw the team, the way we were using puzzle setting, puzzle solving, what he acknowledged was this approach was more like putting your ears on the tracks so that you could anticipate what's coming and then take corrective action. Whereas the way he described OKRs was, it was like looking in the rear view mirror. You either see that you hit targets or you miss them, but there's very little you can do about it.

00:26:27 Mark Smith
Yeah. Lead and lag indicators. I come up against them in businesses A lot. You can either find out what the, once a whistle blows, that data, you can't change it. It's too late. The game's finished, which is, you know, the end of the quarter results or end of the month results. How do you, obviously OKRs have worked, because GE, which I think was the originator of OKRs, and then Google, their whole business runs on an OKR model. Is it because they are big, well-established businesses that have well-established products? And I suppose they also had other levers. So for example, One of their levers ahead was 20% time, where they encouraged people to have 20% of their time to innovate, to create something new. This is why we have products like Gmail and Maps and things like that, because they were 20% projects, but their core business was still run on OKRs. Or do you think OKRs is something for businesses of the past and is a different, and we should be looking at it differently moving forward, particularly in the space we're moving into where the need for innovation and business just to stay competitive is going to be so much more prevalent in the coming years.

00:27:35 Radhika Dutt
Oh, I love this question. And this is what I think is really interesting. It is the biggest illusion that OKRs work. So first of all, let's talk about Google. If you talk to anonymous Googlers, you know, there was one Okay, I won't give any more details. There was someone, anonymous Googler, who said to me, you know, we joke internally that OKRs are a weapon against competition. We evangelized OKRs to make our competition fail. And, you know, this is anonymous Googlers saying this, but, you know, even within Google, it's not that OKRs are what drives success. It's more like, Google has succeeded despite OKRs. And let's even look at examples that were evangelized by John Doerr in his book, right, or his TED Talk. He talks about a Google OKR example. He talks about Google Chrome, that when it was launched, Sundar Pichai, who was heading up that area of Google's product, He had set an OKR. First year, it was something like get to 10 million users, and they didn't achieve that, but oh well. Second year, the OKR was something like get to 50 million, and they didn't get there. But he doubled down in third year, he said, get to 100 million users, and kaboom, they get to 111 million users. And so ta-da, you see, this is how well OKRs work. ta-da indeed, and kaboom, because in 2016, the European Commission slapped Google with its biggest fine to date for anti-competitive behavior. Google's Chrome achieved its OKRs not because they built the best web browser, but because of monopolistic behavior where they forced manufacturers to install Chrome on it. This is what I have tracked of other examples in John Doerr's book, Measure What Matters. Look at the Bill Gates example. There's an example in the book of the Gates Foundation, and the OKR is about malaria. Again, he touts it as this is such an amazing example and look at the progress this led to. Look at what happened in the long term. Actually, so do you want to hear the story?

00:30:17 Mark Smith
Yes, definitely.

00:30:18 Radhika Dutt
Okay, let's, so the OKR was eradicate malaria by 2040. And the key result was, I'll mention one out of three, prove that a radical cure-based approach works. What happened? Arata Kochi, Dr. Arata Kochi, who was the WHO's malaria chief at the time, he said that this had created a cartel of malaria scientists who were often pursuing cure-based solutions, even when sometimes the preventative solution was both more effective and actually safer, right? And it was stifling debate in the industry. The Lancet, leading medical journal, wrote a scathing article called What Has Gates Foundation Done for Global Health? And it talked about how it had created perverse incentives. Perverse incentives because it was taking away resources from other really important things to focus on malaria, because that was what Gates Foundation was driving politicians to do. So it's not true that OKRs work. So let's look at where do goals and targets actually work, because there's been research that has been done on this. What research shows is if you have a repetitive task at hand, which is if you have a task like, let's say, doing crunches in the gym, stuffing envelopes, where there's one right way to do things, then targets and OKRs work well. If on the other hand, what you have is more like a puzzle where there's no single obvious solution, the instruction do your best works better than setting a target.

00:32:03 Mark Smith
Yeah, that's so interesting. So interesting. Thank you so much for coming on the show. As we wrap up, we'll put in the show notes a link to your book. I take it's on Amazon.

00:32:13 Radhika Dutt
Yes, radical product thinking is on Amazon. But I'll also include on the show notes a link to the toolkit so that people can start using this puzzle setting and puzzle solving framework. And also I'll share my LinkedIn profile so that people can reach out to me because I'm working on the book at the moment. And so if people want to share their experiences, I would love to hear from all of our listeners.

00:32:39 Mark Smith
When's your next book out, do you think?

00:32:42 Radhika Dutt
Probably will come out sometime towards the end of 2027. It's still very much work in progress. I'm in the, so I'm still working on it.

00:32:51 Mark Smith
Awesome. Thank you so much for coming on the show.

00:32:53 Radhika Dutt
Thanks so much for having me. It's such a wonderful conversation.

00:32:57 Mark Smith
You've been listening to AI Unfiltered with me, Mark Smith. If you enjoyed this episode and want to share a little kindness, please leave a review. To learn more or connect with today's guest, check out the show notes. Thank you for tuning in. I'll see you next time, where we'll continue to uncover AI's true potential, one conversation at a time.