Episode 65: Shifting How We Measure Success with Jeff Gothelf
Melissa Perri welcomes Jeff Gothelf to this episode of the Product Thinking Podcast. An experienced consultant in the Agile and Lean UX space, Jeff just released the third edition of his popular book, Lean UX. Jeff talks with Melissa about how he’s shifted focus to teaching about OKRs, and why he encourages the companies he consults with to adopt this goal-setting framework to measure success. He breaks down what OKRs are, why they can’t be the only product-led change a company adopts, how many OKRs there should be within an organization, what a good OKR looks like at the executive level, and why OKRs make a “great gateway drug” to organizational agility.
Subscribe via your favorite platform:
Here are some key points you’ll hear Melissa and Jeff talk about:
HR leadership, especially the performance, retention, and promotion factions, is taking an interest in new ways to measure success, Jeff shares. They want to learn how to deploy this new metric across their organizations so they can improve their overall internal and external performances. [5:03]
Jeff shares how he educates clients about integrating their corporate strategy with their OKRs. “These things don’t exist in a vacuum and can’t be manufactured out of thin air,” he says. “They have to be derived from some kind of corporate strategy, product strategy, business unit strategy.” [9:19]
Objectives are the qualitative goals that we would like to achieve; they are aspirational and inspirational, and the value of doing them should be clear. [11:02]
When your teams are too independent, you run the risk of hyperlocal optimization, Jeff advises. “One of the better tactics that I've seen over the years is to take a set of teams and give them the same OKR set to hit,” he adds. “With those teams, we’ve defined what success is.” [19:09]
Jeff describes an exercise he runs with most of his executive clients. They visualize the relationship between impact metrics and leading and lagging indicators in order to identify the outcomes they’re going to work toward. What this exercise ends up becoming is a top-to-bottom customer journey map. [25:44]
Typically, teams get told what to build; they make a roadmap and get it approved. In Jeff’s OKR conversations with clients, he removes the output part of the process. They now have to discover what to build by practicing Lean UX, product discovery, and design thinking. Many organizations either don’t know how to do that, or they do and they make it difficult or impossible to execute the work. [33:09]
Quarterly check-ins allow you to reflect on whether it makes sense to go towards the goals you’ve set for yourself. [38:06]
Being a good storyteller is a key component of being a good product manager. A vast majority of product managers have to rely on bringing people together on a vision they’ve either built themselves or along with a team through storytelling, as they lead without authority. If you can tell a concise and compelling story that ties in the necessary information, that will be a valuable asset. [41:45]
Resources
Jeff Gothelf on LinkedIn | Twitter
Transcript:
Melissa:
Hello, and welcome to the product thinking podcast. Uh, today I'm joined by a very special guest. Jeff Gothelf. Welcome Jeff.
Jeff:
Hey Melissa. It's good to be here. Always fun to chat.
Melissa:
Yeah, it's great to see you again. How is Barcelona treating you?
Jeff:
Uh, it's great. You know, uh, when, when we were in lockdown, it was a good place to be locked down. At least I could see the Mediterranean from my house. Uh, and now that things are open again. It's, it's fantastic. You know, it's, it's interesting. Like you I've been here almost five years and still have that feeling every now and again, where you're walking down a street or a block or riding my bike or whatever, like, oh, wow. I live here. Like it's super cool. Like, and so we're, I'm happy. We're happy.
Melissa:
It's good. That's awesome. Yeah. And I'm sure the weather doesn't hurt either.
Jeff:
It doesn't.
Melissa:
Nope, not at all. That's awesome. If you don't know Jeff Goel, Jeff is a experienced consultant in the agile, in lean UX world. He's been working with product managers and, uh, UX leadership and executive and CEOs all over the world to learn, uh, these techniques about how to make modern software companies. So, Jeff, can you tell us a little bit about what you've been up to, uh, recently and what type of work you've been doing?
Jeff:
Yeah, so the, one of the things I'm really proud of over the last year or two years is Josh Seiden and I got to write and launch the third edition of lean UX. Now I think it's fun. And, and, and I'm proud of it a, because it's the best version of the book yet, right? You get a chance to iterate a book. It's, it's rare, particularly one, you didn't self-publish. Um, but the fact that this is still a part of the conversation is, is exciting and interesting to me. And, um, and it's been, you know, it's been a decade of that book being out on the, the market at this point, and the fact that there's still so much demand for newer versions of it. I'm super proud of that. Um, lately in the last couple of years too, I've shifted a lot of my focus to objectives and key results.
The goal setting framework, that's sweeping the nation, taking the world by storm. And, uh, and so that's been really for one and interesting, and to have some kind of new and different conversations. And it's really interesting too. I'll get into it, I think a little bit later, as well as to who I'm having these conversations with, you mentioned in the intro that I do a lot of work with product managers and UX designers and, and the leadership of those groups. And I have done that a lot in the past. The OKR conversation starts a different place in the organization. It ends up with product and UX and tech and all those folks, but it always starts in a different place in the organization. So I'm doing a lot of that, uh, as well. And, um, and then, you know, the, the best part of the last couple of years is I've transitioned my entire business online. And so I, I work for my home office. I don't travel for work anymore, and that's been, uh, personally and professionally transformational. It's just, it's a whole different way of living and doing work. And, and they've been really good business years. That's the best part of it is that the business hasn't suffered everything's and my personal life has, has benefited tremendously from this. So that's what I've been up to.
Melissa:
Sounds like good things all around. I know we've been talking a lot about the, the not traveling piece of COVID that has been really, really nice. And, uh, after taking 137 flights in one year, I was like, I can't do this anymore. I gotta, I gotta cut back. But I do miss getting out there and, uh, seeing you at conferences specifically, like we used to run into each other what, like five, six times a year, sometimes. In different countries.
Jeff:
I missed that too. Yeah. I think that's the, the part that I missed the most is the, the, um, the serendipity of running into folks in the community. Right. So we see each other on zoom calls and whatever, but, uh, yeah, that part is definitely is missing for me. Just kind of cuz you know, like, like you, like, I have friends all over the world and the only time I really saw those folks is when I traveled for work or for conferences or for events or whatever. Um, and so I haven't seen them in a while, so that the sad side of it,
Melissa:
Me too, I'm looking forward to getting back to that part, but I am excited that we can still keep doing what we love doing online too. So that's pretty fun. But this, uh, OKR framework that we're talking about, the goal setting framework, uh, you just mentioned, it usually comes from a different part of the organization. Uh, I've been doing a lot of work with people in strategy deployment and strategy creation, uh, getting into the OKRs. I think I got an email like 10 minutes before this, about somebody having OKR trouble, where have you seen it start to originate from that conversation?
Jeff:
So the interesting part for me is that the majority of the inbound requests that I get for objective and key results are coming from HR and HR leadership specifically. Yeah. So not product initially. I mean, certainly there's some, but we're talking about sort of the majority, right? The majority has been HR leadership. So folks who are in the performance management business who are in the, um, promotion and incentives and, and the a goal setting business and they are realizing that there's a new way to measure success. There's this idea. It's not that new, frankly, it's 40 years old. Right. But, um, and so there is a tremendous amount of interest from HR, particularly the, the performance and retention and promotion side of that discipline that is really interested in this and wants to learn it and then wants to figure out how to deploy it across the entire organization. Because it's weird, I think to deploy it to only a, a few departments or a few teams or a few folks in the organization. And so that's been the Genesis of a lot of these conversations for me.
Melissa:
That's really interesting. So are they using it to, you know, benchmark people's performance or trying to tie like how they're, how they're working back to the business goals?
Jeff:
Yeah, that's exactly right. So this is, they're trying to see if this is sort of the new 360 review. If this is the new, like, should I incentivize, uh, people to hit their OKRs? Should I promote people for hitting their OKRs? Um, should I, should we, should we celebrate because of OKRs and, and, and what do we measure and what like, is it, is it just, Hey, did you, did you hit the KR or is there more to it? And so that's, that's the beginning of the conversation and it, and it's really interesting and, and actually pleases me because every single agile transformation that I've seen, that I've been a part of, especially big ones. And you and I have worked in some of the same companies, uh, as consultants over the, the last few years, the last thing they do is performance management.
Right? They'll change, they'll change the, the, the physical space they'll change the vocabulary, they'll change where people are sitting and what teams they're on. They're, they'll start talking about know release trains and all this kind of stuff. Right. But the last thing they'll do is actually talk about what do we measure? What do we incentivize? What do we reward? What do we promote? And because it's messy, it's, it's, it's messy. Uh, it's hard and it's risky, right? I've seen this cause I typically work with large organizations. I see this all the time, right? Imagine you've been an organization to say, let's say it's a bank. And you've been at that bank for 20 years and you've played by the rules and you've got your sights set on that corner office. And you've been kind of, you know, inching towards that corner office for 20 years. And then what we're gonna move the goalposts on people I, after 20 years of, of, and, and it's not, it's not just, we're gonna move them further away. We're actually gonna move them laterally to a completely different, uh, set of set of measurements and goals. That's really scary. Like there, there's a lot of risk there for massive backlash, attrition, that type of thing. And so it's not surprising to me that I'm seeing HR come with these questions first and I'm super happy to see it.
Melissa:
That's really interesting. So when I've worked with OKRs, like one misconception that I keep seeing all over the place is leadership kind of going to product managers or going to these teams. And I imagine HR probably does this as well go, what are your OKRs it's like, gimme your goals. And if you know, strategy deployment for these organizations who are, you know, learning strategy deployment, you know, that has to kind of start with the business goals on the top. Um, when you get brought in to help these HR teams set the goals, like what are the misconceptions around it that you run into, and, and how do you start orienting them towards understanding that this is not just like, Hey, product managers go tell me what your goals are so I can measure you against them, but more of like a company wide undertaking, more of like a strategy setting opportunity.
Jeff:
Yeah. So, so they don't know that. Right. So what, what you just said is a big, big part of the initial education, because what you're, what you suggested is exactly right. They're like, oh, okay. This is just a new way to set goals and to measure folks. So just explain it to me and then we'll put it into place and then everything will be fine. And the reality is like, what's the point? Right? Like, like if you're just gonna swap one, one goal setting framework for another and not change anything else, what's the point. There's, there's literally no point in doing that. So the first, the first thing I do is I, we have a fairly significant educational session about just what are OKRs. Now, look, it's not brain surgery. It's a, it's a really simple concept. Um, it's just not easy to implement, but the concept itself is, is really simple, the, the next, and then, so then we explain, Hey, this is what they are.
The second thing we talk about is, okay, great. These things don't exist in a vacuum. And they can't just be sort of manufactured out of thin air. They have to be derived from some kind of corporate strategy, product strategy, business unit strategy, right. Something. So that, that has to exist. That has to be, um, aligned across the organization or the, or the business unit. And people have to know it before they can even think about how to come up up with objectives and key results that make sense to achieve that particular strategy. And so there's a massive conversation to be had ahead of this, that most organizations either dismiss or limit to a select few high ranking individuals, and then sort of assume that everybody else knows.
Melissa:
Yeah. So if you, you were to describe, you know, what is key result? Like what's an OKR for your definition. Let's pretend whoever's listening to this. Doesn't really know what it is like, what's your definition for this?
Jeff:
Yeah. Super simple. So objectives are the qualitative goals that we would like to achieve. So, so to be clear, I said, qualitative, I'll say it again. They're qualitative. They are aspirational, inspirational. They should, the, the value of doing this should be clear. So the, like how does this contribute to the strategic goals of the organization, and their time boxed. So in a good objective statement, you're gonna hear a lot of superlatives, right? The, well, wanna be the best the is efficient, the easiest to use, um, the fastest, um, that type of thing, the, the go-to destination, something like that. Right? Um, uh, so those are the objectives. The key results are the quantitative measures that tell us whether or not we've actually achieved the objective and to be clear, their quantitative measures, say it again. They are metrics. And, and so, and most importantly, they are measures of human behavior.
They are outcomes, right? So those to me, those are the most important and they, and they need to be verifiable with evidence, right? We need to be able to look at a report, analytics, market, research, data, whatever it is that says, this is true. So they're metrics, they're not features, they're not output. And those metrics have to be outcomes measures of, of human behavior. Um, because you can put metrics in there that are not measures of human behavior, for example, a bounce rate, right? I'm sorry, bounce rate is, is, is human behavior. I take it back. Uh, that's always my counter. The one I was thinking of was a homepage load time, right? We want improve homepage load time by 50%, right? That's a metric, right. Um, but it's not a measure of human behavior. It's a measure of the system. If you reduce bounce rate by 50%, that's a measure of human behavior.
And the humans whose behavior we wanna measure are the humans who consume the thing that we're making or the things that we are making. So if you make a consumer facing product, then those consumers are the people whose behavior you wanna change. If you make an internally facing product that other teams use, then those are the people whose behavior you wanna change. And I would argue, and, and we're gonna get a bit meta about this for a second, right? If you, if you make things that are consumed for example, by, uh, by like an executive team inside an organization, you make a report or you make, a budget or a policy, those are the folks whose behavior you wanna change. And so that is what a key result to me is. It's a, it's a, it's a quantity measure of human behavior. That's verifiable by evidence.
Melissa:
So I see a lot of people struggle is around that qualitative side too, like in, in the quantity of side, what you're saying makes total sense to me. Um, and I think we can use a lot of our really great product metrics from it to show like, you know, how people are interacting with it and adopting it. Um, but where I see people struggle is that qualitative of statement because they go, well, let's be the best checkout flow. You're like, what does that mean? Right. Like what does the best mean? So how do you help people who wanna put all those superlatives in there and like shove it with that stuff and make it kind of fluffy? Like what makes a great objective statement when it comes to the qualitative side?
Jeff:
Yeah. So specificity, I think is key. It's one of the things, if you've ever taken a class from me, you've heard that word from me because inevitably as folks kind of go through the process that, that I teach, they start off really big and they do stuff like you just said, like the best checkout process in the world, you're like, really, you wanna be the best checkout process in the world. Let's be specific, right. Like what's, what's happening right now. How can you make that more specific? So is it things like, Hey, we want to be, um, you know, the, I, I don't know the, the easiest, I, I'm trying to think something off the cuff something like the easiest to use, um, uh, grocery delivery service in Spain, for example, right. Something along those lines. Right. Um, so, so the more specific that you can be, and again, it really depends.
We can talk about setting levels, right? But like this assumes that you are the team thats in charge of the entire application, you might say, look, we wanna be the most efficient checkout process for grocery deliveries in Spain, uh, you know, in 2022, something along those lines, that becomes really interesting because now you've actually made a bit more specific about just one portion of the, of the user journey in there as well. And ideally that's a part of the user journey that you and your team can influence as well. And so the more specific that you can be, the, the better the, the objective statement is.
Melissa:
Yeah, I like that a lot. The specificity that, that really, I think, drives home what you're getting at. And you just mentioned the levels too, which is something I think that's really important about OKR is that a lot of people don't anticipate, um, I've seen a lot of teams have like every single person as their own OKR. Sometimes that happens. I find that every level sometimes gets their own OKRs, but then in some organizations like these really large banks, there's like 27 levels of people. And then all of a sudden we have 27 levels of OKRs for our product. What's the right amount of, of levels. How do you think about, like, where does an OKR span and how many should there be in an organization and what should it actually cover?
Jeff:
So let's talk about it from a couple different perspectives. So, first of all, from a leveling perspective, we want to write. So if you're, if you're in a position to write objectives and key results for your team, you wanna write objectives and key results that your team can actually hit. Right? So for example, let's say you are in charge of the authentication process, uh, the authentication portion of the customer journey in your online grocery delivery app. Okay. You should not sign your team up for, um, I average order value metrics or, um, you know, the, uh, some something along those lines, right? Because that's not the part of the world that you control. So you wanna set levels that your team can actually influence, right? So if you're in charge of authentication, you're gonna focus on the authentication piece. You're gonna focus on behaviors like successful authentications, right?
Number of, uh, support calls about authentication, uh, number of password retrievals. Like those are the kinds of things you're gonna, you're gonna sign up for, because those are the things that you can directly influence, right? And the higher up in the organization you go, you might actually end up taking on, right, an entire customer journey, an entire application and entire business unit. It really depends on where you, you sit in the organization now, as you pointed out, particularly in large companies, you may, you, you may, you're gonna have multiple layers. And so the question becomes, when does it become counterproductive to continue and slice and dice the OKRs until every, every person has it? Um, certainly we don't wanna go down to the individual person level, um, at the very worst or the very, the very lowest level we wanna get down to a team should know what their OKRS are.
However, we then start to run the risk of hyper-local optimization. And I I've, I've experienced this myself throughout my career where, uh, a team will have a, a, a key result to hit. And they're just blinders on heads down working towards optimizing that specific customer behavior, regardless of how that customer behavior impacts other things downstream from them, right? Because their job is just to optimize this particular behavior. Meanwhile, they're adversely impacting their colleagues who are two or three steps down the customer journey from them. And so that's one of the biggest risks of slicing this down really, really small. And so I'll give, so I worked at the ladders a thousand years ago, right. In, in New York city. And I worked on the product side of things. One of our main goals, it was a subscription service. One of our main goals was retention. One of our main key results was retaining our customers on a month to month basis.
On the marketing side, the marketing folks had acquisition goals. And whenever the acquisition goal, the acquisition, uh, key result would start to drop a little bit. Those folks had their go-to hammer and their hammer was email marketing, and they would crank it up and they would start sending hundreds and thousands of emails every, uh, every week to get folks into the system and reminding them to use the system. It would murder retention because people would start to get all these, uh, it would drive acquisition up, but they'd send it to existing customers as well. And so our retention numbers would be awful, but their acquisition numbers were great. Right. So they're filling the top of the funnel and were bleeding out the other end. Right. But they didn't care. Right. Because cuz they were solely measured on that acquisition metric and we were, we were being killed by our own colleagues.
It was a real issue. Right. And so when we slice and, and kind of slice and dice our teams into too fine of, of a, uh, of a slice, I guess, right. Um, we run the risk of hyperlocal optimization and not being able to sort of see the forest for the trees. In those situations one of the better tactics that I've seen over the years is to actually take a set of teams, 2, 3, 4, even five teams depending on it and giving them the same objective and key results set to hit right. Now, what those teams, what we've done in that situation is we have defined success for all five of those teams with the same terms. And so now they've gotta self organize. They've gotta communicate, they've gotta be transparent. They've gotta share data. They've gotta be proactive about, Hey, we're doing this and be like, yeah, you're doing that, but it's killing us over here because they win or lose together. And so it's one of those things that you have to watch out for and you have to nip in the bud very, very, very quickly be. Cause it it's, especially in large organizations, it's inevitable that it's going to happen as you start to kind of chop the customer journeys and the processes down to the too fine of a slice.
Melissa:
Yeah. I think that I had the same exact experience by too where, uh, we, our marketing team kept giving credits to people for their first purchase and retention just like hit the ground. So definitely seen that happen before. And I like your idea of, you know, giving a couple teams the same type of metrics and key results. How do you kind of scope that right? Where everybody's kind of working towards the same ones? Like how do you know, let me give a couple of these teams the same one rather than do it individually. What, what types of, um, problems in the organizations would drive you to do that?
Jeff:
Well? So, so again, I think this, this like hyperlocal optimization that negatively impacts other teams is, is a clear sign that we've sliced too thin. I think that's the first thing I think, look again, there's there as you know, there there's no, there's no one size fits all answer to this kind of stuff. But I think, I think there's some common sense to be taken. Like, should we really divide the, um, I don't know, the, the sign in process into, uh, you know, signing in for one team password retrieval for another team customer support for a third, like, like in other words, like there there's some common sense has to be taken into account in these situations that says, look, this is, this is the, the atomic unit of the customer journey and we are not gonna break it down any further. Um, I don't, I don't know what the, the, a rule of thumb here is, but I think there's some common sense that come, that comes into that.
Melissa:
Yeah. It sounds almost like breaking it down into like a job to be done level or some kind of bound by a feature level instead of, you know, getting into the absolute component parts of it, which makes sense because it's really that feature and that, that piece like driving the metric forward, right. It's not like the individual component of the API that does login, that's gonna make login successful. It's like the whole login piece usually has to contribute towards that. So that makes a lot of sense to me. Um, so with the OKRs too, like we were, we've been talking a lot about the team level OKRs. Um, but as we started talking about at the beginning, you know, a lot of this comes from the executives, um, and from figuring out like your business strategy and what's going on that way, what makes a good high level OKR at the executive level? And, you know, what's kind of the scope that we should be thinking about if we're an executive versus a product leader, you know, to make sure that when we get down to the team level, it's not go build this specific feature is our objective, right? From an OKR perspective, it's not like, so bounded that there's no direction, sorry, let me rephrase that. It's not so pigeonholed where they can't actually make any experiments for themselves, or they can't really figure out what solution they should be building.
Jeff:
Right. So just, just as a rule of thumb, and I know this contradicts measure, what matters we don't put output in our key results. Right. And let's talk, I wanna talk about that for just a second. And then I'll answer your actual question. Um, the, the reason why we don't put outputs in key results is because the moment that you put an output in your key result, you're back to fixed time, fixed scope initiatives, right? You're saying, Hey, we want to be the, the most efficient checkout process in online grocery ordering in Spain. Terrific. And then the key result is we'll implement one click ordering by Thursday, right? Like there's no, there's no room for experimentation. There's no room for, uh, diversion. You've assumed that this is going to work. You've assumed that this is what your customers want. Uh, you've assumed a certain way of developing it.
You're gonna ship it by a certain date and you're gonna move on to something else, right. At which point we've added no value by switching your goal setting framework to OKRs. That's, it's completely, um, uh, moot at this point, right? You're just saying, look, it's the same thing. As, as, as waterfall development, we're gonna build this. We're gonna get it done by this date. You did it. Yes. You didn't do it. Sorry. Right. That type of thing. Um, so that's really important to, to remember that, that if, if you're starting to find output in your key results, it breaks the process. Um, now coming back to your other question, so what makes for a good executive level OKR statement? The, the, the terms that we use that I use when I teach this stuff is impact and outcome metrics. So impact metrics are the high level measures of the health of the business.
You'll see them sometimes in KPIs, you'll see them on executive dashboards, sales, profit, revenue, retention, customer satisfaction, that, that type of thing, those are the kinds of things I would expect to see those kinds of metrics in an executive of key, uh, objective and key results statements. For example, let's say there's a strategic objective to expand into north America, right? You're, you're a big player in Europe and you want to expand into north America. So our objective is to become a significant player in our space in north America, by the end of the year, right. That's our objective. And so the key results I would've expect to see for that would be things like 10% market share by the end of the year, at least 15 million in revenue, right. Stuff like that. Like that's the kind of stuff that I would expect to see at the executive level when it comes to a key result.
Now you might argue Jeff, that that's say Jeff, that those aren't human behaviors, I would say, I'd say kind of feels like very high aggregate measures of human behavior, but those are the kinds of, of key results that I, I would expect to see at the, at the executive level. The next conversation then is about outcomes, right? So if those are your impact metrics as key results for the executives, then the outcome metrics become the things that the teams need to achieve. So what are the leading indicators of market share? What are the leading indicators of revenue, right? What are the leading indicators of adoption in this north American marketplace? And, and so those are the kinds of conversations that we wanna have. And generally speaking, those are the behaviors that our users do in the product or in the system. And the exercise that I run with most teams and most executives is a literal visualization of that relationship.
We put the, the impact metrics at the top of a, of a board or whatever. And then we start to say, okay, great. What are the leading indicators of all of these things? And we start to map out a tree of leading and lagging indicators. And then we look at this board, that's filled with post it notes. It's gonna be dozens of things. And we, we say, great. Now strategically, what do we care about for the next six months? And then we try to carve out a slice of that. And it's those slices that the teams can start to work towards, right? So the slice of outcomes, the teams can start to work towards and they'll say, okay, this team's gonna own the teams. Well, the teams can step up and say, we're gonna own this outcome. Right. And we know that it's important because it's a leading indicator of the strategic objective.
And so the relationship here is, um, between outcomes, outcomes, map, as leading indicators to impact metrics and you know, where, you know, point that you stop, I think is what ends up happening with this exercise in reality is that it ends up becoming sort of a top to bottom customer journey map. You kind of work your way towards the bottom. And, and it it's basically starts with like our users do this and then they do this and then they do this and then we make money. And so I think when you get to those first few steps in the customer journey, you've hit the, the bottom of this particular process. And you can still use those outcomes as potential key results, but we're still not into feature, land or outputs at that point.
Melissa:
I like the explanation. I love the customer journey thought about it too. Now I get asked a lot and I'm curious what your take is on it, cuz it, I find it a little hard to explain. It feels like I, I think it's one thing to actually like write these things out and do it. And then it's another thing to try to explain why it works. Um, but I get asked a lot about like platform initiatives. And what about, um, you know, let's say we've been a giant monolith and one of our, our strategies and our OKRs is to move our monolith, to more of a data centered approach where we have a nice platform in the back and we can decouple it into our APIs and our applications that sit on top of it. Um, which theoretically should make it faster for us to develop and, um, get things out to customers and, um, you know, make it easier for us to scale. How do you look at those types of initiatives when it comes to OKR? How do you reconcile pushes like that? Like where do they fit in when we're thinking about these goals framework?
So they work exactly the same. You just have to think of these, the internal teams as your users and your customers. And again, it's coming back to the goal, the objective, why are we doing this in the first place we're, we're trying to become, um, uh, we'd like to become a significantly more agile organization by the end of the year, right? I'm trying to come up with a, a, a good objective or something like that, or we'd like to become, uh, you know, far more responsive to market conditions. Uh, you know, because we're bad at it right now, something along those lines. Well, great. If we're more responsive or if we're more agile about it, how will we know, what will people be doing differently? That's always the key question, right? What will people be doing differently? And, and in this case, people are internal development teams, for example, right?
So we could say that our, our, you know, our average cycle time from idea of production is reduced by 50%. Um, our mean time to recovery right, is, is, is reduced by 90%. Um, we're able to, um, deploy right, a hundred percent faster than we can deploy today. Something along those lines of you're looking for those kinds of internal metrics that indicate that the work that you're doing is actually making the organization more agile, more responsive, more able to react to changing market conditions, right? Your barometer for whether or not the work that you've chosen to do this replatforming, this refactoring of the code base, whatever it is, right? Your barometer is not, did we deploy the new platform or did we refactor the code? It's are people more productive? Are they more responsive? Can they react more efficiently to what's happening in the marketplace? And we know that they are, or they aren't based on these behaviors, right?
That's the, the fundamental difference here you is, the measure of success becomes the behavior change in people externally facing or internally facing. I'll tell you a story. It's, it's related to this question and it's from, um, a mutual client that you and I shared a long time ago will call them big American bank. Right? Uh, big American bank had a very enlightened executive who used to work there. And this platform team came to this enlightened executive at big American bank. And they said, look, and this, the, the platform that this team was in charge of rehabilitated defaulted credit card users. So if you defaulted on your credit card, this was the system that sort of rehabbed you and got you back into the system. So this was an important system, but it was old and it was clunky. And so the team that was in charge of this came to the enlightened executive, and they said to her, they said, we need to replatform this, do a, basically a lift and shift.
We're gonna pick this up off this platform and put all these features down on a new platform. And it's gonna take two years and cost 20 million. Right. And the enlightened executive looked at the team and they said, and she said to them, she said, that's great. I'm gonna give you six months and 5 million, right. Show me value. Right. Show me value after 5 million in six months. And I'll give you another 5 million and another six months. Right? The fundamental difference here is that they came at this with like, here's all the work we need to do. We've estimated it out, right. Is gonna be spot on. I'm sure their assessment, right. And we've estimated this timeframe, which will be spot on as well. Um, and it's gonna take two years and 20 million. And she said, look, that's fine. I believe you. Right. I wanna see value sooner than that.
So, and, and it blew their mind because it got them thinking completely differently about the work that they were about to do. Right. They had to go and do discovery work to figure out, right. Do we actually need a hundred percent of the features from the old system, right? Which of those features do we actually not need at all? Which ones do we need to port over first to start showing value? Right? And these are the kinds of conversations that when you work in a waterfall prescriptive, sort of go build media, these things by Thursday and make them blue approach to, to product development. Um, you don't ask these questions. And so whether it's external users or internal users, you've gotta start asking those types of questions first.
Melissa:
Yeah. I love that. The time to value conversation that they're having there, it makes you really think. And I feel like sometimes people just don't think about, you know, what value you get from a user perspective too, when you do some technical initiatives, like it helps internally, like you were talking about, but, um, if you can release things faster, get things out the door faster, this time to value, time to revenue for the company that also comes from that as well. So I love that exercise that she was thinking through. Um, but also it's making me think too, what you're talking about, it's really sets you up nicely for a lot of the stuff that you teach in lean UX, right? Which is start from the goal and then work yourself backwards into figuring out what customers want. So how, how does this whole process fit in with those types of principles?
Jeff:
So here here's the, the, well, the self-serving part of OKRs for me. So let's say a team gets really good company gets really good at, at writing well written, strategically meaningful, uh, impactful objective and key results statements, right? They've done a great job. They really picked the right metrics. This is gonna make all the difference. What we've removed from that conversation is features, right? Typically teams get told what to build, or they make a roadmap. They say, here's what we're gonna build. And they get that approved right in the OKR conversation. That part doesn't exist because we, we leave output completely out of it. And so a lot of teams are like, okay, great. I got OKRs now, what, what do you want me to build? And, and the answer is you have to go discover what to build. You have to practice lean UX, you have to practice product discovery.
You have to practice design thinking. You've gotta go out there and declare your assumptions and write your hypotheses and design experiments to figure out the best combination code, copy, design value, proposition, pricing, model, business model, whatever it is that's going to achieve. The behavior change that you're looking for. Now, the sad fact is that a lot of organizations don't know how to do that. And so this is a real opportunity to help teach them lean UX product discovery or the even sadder fact is that some organizations do know how to do that. And they make it difficult or impossible for that work to actually take place. And so that's the next difficult part of the conversation in, in OKRs is, okay, we did a great job writing them. Now we actually have to go do work differently. Because like I said, at the beginning of this conversation, if we just swapped out the goal setting framework, and then we just dictated a bunch of features, what's the point we're wasting time. Just, just stick with the old stuff and keep going. Right. So this forces a conversation around discovery research lean UX that most organizations haven't had yet.
Melissa:
Yeah. One of the things that always gets me too, when we talk about like strategy deployment, especially the OKR framework is I think people walk into this with this notion that, oh, we're just gonna get in a room in November and just everybody's gonna write their OKRs for the next year. Right. Instead of like, no, when we talk about deployment, like literally you have to deploy it in order to create it, right. It's not OKR writing. It's OKR. Start from the top, go out, do the research to figure out the next level, go out, do the research to figure out the next level. And there's so much experimentation and research and data gathering that happens at each point of when you have to set an OKR that cascades all the way down to the teams and then comes back and refines it all the way back up the organization to make sure it's all aligned across the board.
Um, and that it's like, a consistent process. It's not a thing that we do in November, right. To me, like one of the biggest misconceptions about OKRs is it's not just like a way that we measure success, but it's a way that we should work. Right. We should be constantly reevaluating. Where are we in relation to our O KRS doing the work to get there and then saying, Hey, did we reach it? If so, what's the next one we're gonna set instead of waiting for this one weird point of time, that happens to be November usually and saying, right, Hey, let's do it again. Like, let's just write those things, um, once a year. So what types of like cadences do you work with, with teams? How do you kind of get them into the mentality? This is a continuous thing. This is not just Ta-da, like we've got are things for 2022, let's go.
Jeff:
Right. So spot on. Right. So OKRs as much as they're a, a gateway drug to lean UX, they're a great gateway drug to organizational agility as well, because to your point, right, we set a goal and that goal is behavior change. And we say, okay. So our goal for the next quarter is to, um, you know, to increase, uh, average order value for our, our high value customers by 20%. Okay, great. And, uh, we work on it for 12 weeks for a quarter and we did it cool. What do you wanna do next? Right. I got three more quarter this year. Right. Um, or maybe the team comes back and says, look, we are pounding our heads against the wall here. I think we can squeeze 8%, maybe 9% average order value increase. But realistically we're maxing out these folks, right. We're never gonna hit 20%.
Cool. You've got evidence to prove that what do you wanna do for the next three quarters? Right. That type. And so it, uh, it really provides an evidence based objective conversation about how to focus and, and, and, um, prioritize the team's work. Now what cadence to me start quarterly and why quarterly simply because it's less than a year, right? Like you said, we've got this season in November, typically where everybody sets everything up, they hang around for a year and then we come back and visit it, revisit it in a year. But maybe we, we hit the goals already ahead of the year. Maybe we'll never hit the goals. Right. And so quarterly check-ins allow us to reflect on whether or not it still makes sense to, to, to go towards these goals that we've set for ourselves. Right. If it does terrific. Right. But if it doesn't, well, then let's decide where to change and how to change.
And to be super clear about this, this conversation, this quarterly conversation should not be a surprise. Everybody coming to that quarterly OKR check-in should be fully aware of what's happened along the way. So there's a responsibility here on the team to communicate proactively and transparently both up and out so that when we all show up with a quarterly check, it's not like, oh my God, you only hit 8%. I thought you were gonna hit 20. Right. Everybody should be fully aware about why we're only hitting 8%. What we've tried and the conversation should be forward. Looking not backwards, looking in those quarterly. Check-ins now look quarterly. It's just look, it's just a number. I, I mean, it's just a, it's just a, a cadence that works for a lot of, but there are people who work in the field or are an industry where they can't get data fast enough in a quarter.
Right. Or their production cycle times are, are much longer. That's fine. Right. Just do it sooner than you normally do. Right. Cut, cut your current process in half, cut it into quarters. Like if it's two year cycle that you like a product development cycle that you work on right now, right. Check in every six months, anything that's, that's sooner gives you a greater opportunity for course, correction based on evidence. And it makes it infinitely easier to make that course correction because you've invested less in your current thinking, right. If you spend a year on something, you're gonna hate having a change course, you're not gonna change course. No one's gonna change course. Everyone's just gonna, you know, die on that ill because we burned a year's worth of budget on it. Right. But if you spent three months on it and you find out that it's the bad idea, it's just not gonna work. Cool. We got nine more months. Right. So, so there's, there's a phenomenal benefit here to shorter, uh, shorter cycles, not just in your product development, obviously, but in your, in your OKR cycle as well.
Melissa:
Yeah. I find like so many people look at what we talk about with lean UX and experimentation. They're like, well, I work for like this giant aerospace company. I can't possibly experiment in a month. And it's like, you don't have to, if it took you five years to do something before, maybe get some information in one year instead, right. Like that's still better than five years. Um, it's
Jeff:
Five times faster.
Melissa:
Yeah. And then you can actually do something with that. Which, which I like, I, I just think that mentality is like so important to understand. It's not, it's not about doing something in a week. It's about doing things in less time, like you were talking about than what it normally takes. Um, but you also just mentioned something about communication too. And like going into this meeting, uh, this is where I get a lot of questions about how do you communicate your OKRs and your strategy. So, um, I've seen like a lot of people throw 'em on a dashboard and I think that's great, but I don't think that really communicates the intent between behind the entire O OKR. Right. For a lot of situations. So, um, I make people write memos. Uh, I have them write like two, three page memos on, you know, what's their strategy and why do we pick these OKRs?
And where were we coming from? And what's the vision and what are we doing versus what, what we're not doing? Um, sometimes that's the only time executives have actually sat down and wrote out exactly what they want to do for the next year or so. But I feel like that communication piece is so important. What, what have you done that works with your teams and your executives and your product managers at every level to help them really explain these things so that when you do walk into the quarterly meeting, people understand what you're talking about.
Jeff:
Yeah. So the, I I'd take a slightly more generic rather than tactical approach, but you're, I might steal your approach. Cause I really like it. It's a, it's a tactical person, but generally speaking, look, I, I talk to product managers explicitly about storytelling and, and being a good storyteller as a key component of being a good product manager, right? The overwhelming majority of product managers lead without authority have to bring people along with them on a vision that either they've come up with or that they've, they've built together with a team. And why should I go with you? Why should I believe in these goals? Why did you choose these, these behaviors as the ones that we focus on? If you can tell a concise and compelling story that ties in the strategy, user data, the, the reason why you chose these and the benefits that you think will achieve.
If we hit these goals, to me, that's that can become a, an extremely compelling part of the conversation. And, and it takes practice and it takes discipline and it takes work, frankly, like you have to go do the research, right? You can't just say, oh, these three look good. And let's just do these KRs right. You really got, gotta make a compelling case for where the problem that we're seeing. Right. So, so in lean, UX, we always start with the problem statement, right? What's the problem statement that we're trying to solve, right. That's a great way to start the conversation, right? What we've observed over the last six months is that people sign up for the service. They place one order for grocery deliveries, and then we never see them again. Right. That's a problem for us because our cost per acquisition is $15 per user.
Right. And if they don't buy at least $15 worth of stuff on their first order, we're losing money on every one of these folks. And this is 75% of our user base. Right? So that, like, in other words, like this is become a good story, right? You're you're coming along. You're like, okay, this is, this is worth solving right. This is fundamental to the business. And so if you can tell a compelling story, you can sell the OKRs to your team, your executives, your stakeholders, whoever's paying attention to it in a way that motivates them to help you achieve them. So storytelling, absolutely. The key to all of this.
Melissa:
Yeah. I think that's probably one of the most, some important tools for a product and, uh, using the data to make it colorful and to really bring it to life. Absolutely key for explaining your strategy and what you should be doing as a product manager and why you should be doing it. And we always love to talk about like bringing developers on the journey and, you know, making your team sight about it. A lot of product managers out, out there asking about like, how do I get my developers on board with this? How do I get my stakeholders on board with it? And I think your advice is absolutely key for hitting on that for really making sure that people come along with you on that journey. Uh, so this has been fantastic. Jeff, thank you so much for sharing all of your O OKR advice with us. Uh, if people wanna learn more about OKRs and your work and lean UX and everything, where can they go to read more about it?
Jeff:
So the best place you can go is Jeffgothelf.com your source for all things Jeff Gothelf. Uh, Nope, but that's, that's really, I, I write up the blog there. I, I write a weekly blog. The overwhelming majority of the topics in the last couple of years have been OKRs. And, uh, and you can always please connect with me on LinkedIn, very active on there as well. And, and you'll find links to, to blog posts and my OKR course and, and everything else on there. So just go to my website. I'd love to see there.
Melissa:
Great. Well, thank you so much for joining us, Jeff, and thank you all for listening to this podcast. Um, if you enjoyed this podcast, please subscribe to the product thinking podcast. So you get a new one in your inbox every Wednesday. Uh, we'll have some more great guests and we will see you next Wednesday.