#4 Radhika Dutt - Radically Rethinking OKRs

Sebastian: Welcome back to the Product Masterclass podcast. In today's episode, Radhika Dut joins us to explore the limitations of metrics and OKRs, the power of narrative in goal setting, and a fresh approach, OHLs objectives, hypotheses and learnings. And we'll also discuss why experimentation matters more than rigid targets, and how real change starts from within the teams. So let's dive in. Hi, everybody, and thank you so much for joining us again. Today is actually super, super special to me because we have Radhika with us. Thank you so much for joining us, Radhika.

Radhika Dutt: Thank you for organizing this, Sebastian.

Sebastian: I actually. So before we get started, I actually have to tell, um, a little, uh, private anecdote how Radhika and, uh, me met. So, uh, that was last year in October, I believe. And, uh, we were both speaking at this conference in Romania called Pro. And, um, I think we already met, like, during the rehearsal of the speakers, which was a lot of fun. So there were basically, like 10 people in the room. It was already a lot of fun. And then there was this cocktail reception, um, that we had in the evenings. And we're just like, standing at this table and pretty, um, much talking about everything product. And it was so much fun. That's why I'm super delighted that, um, Radhika is here today and can join us.

Radhika Dutt: Likewise. That was one of the best conferences. And it was really because of all the people that we met there, including you and, uh, other people, that, um, those speakers there felt like authentic connections. Right. It was just really wonderful.

Sebastian: Absolutely. And, um, if by any chance you don't know Radhika yet, which is impossible, to be honest, but, like, in case you haven't met her yet, um, so Radhika is the author of a book called Radical Product Thinking. And, um, first of all, brilliant book. If you, if you haven't. If you haven't ordered it yet, like, go to Amazon right now and just, like, order the book and then read it and then you can thank me later. And Radical, of course. Um, but the reason why we're actually here today is, uh, to talk about a new book you are currently writing. Um, which is. And I'm not, like, spoiling it because you can explain it way better than I can, which is in, like, the rough outline is basically rethinking metrics and how we use them.

Radhika Dutt: Right, Exactly. So I'll share a little bit about why this book even came about. Um, so, you know, one experience that I've had for years and years now is I find that Companies are goal oriented. They set up okrs. It's all about setting targets and saying, okay, as a product team, you know, you have to hit. Hit XYZ targets for these different metrics, or as a company, we have to hit these different metrics. And what I found is, right, whenever you have targets, metrics tend to get corrupted. Uh, and there are like three or four common phenomena that I keep seeing. So the first one is sandbagging. I was observing okrs being, uh, set up in an organization at the beginning of the year. And, you know, all these leaders were negotiating OKRs. And finally, like, one leader gets up and goes like, okay, guys, I cannot sign up my team for failure by committing to these okrs, right? And no matter how ambitious you are, it feels like you're committing to. It's like you're setting a final exam for yourself and you don't want to fail that final exam, right? So even for the most ambitious people, you feel like when you have a target, it's a failure if you don't hit that target. And I know people say, oh, you're not supposed to achieve your okrs. They're supposed to be, um, aspirational, etc, but it's not grounded in science at all. This is not how it actually works. This is not how people actually feel like, you know, people are not okay with getting, like, especially ambitious people. They're not okay with getting anything less than an A grade. Right. Or A plus. Right. Another common thing that I see is.

Sebastian: Can I ask something before we jump into the next thing? So isn't, I mean, so saying, uh, okay, they should be aspirational, Isn't it? Isn't this also like a cultural thing in terms of company culture? Or would you. Would you say otherwise?

Radhika Dutt: The. It's the book, um, Measure what Matters, written by John Doerr, that about okrs that actually says this, that, oh, they're supposed to be aspirational. You're not supposed to ever hit that okr, right? Like, if you get to 80%, that's supposed to be great. It's nice to say all that, that you're not expected to hit it. The reality is that's not how people actually feel. That's not how psychology works. Uh, so I don't know that it's about company culture. I think leaders then read such books and such statements by famous billionaires and then feel like this is what they're supposed to do, but it's not grounded in science or psychology.

Sebastian: So is that. Once again, this. All right, it worked. This One time and then we are kind of adapting it. So is it the Spotify Squad model again?

Radhika Dutt: Exactly. It is like the Spotify model that even Spotify doesn't use. Right. Um, by the way, in fact, Spotify, uh, used to do individual okrs and they dropped that.

Sebastian: Right.

Radhika Dutt: Um, it's just an interesting, uh, tidbit. But going back to this, right, the whole premise that OKRs brought success to Google and the Bill Gates foundation and blah, blah, and all of them attributing success to okaris, it's kind of like saying, you know, Steve Jobs was successful because he wore a black turtleneck. It's like he was successful and he could have worn whatever the hell he chose. Right. It's not that black turtleneck made him successful. So it's kind of like that about okrs.

Sebastian: Got it. So what you're essentially saying is like, all right, these, these organizations would have been successful because of their principles, because of their culture, because of the people that work there had nothing to do with OKRs. I mean, this is again like the kind of the discussion about like, should you adopt these frameworks? Is it more important to use the frameworks or to apply to these principles?

Radhika Dutt: Right, right, exactly. I think what happens is you have people with a lot of authority and fame who have been successful. It's like, you know, John Doerr is, is kind of like the Midas of uh, of Silicon Valley when he says, this is what you should do. You know, it carries authority. And so we listen to that, but we haven't necessarily looked at it scientifically and said, okay, you know, do okrs actually work? And throughout my career I've seen issues. And what's interesting is that there have also been research papers. Like there is a really famous paper called Goals Gone Wild, and it talks about how goal setting actually is detrimental to companies. And despite, uh, uh, by the way, what's funny is even John Doerr in the book Measure what Matters, he talks about this really well known paper that talks about the issues with goal setting and he dismisses it, saying, if you just set okrs right, then you overcome these issues. Right? And um, this is the thing, like every time anyone says, well, okrs aren't working for me, people go like, oh, then you're just doing them wrong. Right? And this is why I want. I felt like there was a need for booked, right, to really talk about what are the issues with okrs. Um, and so I talked about sandbagging is the first one. Shall I go on to the next one?

Sebastian: Yeah, absolutely. Maybe Maybe one. One lot. Yeah, because I'm, I'm actually super interested because, I mean, we also have people coming to us and saying, like, okrs don't work. There's companies that I've seen. For them, it worked, like, fairly. Okay. Um, so the, the, the question, I mean, obviously it's not only about okrs, right? So okrs is just like the framework that everybody wants to adopt right now, right? It's, it's more about, um, how should you set goals instead. Right? So this is. Or like, how should we set goals? So this as like a. More like, like a broader question.

Radhika Dutt: To me, it's more about how do we use metrics? Because when you set goals or targets, right, let's talk about real examples. Let's say you set a target of get 20,000 signups, uh, by the end of the year. So what happens when you set such a goal? Teams are all working to prove that. Look, I hit that number that you set for me because nobody wants to fail, right? So I'm going to work as hard as I can to show you I've hit 20,000 signups. What do you want instead? As a leader, you really want to figure out the fundamentals of your business. Like you, like if you're a startup and you're trying to get all these signups, you need to figure out, is my message resonating with Personas that I'm targeting? Am I targeting the right Personas and like, what is working? What is not working in terms of marketing? Right. I need my team to experiment and solve this puzzle. So what research papers show is when you have a puzzle and there is no single obvious strategy that works, goal setting actually gets in the way of progress that you're less successful in solving those puzzles compared to a, uh, do Your best instruction vs 20,000 signups by end of the year. And so if I go back to this model of goals versus metrics, right? The whole point is don't set targets. But metrics are super important. You should be measuring, like, you should be figuring out, you know, what is, what is working, what is not, which Personas am I targeting? What is. How do I figure out if my message is resonating for those Personas or not? What are the right sales channels and like, what marketing message do I try on these different sales channels? How do I optimize for my whole conversion funnel starting at the top, right? There's a whole bunch of stuff to be measured. Um, but what you don't want, which is when you have targets, is you don't want a corruption of metrics m where people are just trying to show you tada. Uh, look, I achieved what you told me to. And if it's just 20,000 signups, you know, I can game the system. I'm smart. I'll figure, like, I'll just include all the bot signups, I'll include whatever crap I need to, right? Like whether or not it's the right Personas. But that's not what you want from your team.

Sebastian: Okay, so understood. So you should have metrics in place, you should measure so that let's. You made this one clear, right? So it's not about not measuring, it's just that sometimes goal setting in the situations that you described, um, in the example, get in the way. All right then. Full sandbagging, guys. So sorry, I'm just like. So the reason, I mean, the thing is the reason why we do these events because they're, uh, interesting for us, right? So hopefully also to the people who are watching and listening. But, um, um, yeah, that's, that's why I'm asking all these questions because it's like, super interesting for me. And um, obviously I also don't have, um, Like, I don't really have a set opinion in this case. Um, I just find it like, super interesting to, to get a different perspective on goal setting and metrics.

Radhika Dutt: And it is such a different perspective. Right. So I really appreciate all of your insightful questions because it really is a different mindset. And here's a really interesting thing that I discovered just literally in the last week, uh, so hot off the press, like, here's why we think differently when we don't use targets. I was talking to a colleague, an ex colleague of mine who I had coffee with and I had seen him in like more than 10 years, right? And so we started talking and he, we connected because he read one of my posts on LinkedIn and about, um, instead of using okrs, using uh, ohls or objectives, hypotheses and learnings. And it really resonated. So he said, you know, let's get together. So we had coffee and he was telling me how in 2020 he had a stroke and it affected the left side of his brain. And one thing that he realized is that it shifted how he thought about product or how he was looking at metrics, that initially he was really focused on the analytics and the numbers and just, you know, like, and this is what happens when you set targets. It's always about, am I hitting this 20,000 signups or not? This is kind of the one metric. It makes you very focused on that one task of 20,000 signups. Right. And what he said was after the stroke it, he started using more of different parts of his brain where it was more about the narrative and the strategy. And so instead of uh, focusing on just numbers in the metric, it was more about the story of what are these metrics telling you? And this hit me because, you know, when I work with product teams sometimes, like I literally had this last week where I was talking to a product manager and she was telling me about weekly active users. And I said, okay, forget the number of weekly active users for a moment, tell me what are these people doing on the app? And walk me through exactly what you're learning from all the analytics and the metrics. Tell me the story. And then as she started telling me the story, there was a whole picture that came together. A story, right. And from that story I was able to say, okay, now that you've learned this, if I gave you a magic wand, what would you ask for in terms of a new landing page for mobile and like it, it led to her answering the questions, is it working? Then, uh, you know, what have we learned? And then finally, what are we going to do differently next time?

Sebastian: Right.

Radhika Dutt: This is what I want out of teams. On the other hand, when you have okrs, what happens is you're just focused on the task at hand. I have to 20,000 signups or get X number of weekly active users. And that's like the focus and it leads to the loss of this narrative. You're not using this part of your brain where you're really learning what exactly am I, what's the story behind the metrics? What are they telling me?

Sebastian: Mhm. Okay, understood. So it's, it's also. So, so basically the criticism is also like in that direction that you focus too much on like the absolute numbers instead of focusing on the qualitative, uh, qualitative insights that you can generate through like keeping like the results a little more open. But isn't that also like being the devil's, uh, advocate here for a moment? Isn't that also something you could incentivize by coming back to what you said earlier, setting the okrs right.

Radhika Dutt: So my, the setting the okrs right. Like, and this is an example that, you know, I was talking to, um, Tammy Rice, um, who uh, is a product coach and she was sharing from her own experience. She was working at a company where they, instead of setting a target, they set the goal of I want you to figure out this puzzle of figuring out the conversion, uh, funnel. She described to me how giving the team this puzzle of figuring out the conversion funnel led, uh, to a very different approach in terms of experimentation. If you said get 20,000 signups, what you would focus on is a certain part of the funnel. So if I say, you know, get this conversion funnel, figure out the puzzle of this conversion funnel, show me that, you know. But is that, is that the right sort of okr? Like typically, okrs are supposed to be measurable, you know, a target that you hit. They're numbers that are supposed to be, ah, something that like John Doerr describes it as. You can ask in a binary way, have you or haven't you achieved this? Um, and that's the problem, you know, where it leads to less experimentation and it leads to focusing on things that limits what the team is going to work on. I'll give you an example. There was an OKR that John Doerr describes in his book as an example of how. This is amazing. Uh, it was an OKR from the Bill Gates foundation about malaria. And the objective was, um, eradicate malaria. And some of the key results was show that drug, um, based approaches are effective in eradicating malaria. Well, how well did this work? This is, according to John Doerr, what is a good okr, right? What actually happened in reality? And there have been, uh, so someone, um, who is heading up malaria research within the WHO actually said that it created a cartel of malaria scientists who are pushing for one approach, a drug based approach. Why? Because that was the okr. What if instead you ask the puzzle question of like, okay, how do we solve this problem that leads to better solutions that are more practical, not just drug based. Even the objectives sometimes, right? Sometimes as leaders we decide that we know this is the right thing to solve. Even this example of malaria, right. Bill Gates foundation decided that this is the problem that needs to be solved in the developing world. And there was a publication, an uh, article on Lancet, a well known medical, uh, journal that said in a lot of the developing countries, malaria is a very small percentage of what causes deaths. Like there are a ton of other diseases like including just, um, you know, uh, gastrointestinal issues, even, you know, road deaths because of traffic accidents that need to be solved. Like those are higher percentage of, you know, causes of mortality compared to malaria. But as a leader, you've decided this is the puzzle that we have to solve. And so this is where instead of an object okrs approach, where leaders decide this is what we do this is the solution. Go for it. We need a more open approach, which is as a leader we figure out what are the puzzles that we need to solve. Involve your team in figuring out what are those puzzles and then use a hypothesis and learnings based approach. And that is what I talk about in the book, like what we need to do instead of okrs.

Sebastian: Got it. So I have lost track a bit here. So we were with the sandbagging, right. So that's where we started.

Radhika Dutt: Those were the issues with. That's right.

Sebastian: So, so yeah, so those are the issues with okrs. So to wrap that up. So there's. And you have to help me with that. So there's sandbagging, basically the issue.

Radhika Dutt: Yeah.

Sebastian: Oh, you go ahead.

Radhika Dutt: So there's sandbagging, which is like, yes, that even ambitious people need to feel like they need to set easier go. The second is it focuses you on very specific tasks as opposed to looking more broadly. So a lot of people describe it as. I feel like I'm constrained to just work on those things. The third thing is it leads to very often teams working on um, outdated goals. So basically, if you look at large organizations, when you're setting OKRs, you negotiate OKRs across cross functional teams. You do that once in a year and once you've set those goals, you're kind of stuck with them. And you know, this is where OKR experts go like, oh, you know, you just need to set them every quarter, so you just do it more often. And I heard an executive respond to that going like, you know, we would die if we had to set these goals multiple times a year and negotiate those. And so I would much rather be stuck with, you know, goals that have been set. And uh, and they may be wrong, but like we would die if we had to take the other approach of resetting them regularly. And then one last problem is this issue which I talked about is, um, how do I frame these metrics for management trying. This is the corruption of metrics that I was talking about. How do I show management that this is working as opposed to genuinely questioning is this working? What have I learned? What are we going to do differently?

Sebastian: Okay, got it. So I mean that's quite a handful of. Right. So and let's, let's see if you get, get mail after the event. Because I know there's like a lot of people out there that are trying to uh, adopt okrs. Um, I'm also not gonna lie. It's like you are not the only person out there and also not the only person I have talked to that um, is basically claiming that we as a community are either using okrs wrong or should be doing something else instead. Um, so yeah, uh, it would be super interesting uh, to see what kind of feedback we get um, after the episode. Okay, so now we have like all these issues and you already hinted at like you already hinted a Solution with uh, OHLs, but can we maybe get like a little more like a little deeper into it? So if you would like walk me through what we should do instead. So instead of these quarterly uh, OKR planning thingies, the objectives being set by management, the key results being set by the teams, like basically cascading into the overall company objectives. So what should we do instead?

Radhika Dutt: Yeah, um, the approach that I advocate for is as a team we are a lot more motivated in solving puzzles together and feeling like we're making progress in solving this puzzle. So as leadership we have to start by defining problem statements and leading questions. Let's look at um, ah, what these problem statements might look like. Uh, here's one. So our customer services organization is not self sufficient enough. You know, they, every time a customer call comes in they have to go to engineering. And uh, so our engineering team gets pulled into a ton of issues. So you can say that as a problem statement. And so instead of setting targets and goals like okay, you know, reduce customer service uh, calls to engineering by 50% like what is that going to do? That's just going to mean that the customer service team is going to feel like they can't reach out to engineering. But that's only going to mean that you know, the customer suffers. Right. Or um, issues take longer to get resolved, um, or like things go underground. Just don't record things, just go to engineering and so on. Right. Uh, so what do you do instead? So describe this problem statement and instead of setting key results, you then set leading questions like okay, is this because our customer service team needs more training or is it because our product is not stable enough in the field? Is it because our um, uh, our product maybe is too complicated and we need more logging and self service tools for customer service to be able to do the self service and figure out what's wrong, like set a bunch of questions and now you've given this puzzle to your team. And so as a team, right, you have, you can uh, work with leadership and set an objective. Like this is what we want to solve for. We want engineering to be able to not constantly be pulled into customer service calls. And so if that's the Objective. Right. Um, in terms of uh, hypotheses, you know, you can write a few hypotheses that you believe are the key problems and then the learnings is where you go and figure out is this hypothesis working. So one hypothesis might be okay, if we um, create logging tools for customer service, they'll be able to do self service and not pull in engineering because they have access to figuring out what's wrong. Now that's a hypothesis and you're basically describing if we do this experiment, this is what I expect as an outcome because this is the connection. Now the next step is learnings. You go work on logging and you try to see have we made this hypothesis any better. Then you might discover the next step. Okay, the team is doing a ton of logging, uh, or is able to access all the logs. It just turns out to be a really shitty product in the field that has so many issues that, you know, they're overwhelmed. And so then you have the next set of hypotheses and learnings. Right. And so what we want is a system where we are describing objectives, hypotheses, and we're talking about learnings really openly between the teams that are doing the execution and with leadership and we discover what's working, what's not, and what are we going to do differently together next. So that's the kind of collaborative learning we want to have happening through OHLs. And at that point it's like saying, you know, I don't need to set a target to see if this is working or not. It's the communication and constant feedback loops that make you realize, yes, we've solved the puzzle or we've not solved it. Right.

Sebastian: What I find interesting in that regard is, I mean, how um, do you do expectation management in these scenarios when you do. Oh, because I mean, at the end of the day with OKRs, there's always, let's say, a certain expectation to it. Right. Even if it's kind of like, like a stretch goal or like should be. I mean, let's be fair, not all organizations are doing it that way. Like some say we, we set realistic goals and not, not um, very ambitious goals. Right. But uh, like how, how do you manage expectations? When you, when you simply define like a problem statement, you come up with a couple of hypotheses that you iterate on and share the learnings. But it like. Yeah, I think, yeah, I think like, I mean it's, it's all. At the end of the day, I get that. It's all in the communication Right, but how would you expect this expectation management to be done? Especially towards um, top management? Because at the end of the day, I mean they are not in the trenches every day, right? And they expect certain results. And with okrs it's really easy to set a goal saying um, get X from A to B. Right? And here it's more like, okay, we have a problem like you described like in customer service and now we come up with a couple of hypotheses how we might tackle this problem and then we share the learnings, but it's not really telling us anything about okay, how good do we think this is going to get, Right?

Radhika Dutt: Yeah, but see herein lies the assumption that when you set these targets and goals, yes, it feels easy as a leader to set that target and a goal, right? But the assumption is this is what the team needs, that you already know what is the right number. Right? Um, if you don't. So I'll give you the example going back to that 20,000 signups. When you set that number at the beginning, you don't know is 20,000. The right number is 1 million. The right number like where in the middle is like what is the right number? So you set some arbitrary number and the team is supposed to hit that. Right? Which is why you get all of this gaming. The point is it feels easier to do management by objectives. This is what we've learned all along in business school that you do management by objectives. But the reality is, you know, if you, whatever number you set, you're going to discover that it's either right or wrong or you need to adjust along the way. OKRs just prevent or make that adjustment much harder. What you really want is the open conversations where leaders continue to give feedback. You know, you might go off and try to solve this puzzle with customer service and like trying to include logging. And as a leader I might be able to tell you like look, this is not working, I need you to work faster. Or this is, you know, the solution you've come up with like adding all of this logging. Yes, but how customer service has to access these logs is a bunch of bullshit. Like you know, uh, it requires so much training, it's not practical. Like do something better. This sort of feedback is, needs to come as feedback. It's not going to happen because of goals or just setting targets. Like I could set a target like okay, go uh, give logs to customer service, but I need to give you feedback if you've done it in a shitty way that is so half assed that it's not going to actually help the team. So it's more about communication. It feels easy as a leader to think that if I just set a target. I'm sorry, setting you off on a mission. But you know, it's kind of like the casino where you know how they say the house always wins. I feel like the product team always has the metrics. The product team always wins. Like you can set, you can show whatever stats you want to show. You know that famous statement also lies. Lies and statistics. I can show data to show whatever you want me to show. The reality is you don't want me to just show bullshit. Like you really want me to solve a puzzle and that's what you want to incentive or motivate me towards. Right? You want to motivate me to, to solve puzzles.

Sebastian: M. I mean that's, that's completely fair. Um, I'm just wondering like the. So I mean at the end of the day, yes, setting a goal and setting like a random number, I still feel like sometimes like, hopefully that number is not too random. Hopefully you have seen in like former experiments that you were able to increase a number. You know, for instance, let's take the signups again that uh, marketing is running a huge campaign next quarter. And um, you are not only looking at the sign up rates, but hopefully also how well qualified, uh, the signups are, all these kind of things. And I get it, that you can basically fake, um, or like basically try to bend every metric by doing certain things. Right. But still, I mean, like in practice, like, does this really work for like, well, let me, let me phrase it um, the other way around. Like, how do you need to sell it in big quotations to top management? Because at the end of the day it's also all these things are always a budgeting discussion, right? So for how long, like how much budget do we get to work on this problem? Like for how long can we keep working on this problem? And how well is kind of good enough? I guess that's kind of where I'm going. And this is not like clearly defined in the beginning when you do OHLs,

Radhika Dutt: right, but it's not defined in OKRs either. Right? Like, because you say, this is what I need. You say, go solve this, um, customer service problem. You might set whatever okr, like deliver, uh, tracking or logging for the team. And okay, so I show you that I've done it, I've met my okr, but what then, right? So this is this point that you bring up that doesn't. Don't you need to have such, uh, benchmarks anyway, like the budget question, the question of how much time, all of that, you know, you have to talk about anyway. You have to give feedback to the team. What I see with okrs is that it reduces experimentation and increases postmortems. I'll give you an example, right? Like this, the same example of weekly active users. Um, I was working. So at this company that, you know, I was talking about with the weekly active users, they had okrs for get X percentage increase in weekly active users. It wasn't, it wasn't, you know, it was a very realistic goal. I hadn't, like, if you look at it as an okr, there's nothing you can say is wrong with that. Right. But what I was actually observing teams doing is they were so focused on the weekly active users metric that I wasn't seeing the sort of experimentation and narrative.

Sebastian: Right.

Radhika Dutt: It was all about, I have to show that we're getting this number of weekly active users. So it was, how can I do more promotion of the app? Like, we need to get more people to start using it. It was all, ah, focused on that. Um, you know, if you ask very different questions, like you're saying, okay, forget the number for a second, just talk to me about what is working. I want that experimentation. I want people to figure out, what have we learned and what are we going to do differently. That's actually those two questions, right? Uh, or three questions, is it working? What have we learned? What are we going to do differently? Those are the essential questions for innovation and those questions get skirted. When you have okrs. When you have okrs, I just want to show you what's working. I want to show you a high number for weekly active users. My goal is to show you good numbers and I'm going to hide the bad numbers. Whereas the reality is you want your team to show you the bad numbers and say what's not working? And when you can talk about what's not working, you as a leader have more leverage to be able to say, okay, I understand what's not working. You get to have more granularity in sharing. Am I going to give you more budget to solve that problem or is this not worth it for me? Right? You have more information to make guided decisions as opposed to things being sweeped under the rug. I might be spending a lot on solving this weekly active users problem. If I discover that the problem is so fundamental, maybe it's not worth spending money on that app at all. Right? If it turns out like Our, our Personas don't use a mobile app. Like this is a B2B thing where they only use the desktop. This is their mindset because they're not tech savvy, whatever it is, right. When you are getting people to show you good and bad numbers, true learnings, then you can truly make the right decisions as opposed to what you say you want and people will just show you that they have achieved what you want.

Sebastian: Got it. Understood. It's, it's actually quite interesting because we like two years ago, um, yeah I think two years ago we had a talk on, on our conference, um, by um, Anya someone, um, who used to work at Tinder and she actually talked about how they started doing um, experiments that were basically going toward qualitative insights. Again because they were getting stuck in a B testing. So because they were just focusing on the numbers, they were not really sure anymore whether they were actually optimizing a local maxima. And um, uh, to me it feels a bit like your approach is also going in that, that direction. So basically we widening your horizon to basically figure out is this really the right thing to solve or could we solve it some other way other than being completely narrowly focused on, on, on metrics.

Radhika Dutt: But um, I love what you just said.

Sebastian: Right.

Radhika Dutt: Like let's. Even algorithmically there is a term for this and algorithms, right. When you know something is down this path, you do a depth first search. When you don't know that it's down that path, you do a breadth first search. And what happens in okrs is we're making teams do a depth first search and you don't always know. As a, as a leader you're assuming that this is where the answer lies. And that's the malaria example that I gave.

Sebastian: Right.

Radhika Dutt: Whereas sometimes what you really not sometimes like generally what is preferable is this is the puzzle I need you to solve. So it's a breadth first search.

Sebastian: Got it. So if people are listening and they think, wow, that sounds great, but I have no idea how to get started. So how, how do you start kind of getting that into a company, even maybe like an experiment if they are for instance currently working with OKRs, or they might not. But like how do you get started implementing or like experimenting with it in a company?

Radhika Dutt: Yeah, I love this question, right. Because um, as a leader, if I tell you abandon okrs, it feels super scary for so many reasons. Like you've only ever learned that you do management by objectives. You need numbers, targets, et cetera. And you know, you don't know As a leader is my team really adept at ah, being able to measure correctly and be that rigorous about measurement. If I don't tell them this is the target I want you to hit. So I understand that it's hard to move away as a leader. So what you can do, like one approach is, you know, if you must set okrs, set them, but start to de emphasize uh, the key results and start to not focus so much on is my team hitting this or not? Stop doing the scoring of teams or individuals, um, et cetera, against those key results because that'll get you the behavior of just doing postmortems that we didn't hit the weekly active users, but it won't get you learnings. Right. So that's the first step. What you want to encourage is collaborative learning and open discussions. So as a leader, you know, you start to de emphasize and what you insist your team does instead is share with you their plans for experimentation, hypotheses and learnings. And um, then you know, what you need is not just leaders making this change, but you need the individual contributors and teams as well to adopt a slightly different way of thinking. Like instead of just using numbers to prove um, that you've achieved something, you really need to genuinely question, is this working? What have we learned and what are we going to do differently? And so as a team, like being able to ask this question of yourself to say what is my hypothesis? And then what are some leading and lagging indicators? The leading indicators are the ones that you can tell quickly where you know if, uh, so sales for example, is a lagging indicator. If you get sales or revenue, it doesn't, you don't know kind of what exactly attributed to it. Right. Um, whereas, you know, maybe no for conversion funnel, if you know someone is uh, like more people clicking on something or moving to that next step, those are leading indicators. So think about leading and lagging indicators. And then secondly, don't just stop there because if you only focus on, oh, I've picked out these few metrics, then you're again going to be super task focused. Think a little bit more broadly questioning what are the metrics telling me. Try to weave a whole story around the metrics and that leads to sort of broader, more creative thinking. And then finally you ask this question of what are we going to do differently? If I had a magic wand, what would I do? And so this is how you can bring it into the organization. It's both leaders deciding that they want more experimentation, realizing that they don't have the answers to everything. It's not a depth first search, it's puzzles that need to be solved and then individual contributors and product teams saying we're going to measure uh, extensively and rigorously but it's more towards this approach of what's working, what's not, what have we learned and what are we going to do differently.

Sebastian: So what you're basically um, so what I'm basically hearing is that you could also try this with one team, start this so basically as an experimentation to see how this is going. Because at the end of the day like big change is always super scary. And also I would think that after maybe years of fight, um, some people have managed to convince their organization to adapt OKRs to finally become um, outcome based and then now it's basically a different thing they should try. So what I'm hearing is you could also start doing it on a team level. You could also start doing it in an OKR setup just to get started and try this out, whether this actually yields better results.

Radhika Dutt: I so agree with you. And even as a team, even if you're doing this at a grassroots level, maybe your leadership doesn't know about OHLs, but you just want less soul sucking work. You just want to feel like you're really making a difference within your team. Just start to use this OHLS approach. Think about it as a puzzle, hypotheses and learnings, right? Because one thing that happens in goal driven organizations is that this muscle of figuring out puzzles and saying what's working, what's not, what are we going to do differently, what have we learned? Like that muscle has atrophied a little bit because we're just very task focused, focused on those 20,000 signups or get logging to customer service or something like that. And so just you know, even with your within your team, just as at a grassroots level, start to first get really good at this and then you can communicate upwards to leadership saying, you know, these are the hypotheses and this is what we learned and this is what we're going to do differently. You can take that approach at a grassroots level as well to be able to bring about change and start to create momentum and even if you don't share it, you'll just feel happier when you within your little bubble by actually learning together as a team.

Sebastian: K, thank you so much for sharing this. Um, that, that was definitely interesting and um, probably also a little bit thought provoking because uh, for the past I feel three, four years I've only uh, been intensively talking about OKRS and, um, now I have a feeling I will need to look into OHLS as well. So, uh, thank you so much for this.

Radhika Dutt: Thank you so much. Thank you for all of these insightful questions. Um, you know, something you just said, if I can add one more thing, is, um, you know, you said about now I'm going to have to start looking at ohls. You know, I feel like I can hear this pain that now it's a new methodology. I think of it more like a mindset. Right. Like this muscle that's so important to build anyway. So it's not so much a methodology, more like a mindset of experimentation, of asking questions and then figuring out learnings. So maybe I'd like to frame it as this is kind of the most natural way of working and that we're sort of bringing this back in into an environment that has lost this natural way of working. Working. But thank you so much for these insightful questions, Sebastian. It really made me kind of also frame it slightly differently at the end of this.

Sebastian: Right. Thanks for tuning into the Product masterclass podcast. If you enjoyed today's episode, make sure to subscribe and leave us a review. It really helps us reach more product people like you. If you have questions or topics you'd love to hear about, reach out to us on LinkedIn, YouTube or visit productmasterclass.com.