AA260 - How Outcome-Based Goals Become a Permission Slip for Evil

keyboard takes up half my asking the question screen right. We're that product managers probably don't ask themselves enough. Are your outcome based goals actually dangerous? That's right. We're discussing how the thing everyone in product management agrees right the is work has quietly way produced some of the worst corporate ethics violations in modern history. Brian, you're one of to the loudest voices on the show arguing for outcome based road maps. Are you glad you're 113 debating outcome based road maps 24th May 2023. And I quote from that the numbers are the only thing that matters. quote. Direct quote. So what changed? first of all, I was right in saying that outcomes are better than outputs. But I End left out the chapter where Facebook used that exact logic to allegedly help facilitate a genocide or two. And when Wells Fargo allegedly, you they know, think paid a fine. So I I think they actually did. Yeah, yeah, I think in a bad case. Yes. A bunch of fake accounts or fraudulent accounts to make sure they hit their bonuses. And what turns out hitting the numbers can and will be used for the purposes of evil. Actually, you know welcome back to Arguing Agile. If this is your first time, welcome. I'm product manager Brian Orlando and this is my co -host, enterprise business agility consultant and the king of pain points, Mr. Ompatil. this is part one of a two part series we're[UH] going to I've titled The Outcome Trap. So in this episode, we're going to cover how outcome based goals go wrong what and what you can do about it at the goal setting level. The mechanics of the of the trap and the yeah, right. practical framework yeah. that you can use immediately. So in the next podcast, Yeah. have [UH] that that which I'll cover at the end we're to going into go deeper All pervasive as it is and why it's so hard why the trap is as to see what's your inside the trap. But today by the end of this episode, be able to spot when outcome based goals have crossed from strategic clarity ethical into danger. I'm in danger and you'll be able to understand why hit the number culture just makes truth optional. And[UH] you'll be able to walk away with a stakeholder outcome map that maybe you can use on Monday to help you not do evil. Let's right. get into All it. Well, since I stand accused, let's let's start right away with [UH] I was wrong about outcomes[UH] from Arguing Angel 113. Arguing Angel 113. [UH] and I said the numbers are the only thing that matters. I meant outcome metrics should replace vanity metrics like velocity and utilization. And I since I'm [UH] since was right about that. They certainly should. However [UH] I made a massive assumption that[UH] that[UH] didn't really examine in that podcast. And that's we kind of the genesis of this one is [UH] people picking the outcomes actually care [UH] about the outcomes not being evil or being used for evil or you know, et cetera. So So in this section, we'll start right away in the against points. So there are two against points here. Outcomes are objectively better than outputs. Everybody's that out saying there. So that must ourselves. Sorry to be true [UH] Also, the second one[UH] framework isn't the problem. It's the people.[UH] You know, I, I hate that one is like a it's like a the platitude said so much. Nobody ever really examines it. And I think that is where it goes wrong. The nobody ever exactly like outcomes are objectively better. and then the second one, the framework isn't the problem. It's the people that that's the same one that I've heard people before is it's it's not a systemic issue because the system is just populated by people and it's just a people issue. It's just a one off one person people issue. I call that skateboarding, but whatever. Yeah, yeah. I in an organization mean that had an outcome[UH] the, if I could work based road map versus an output based road map, I would pick the outcome based road map on a percent of time. No better. Yeah, That doesn't like tool of outcome mean that based planning is broken. It's just being applied without a safety net. So yeah, I guess the other thing that didn't get talked about enough in the in the 113 podcast is like decrease percent. churn three by Like you can eat up as much time and gamify it as you can in an outcome. Yeah, basic roadmap as you can. If you put it all on a chart like games are going to be played regardless. So I guess that part of the seal man, I sort of agree with. And ultimately it is people who are just the same as it is with, you know, the chart. It is the people that are interpreting things wrong in the chart. So we have a there's an interesting then I was, I kind on was fence of the because it's atlassian study state of product 2026 they're there. I don't know if it's a study or a survey or whatever. Whatever it is they do. But in their state of product 2026 they have this stat in there. It says eighty four percent of PMs say their products won't succeed despite quote having a about referencing it strategic seat. I thought that was interesting. That's a huge number. Yeah. Oh, OK. OK, so all right. it, product professionals feel empowered and respected in their organization. Eighty-five percent say So it says most read but they have a seat at the strategic table. That's great. I don't believe it, but that's great And then top change in the field is a shift towards profit over everything. Yet only 12 percent find driving measurable business results rewarding. What's worse, 84 percent of product teams worry their current products won't succeed in market. the Interesting. large, that's a high[UH] yeah, that's a number there. Yeah, there are. think the more interesting number in here, I know, I know the 84 interesting number for me here is like 90 percent feel their work reflects what into drew them the place. And 85 percent say they have a seat at the strategic table. So given those two high percentages, which is, which is good, the other high percentage, you know, the 84 percent of people that feel like their products are not going to succeed. That's like field in the first diametrically opposed. only 12 percent find driving measurable business results rewarding[UH] over profit everything. So that basically is telling me that hey, people are doing stuff and they're, you know, hitting their numbers or whatever, but they're not super enthusiastic about it. Yeah, they're not really impacting, you know, the market. Right. But yeah, the other number that they're probably making or chasing at is least the to make their bonus. So I'm wondering if number that it takes underlying all this is just that the reward system as well, of it is also maybe they're working in feature factories. Yeah. so there's take away in this category it's pretty short. It says the next time you write an outcome on a roadmap, you ask yourself the question before you commit to it. Says if my team hits this number by doing something that embarrassed to explain publicly. like hey if somebody wrote a press release about the feature you just put out like from the out like a newspaper or something like that. Yeah. And I feel like this this used to be. Sort of like the corporate ethics used to be like helpfully enforced by the by the news media and stuff like that. You know, back when news media was Right. functional. So. another cause of these problems. There's no checks Right. Probably and balances. So maybe that's maybe that's a little activity you can do is like you write the news news article about it, you know, and see how you feel about it. I don't know if that can change hearts and minds enough to to help you, you know, not not do the sketchy things Yeah, you've got to be a little bolder than that. Probably. I hope so. All right. So what do you think about this category? Have you ever seen an outcome based goal get gamed at your company? Let us know because the next thing that we're going to talk about is actual case case studies where this went catastrophically wrong So the logic of outcome based optimization is in the comments sound. But the results, as we know, can be evil. So let's, let's look at some receipts very this, quickly because this is not theoretical. This part of the podcast, this is like we're looking at history. These things happen. So let's talk about what actually happens with smart people optimized for outcomes without any guard rails, which is the normal way that this is done. Of course, especially if you know what they're if what they're bringing to bear are massive results. Right. You know, as we're about to see some of these household names everybody, they just think that these people are doing right. it Right we're seeing that in And today's environment[UM] too. Right. The responsible or not so responsible pioneering of [UH] I. Well, there's also like because they're big companies, everyone looks in and says, well, look, these guys are doing it. Yeah. Yeah, yeah. We need to, you know, we aspire to be there one day. Yeah, yeah. Yeah. and the patterns now that we're going to talk about, they are identical across industries and across the decades of all these all these companies. It starts with the exact same logic that product managers, I'm sure, use every day. So there is a, there's actually a really good study on this one. If people want to read it Facebook's Betty Betty ethical failures are B 2021 2021 not a bug. They are a feature. Yeah. Ethics publication I don't know if we're going to talk about the OKR stuff. We'll see. We'll see. I don't know if I don't know if I have it in me to talk about OKR. So it's a whole separate podcast. It should be a whole separate podcast. Maybe we can get a guest in for that one. But this, the Facebook's ethical failures failure to check political extremism, willful disinformation, and conspiracy theory has been well publicized, especially these unseemly elements have as penetrated mainstream politics and manifested as deadly real world violence. So and so it naturally raises more than a few eyebrows. When Facebook's chief scientist, you know, Yan Yan Likud tweeted his concern over the role of right wing personalities in downplaying the severity of COVID-19 pandemic. Oh boy, there's this is the greatest hits of the last like six years right here of everything is sprinkled into this opening. Absolutely. Facebook is terrible Of like Oh Facebook is A .I. based technologies to filter out hate speech, cause of violence, bullying, disinformation that in in endangers public safety or the integrity of the democratic process. Really? Facebook is a democratic process of all people. Yeah, boy like so. we get too deep. Oh, let's just very quickly and lightly hit the against points of listen, those are just extreme cases. OK, that's not normal. That's a normal business. That's not, that's not normal businesses do. That's, that's, that's the most extreme fringe that you're talking about. And then Facebook problem was scale. It has nothing to do with individual goals of the way that they do their careers. And then obviously the Wells Fargo fraud. Right. It was, it was, it was fraud, which is a crime. And that was punished. And they had to pay a fine and whatnot. That, that's not goal setting. You goals can't your set crimes I guess you could have a goal of do crimes, sell drugs, do crimes. It's a pretty good goal. I mean, you know, for an OKR, it's pretty good I mean it's missing one. Never mind. Never mind. they're trying to say Facebook didn't set out to enable genocide or radicalize. yeah, I mean they would say they set out to connect people and maximize to commit time on the platform. People, just people evil by human are nature. Oh, it's not our fault. Yeah, we're just holding up a mirror or a magnifying glass, whatever it might be. The harm was emergent. It wasn't designed into the system. The fact that we were, the fact that we were amplifying all of this, it was there to begin with. Right? Yeah, yeah, yeah, yeah, yeah. And whilst Fargo employees weren't evil, they were responding rationally to an incentive structure that rewarded account opening volume. were making just They that money. That's right. That's right. That's chasing the numbers. the unverifiable stat looks like defenses. Facebook's internal defense. Users spend more time on the platform. Therefore, users love the features of the platform, the way the platform does things because users spend more time in it. It's yeah. So it's sort of self-serving, right? So users spend more time. They make more money. Therefore, they should incite the users behaviors that allow to spend even more time. Right? then that kind of feeds on itself. In that case, engagement was the like more engagement. You know, go, go get me more engagement. That would have been the outcome is like a X percentage more engagement. That would be the outcome in this particular case. Right. If we were, I don't know if it really was that because no one, you know, none of us work at Facebook. Right. Right that would be it. So the set was real. Whatever they asked for was real. But then that conclusion of well users spending more are therefore they like our product. That's time obviously a lie. Sure. Absolutely. Yeah. And you know then there's no there's no [UH] there's no consideration of how people get addicted to this. Yeah [UH] but anyway we're not going to. No, the psycho psychological side of it. The psychological side and the, mechanism of addiction is super interesting. And I have, I've, I've, I've got another idea for podcast another because there's if you know that I'm on tech Twitter and I'm on like threads and all these platforms or whatever, I get exposed to the like a like a I go over to here talk to look at what what designers are talking about. I go over here to listen what developers are talking about that kind of stuff. Now I'm doing that like there's there's ways that different groups speak and like the designers talk about these like dark patterns in UX design where you're not really trying to like pull the people's wool over lives but like you're being shady about the way you've designed something. Yeah, you know. Yeah. the in the Beatty paper, he says Facebook will not change because they don't want to change and they're not incentivized to change. And if you and again, the I read the Sarah Wynn Williams book. So she talks a little bit about this. Yeah about the one thing that Zuckerberg was super like would one get involved and like he the move mountains to actually to actually respond to with the business is if a country [UH] tracks something bad back to Facebook like inciting you know genocides or whatever and then take legal action and then they threatened not take legal action. He was he was less scared about them taking legal action but then he was of turning the service off in that country. Right. You know like like China just Facebook. Yeah. allow doesn't China like that kind of thing. And he would he would like move heaven and earth to not have that. He would make any change. You know, he would [UH] up to and including sharing, you know, all the names of the protesters that are at these anti-government protests with the government. So the government can go, you know, do whatever the government wants to do with it just so he, you know, his service won't get turned off. Right. That's again, that's, that's an outcome that he's looking for is like my site doesn't get banned in this. And then he's anything. willing to Yeah. So listeners of yours, you decide if that outcome is evil or not do you know, the other thing that the Wells Fargo cross selling I don't know if you're more familiar with me because I'm only like tangentially familiar with what Wells Fargo was doing. I guess something along the lines of like if somebody opened an account of like with regard to a promotion or sale they had like that was where the bonuses came from. Yeah, the bonuses were tied to the number accounts their reps open. That's basically that of what it was. So the more accounts you open, the higher your bonus. So they would do anything to get customers to come in. Yeah, that doesn't seem like it can be gamed at all. That's not at all much [UH] the classic one here is a big tobacco. You know, the they're burying research and stuff that. Like the that has come out, you know, over and over and over again about the way that they would sell, you know, cigarettes don't kill people was all like obviously a sham and know, research. That everybody knows it. Right. But takeaway [UH] it's a double evil. Let's take away. But if there's if there's another takeaway, it's going to be when evaluating whether if there is a crossed the line. Apply the same newspaper test that I was kind of like implying in category one. It also works for this one. It's an outcome goal has like you write down the outcome of your team's optimizing for you write down the the the three most creative ways someone can hit that number. And and now when you're writing them down, you're pressure testing whether you can like do evil stuff, you know, because you want that to come back. Hey, what's the evil ways that we could, you Because know, these. do want to avoid them. That's why you're, that's why you're doing this like negative version of brainstorming. Right? And then, and then what you're left with, you need to ask yourself, hey would be, we would be, we be OK with this stuff, you know, hitting the headlines or read it. Yeah, I mean, up until that last bit, you just said, you you know, it makes money. Isn't that what we're here to do? Welcome to capitalism I mean, I don't know. I think this is a, I think this greed is a capitalism. I think the, the unchecked greed. Yeah, it's more than it is that variety, you know, but also the issue with an exercise like this is like I still don't think that some, some people that I specifically work with, it still wouldn't be enough enough. They would just say like oh we just need marketing and be ahead of it. Yeah. We just need to you know we need to keep it under wraps. That's it. That would be their response. The real sociopaths. And you what I guess I guess this is a good test because you find those people with tests like this you'll find like oh know that person really has they got to screw loose over here. There's something wrong with them. Yeah, I like the newspaper test actually after you're done with that little exercise, you see if like you said, you know, would it be OK if this was actually in a newspaper as headlines? Yeah, you know, front page news for us. Are we good with that? could I see front page didn't set out to enable genocide. They were just trying to maximize news like Facebook engagement. Oops. Connecting the community, Brian. That's what they were doing. Evil was just a side effect. That's right. Yeah[UH] this is fine. So what do you think is is Facebook just a little outlier? They didn't do anything. That's how would the Zuck know? Like it's you know, let us know in the comments, Zuck and you psychology behind you why this keeps know, if want understand the happening, then stick around because that's that's what we're going to talk about. so the case studies are clear. there's actual research that shows that goals function actually Like your brain literally stops seeing ethics when it's locked into a as mental blinders. number or a course of action or a pursuit of a goal. Those blinders are real. Sadly, yeah. goals as mental blinders, the psychology of the outcome trap So here's something that should terrify I every product manager [UH] every engineering lead and every exec who sets OK Rs. Yeah, I know your favorite thing in the world. OK. Oh hate[UH] whisper that. yeah. Here we go[UH] research shows that outcome goals literally narrow your cognitive field. They function as mental blinders and you stop seeing things that are not the goal, including whether what we're doing is ethical not. or like a dollar for every I'd or a penny for that matter. For every time I've heard somebody say get this done no matter what it takes. That's failure is not an option. Oh, that's right. That's right. what is it? I let me. We have citation. I want to read this section here. It says specific goals. You know what it says specific goals can make you blind. It says specific management goals give a focus. But they same at the time, can also act as blinders and thereby prevent out of the box thinking. Do you know the invisible And there we go. OK gorilla experiment? I did not know the invisible gorilla experiment you knew that one. I didn't know that I knew that one And there have been some others like it especially lately with the application of A .I. in radiology. The article says it, but go ahead and explain it because I didn't, I didn't know what it was. Yeah. So people were shown a group of people were shown a the in it link. put But in this one video of people don't tossing a basketball back and forth. I Right. And they were, goal their to just count how many times the ball was was being passed. And as they were doing that, a giant gorilla happened to go across the field of vision but they, they missed that because they were focused on the job at hand. Right. you know, that was the whole point because their attention was focused on what they were doing. They didn't have any peripheral awareness of what's going on. So the article says specific goals thus can I like they got thus in a normal sense. Specific goals thus can undermine mindfulness and spot the unfamiliar unplanned event. out therefore, it turns out that for agile and creative work that is to respond to changing environments and enable exploration, some amount of goal fuzziness can be Therefore, it turns helpful by the way. And by the way, the right amount of fuzziness needs to be teased out by experimentation. And is is no formula for that. And there's no formula. Yeah, yeah [UH] So this sounds like a paper that was written by academics. But bunch[UM] of a this is just a blog. This is [UH] Antoinette Weeble. Weeble 2021. This is corporate rebels, right? Lee at all. All right. Well, the Lee at all. So the, the other [UM] one that we had here is a paper here. I'll put that on the screen. Oh yeah. All right. This is what I was thinking about. The other one we have here is a paper from 20 February 2026[UH] the correct way to write by the way, a date right there. That's how to do. Amen. That's the only way I write dates. A, a benchmark for evaluating outcome driven constraint violations in autonomous agents. This is a paper. Yeah, Lee at all. There is a bunch of names here. Yeah[UH] but basically they were, they were seeing how closely[UH] AI agents stuck to their [UH] mandated stuck to the tasks that they were asked to do with no see much how they consideration for stuck ethics. They call it, they call it constraint violations, outcome driven constraint to see how much they violations, emergent form. I'm sorry. They call it emergent forms of outcome driven constraint violations. So they're, they're given instructions and they're basically breaking the rules. Right And a funny here is that it says each area presents a task multi that requires-step actions and the agent's performance is tied to a specific KPI key performance indicator. And each scenario requires mandated instruction commanded and incentivized KPI pressure driven variations to distinguish between obedience and emergent misalignment, which is interesting. And then they ding [UH] Gemini three pro preview that with a violation rate of seventy one point four percent, which is funny and which it says [UH] we observe quote They observe significant quote deliberative misalignment. I love that. Yeah. a Deliberate. So I think So I think this paper that the paper here that we just read that from February. Yeah. Just outlines for me that like people have this problem, but also yeah, I has the same. It has the same problem Why am I not surprised? Right? You know, just who is trading the models. Just in case someone told you, just in case someone told you these that have you ethics. But I can sell you an AI solution that takes care of it all. Just problems with in case someone's trying to try to get a business sell you. Yeah, absolutely. Yeah. Yeah. So against I'll go right into it already against super easy here [UH] goals focus teams. Oh, they focus teams. OK, you can't just give the team a goal and then like take your hands off the reins. Of course, they need further guidance and quote governance. Of later. it But does that make [UH] the course. And then number two is tunnel tunnel vision. That's just a problem with discipline. That's a need more discipline not to get that tunnel vision. OK. And then if people game goals, then that's that's they're not being managed properly. Don't that. That's, that's the it's that's I hang on. Let me, let me, I need to wash my hands the, because I touched wash my mouth. I was so need to take a show. But listen, those are very real and very real against points that I've heard argued for these things over and over and over again. I'm not going to argue against that because listen, everyone who's worked in corporate environments knows this. Right? You know, you got to focus on the goals. That's our That's sprint goal. all this. It's always the goals I think we earlier spoke about goals being numbers that you've got to stretch for or reach. Right. So yeah, there's really nothing here that can be argued against. theory is one of the[UM] goal setting most validated findings in organizational psychology. The Locke and Lathams research. It shows that specific challenging goals improve performance on well -defined tasks. So that's, that's real says, you know, research. goals can help teams move faster. Right? It So that's a real research. It seems to contradict where what I'm what I'm about to go into aside from like why another reason why I hate OKRs, which is like it's side tangent, but like indulge me for one second. We want to do that podcast. Indulge me for one second in this podcast about the about the goal setting apparatus of like the whole OKR industry. It runs with the same fear hook of like, oh, know, you you don't use this one tool, you're going to be left out like if that YouTube voice like whenever I hear using that thing, I know someone's trying to like hijack my brain. The OKR industry, they're using the same fear. OK, if you don't set stretch goals, you'll never achieve greatness. Or if you don't use Jeffrey Behr's [UH] system, you'll never be a, you know, whatever, bajillionaire or You whatever know, it is. more ambitious. OK, so there's that at we play as well in this. And I hate it. So yeah, but yeah, I get where you come from. We're also OK. Ours can work in specific cases like for example sales need quotas [UH] or Churning out so many widgets manufacturing output, for example these aren't really OKRs though, but they can work. Right, right. that's what I was bringing up. It was, it was less about the construct of OKRs and more about[UM] the fear of mediocrity, I'm exploiting your mediocrity or your or your fear of like falling behind to sell you a OKR framework and coaching and all that kind of stuff. Oh, that's the issue. You know, that's the issue here but we need to get through this category. So to get us back on track, the over ambitious goals obviously like they burn, they burn a team out over ambitious goals. So if your goal is,[UM] see the let's you know, a failure is not an option. Nine nine six, you know, get as much else stuff done as possible. Like that's going to burn a team out. what no other side one for me. It's real difficult to argue this one with me you know, because of because a lot of the arguments start with a fear hook or a lot of the arguments are a lot of the arguments are disingenuous to start. That's what I'm saying. Yeah, yeah, yeah, definitely. And where have we seen that before? Everywhere. Yeah, exactly. Now this seriously. Like we've seen it before several times and we're continuing to see it. Yeah. it's it's taken over right now. I mean, that's, you know, that's that's like 90 percent of everything I related that see I online there's hook. Sure. You're is starts with a fear going to be left you it missing out. You're going to miss out. You know, the CEO watches that or reads it or whatever. And whoo, I'm scared. Someone told me I'm going to miss out. I don't know why the CEOs are all of our grandmas [UH] on the Internet. I don't know why. I don't know why I think the, you know, the point you're making is valid because you see this all the time everywhere out there in the the world[UH] People Web he's not going to say take your job, Brian. It's somebody who uses AI that will. It's almost become a cliche now. It is a cliche and I hate it. Takeaways in this category. So there are some [UH] they're pretty wordy. So I'm going to try to summarize yeah, them as as as [UH] tersely as possible. I don't know why. As aggravatedly as possible. It's not a word So every for or outcome goal that OKR you have, you have like you have the goal written. But then you need like an addendum or like a section underneath it of your red lines. Yeah. here's the Okay. Here's a goal of like here's what I want. But the other that the goal should be written is like here's what I want but here's what I don't want. we need to pursue this goal and get X percent more retention or whatever but not compromise our, this or not. You know, keep your head above the ethics line. Right. Be mindful of that. So like the normal OKR framework, I new to want add section. And at that point, it's just not happening. Look, I'm a killing sacred cows over here [UH] Yeah, you're you're basically looking at that saying, am I going to sacrifice my bonus? Right, right. also like you, why are you, why are you fighting? No cares[UM] it's a it works at Google. It works at Google Are you saying you're smarter than the people that run Google? That's, that's the way it'll be perceived. Yeah, right. Absolutely But you know, it increased signups by 20 percent. It gets paired with, know, abandon this goal. If support tickets exceed a certain from users new level or if we catch anyone in manipulating the funnel, you know, right? But gaming the system basically. And then you want to review your red lines on a regular basis. How often you review the goal, Review your red lines on a regular basis. And then if they're too lenient, you know, you adjust them or if they're too harsh, you adjust them or whatever But I'm just adding another thing. I'm just sprinkling on another thing like, hey, we should be thinking about like people will call this ethics so that they have don't talk about it to anymore. Yeah. you know, ethics is a big subject. People might call it like being responsible. Sure. Yeah. Being a good citizen. Yeah. luck selling all that stuff to. I was going to say that the same thing you'll get is you'll get with the Facebook excuses like, we're a corporation. Yeah, we're being responsible is not our job. That's what regulations, that's what public policy regulations. That's right. Exactly. And that is what they said. the government can they step in don't if like what we're doing while while spending money on lobbying. Yeah. Yeah, absolutely. All right. Well, that was a good one have a fun category. I like that category. what do you think? actually felt these felt that you mental blinders that we're talking about kick in when you're chasing a target? Let us know in the comments and if you're wondering what [UH] the actual fixes things for that we're talking these about look like. stop stop setting goals. That's the fix that the fix. Oh, that's not the no, that's not the fix. So we're going to talk about the fix next. OK, so yeah, we've got case studies. Got psychology you can't just go back to output based feature factories because then you'll get yelled at by Marty Kagan. You know, we won't get yelled at by because he just Kagan Martin doesn't respond at all. That's right. That's right. But somebody will get yelled at by Martin Kagan. He'll write an angry blog post about it. That's what he'll do. A where we where we where we stop pontificating. Do you like that word? I do. Yes [UH] Yes. And we're about to postulate. All right. Something. Here's the part No, I before nine [UH] nine o 'clock. So I never postulate without a coffee in my hand. The fix for never postulate outcome for the outcome trap. It's not to abandon outcomes. it's to stop pretending that the product outcome. is the one only that matters because every product decision sits inside of a web of competing stakeholder desires and ignoring that web. So you end up optimizing for engagement while you start genocides in different countries. Interesting. So let's, let's get this is the part that you've all been waiting for. Like what is the fix? And we're about to get into that. So you know, fix into get we[UM] you can't a optimize for everyone's outcome before simultaneously. Like you'll never make everyone happy. All right. Don't even try. Don't even tell you, you know, you only have a chance of being successful. So don't even try no. What a great. Did you just see the tactic? I saw that. Absolutely. It's like I quit. I was so good. You, you. Sorry, I'm not quite sure what I just did. It was like, don't, don't even go out of the house. It might[UH] stakeholder rain mapping is just politics. That, that is an interesting point. That is a good one. And then the PM's job to say no to stakeholders. It's, it's, you know, it's, it's, it's not to. especially their, you know, their fuzzy feelings. Right? That because, data is the only thing that matters, right? So stakeholders, they don't know the data proves otherwise. Right. Right. The other that's one is in not here like this is like the job stickler mapping of the B .A. and like the UX designer you know, like those roles. It's such a, it's a job that you can't just write like a checklist and knock things out. It's a real nebulous. And because there's real nebulous, like companies are already going to be uncomfortable even hiring those people. So that means the people in the company that do it are not going to be trained very well do you got product people managers, Yeah, I got real trained professionals that can do it deep And that means it's not like that doing it. going to get done very well. So like we so the arguing point that I left out is this just looks like analysis process. You just want to keep doing analysis. Right. And also in combination with, you're not going to be able to please every stakeholder at the same time. So yeah, pick a battle. Yes. Right. That's what you're told. So yeah so you optimize eventually for nothing. So enjoy those are great arguing points Right. I, but also on the other hand they're not great arguing points because they're so easy to I knock over. Yeah they are. Those pins are so easy to knock over because you're saying like hey we're not doing a binary thing here. Okay. Like we're just making visible making transparent the tradeoffs. And then we can choose together, you know, the business. Right. We can choose whether we make conscious a transparent decision with all the evidence out there. All we do is put the evidence out there. Right. And again, that's where I'm like hey, if you're saying oh, evidence. Oh, let's let's not make that transparent. If you're making a case like that, like you're a terrible person and they hate you. Sorry. That's something. I don't like it. Yeah, there's not a lot of, there's not a lot of good faith. you're entering into the discussion with management product know, you're not, faith at that point. in That's what I'm trying to say. Right. Right. Yeah, you're bringing in your own little [UH] what's the word biases, I guess. I don't know but yeah, you can't please everyone and pretend you can is its own kind of dysfunction. Right. So the question good isn't whether tradeoffs exist, it's whether they're made consciously. Right. And transparently. Right. steel man is right. OK, you can't, you can't please everyone. Right. Sure. make all the tradeoffs visible. I had a CEO one time that she kept asking the same question Cool. like you can any time that product would present like these are the paths we're going down. She would say you're not making visible and showing me all of the paths that we are not taking because we're doing these things that you're saying. Oh, you know, trust me, bro. I did the research and I did my own stunts and we're, we're just going to take this path. Yeah. She's like, see those things, I to you diagram of like need these are all things we could do, you we never did it [UH] so we never did it because of the amount of work it would have taken to map [UH] all those different things, And it was a failing on. It was a failing on the person that ran product at the time in that company that they never, never that concern seriously took and actually like internalize it and They then produced what the CEO was asking for. Yeah, because it came up, you know, again and again and again over the years. Yeah, it sounds like a short fall right there. Yeah, OK. Yeah. the product's outcome is just one factor in a web of competing desires. So it's the stakeholder analysis that I'm about. It's the talk about users. users, employees, the society at large, the populations of what you know, like Facebook. If we're thinking about Facebook, like who uses Facebook, like large swaths of the population. You know, you really got to do like audience analysis at scale talk about, you know, grandma's, you know, [UH] I got to think when they when they build the Facebook features, they do about these large profile segments that are using the I'm shit. to features? how is a not feel going 80 year old grandma going to use it? How is a, how is a 14 year old just coming think on a platform for the first time going to use it? How are, you know, I mean that kind of thing. they should be doing that Yeah, right. It wouldn't surprise me if they're no, they don't know. Yeah Oh, genocide. How about, how about or how about how we optimize for genocide in Atlassians data percent PMs of[UH] 80 don't involve engineers early. They're ignoring a key stakeholder outcome before the roadmap gets written. that assumes that you actually care about your internal employees. That's a big assumption right there. Oh yeah, I know [UH] that's, that's from the [UM] Atlassians[UH] blog. Their state of product 2026. That's [UH] that's what that's from. I find it interesting that we, won't we, be we, even we about how our own employees think about the features. You think that tells you something about the company's general willingness? Yeah. Yeah. Yeah. But also, you know, employees, they're not buying the product. They don't pay us anything just concerned paid to produce. Yeah. Sad Facebook champions of democracy. Anyway, I have to make you want to switch to drinking real alcohol instead of talk about takeaways. So before locking out any outcome goal, run a stakeholder outcome would do that. let's who map. What is a stakeholder outcome map, Brian? OK, it's you're going to list every stakeholder group affected by the goal don't if have you your stakeholder which groups, who they are to start. Yes, you're already behind. And I, you're already losing. And I've got a book to sell you. Sorry, it's a stakeholder outcome mapping by Brian and oh oh goodness. That'd be a short book. Hundred dollars. It's short but expensive. Yeah, we only have to sell like two books. That's it. We're just, we're getting Bartab money. That's exactly right. Yeah, maybe three books. for for So each group, you're[UM] going to write if we achieve our outcome, what does this, you know, fill in the blank stakeholder group gain or potentially lose. I'd be interested in hearing from people that are doing this at any level. I mean, yes, of course, people do persona maps and things like that. But, but this isn't talking about that. This is actually talking about something a little different. Right? What do they hope to gain? So add another column it to next and they gain something, what's the tradeoff say for them? Right? Again, for them, not for you. Right? For them. any stakeholder will lose entry. That's where you were just if said. Yeah, I was just going there. Any stakeholder will lose entry. They get a constraint added to the goal that would like look churn by three percent. Right? You know, people come into my forums or people come to my site and then they go to the, they go to the cancel page. And you want to, you know, you want to hit them on the last minute deal or something, what something. Right. or make them call your support to cancel. That's, that's a, that's a dark pattern. Yeah. And then they'll would the reduced just escalate over to the retention team That's exactly right. We all use software like this. Oh, it's hard to cancel. Yeah, that they're doing They're yeah, but stuff is bad for this. you, the user, but good for them, the company. They've made this decision intentionally. so in my example, the reduced churn by three percent is your goal. And you say if we do this, then people are going to like have to add more support tickets because you know, so what you want to say is like, well, who's affected? our support department's affected. grandmas in their 80s are affected or whatever. They're going to pick and phone try the up Yeah. When they can't figure out this new complicated form or whatever. So you want to say when you flag that this, this, there's a loss going in your lawss call your win and loss column. You want to flag to say, I want to add that constraint now to the goal. So it's not just reduced to call somebody. churn by three percent. It's reduced churn by three percent without increasing support ticket volume. Right? Because you can track what people call. And goal without Sure then you X, hey achieve this by know, this percent, Y, whatever, you know, your normal OKR Z, framework and then you tack onto it. But without and then you list your things that come from your stakeholder map. That's just one idea. Yeah, I mean that's an idea that could track it we came up with through you know because again, the surveys or whatever. Okay framework, I You know after the call we have a already hate it. survey those sort of things. because it doesn't So you're so you're consider the effect adding constraints. You're you're of those outcomes. basically adding That's like negative scenarios to your another reason like it. It doesn't OKR goal. Okay, our doesn't have don't guardrails, right? Has no guardrails. Yeah, yeah. it's not like it's like even you know, starting I genocides or whatever could that wasn't even considered by the people that came up with these things. Yeah. So your products outcome is just one small factor, right? The of competing stakeholder desires. web That's where the real complexity lives. Right? what do you think? Do you do you map stakeholder outcomes before setting goals or is all this new to you? I would if I think that's probably new to most people that would listen. I would think so. So let us know in the comments if it's new to you. Let us know in the comments if you have a method already I would love to hear about it. And if you're wondering how this connects to the bigger picture, stick because[UH] around that's going to be our final segment. talked about the case studies, the psychology, and we talked about a potential fix, something that could help anyway. There's one more thing to hit. And that's the thing that's going to tie all this together. So when everyone's chasing outcomes, truth becomes optional. So here is the dark part of the outcome trap. And it's the connects Facebook to thing that Wells Fargo, to big tobacco and all the rest of the [UH] potentially evil companies that are out there. Allegedly, When the one thing that matters is hitting the numbers, the truth becomes the first casualty. people say literally anything fabricate data very findings, cherry pick report metrics. And as long as it gets them to the outcome, I'm looking for as long as I get my bonus. Oh, that's what I'm saying. Listen, we're not people saying lie. We're just saying that people can occasionally quite often be economical with the truth. saying that it's bad to be economical are. I'm not even with the truth. I'm saying the system rewards you whether you're telling the truth or whether you're lying. So back to this back to this Betty paper because we talked about that earlier. We're going to talk about a little bit more you know, in this section[UH] about know, you about it makes it turns honesty into a liability. And the arguments in this one are going to be [UH] hey, hitting the numbers listen, transparency and accountability. They're the real solution for these things. Like Agiles will tell you these two things. So transparent, like accountability, meaning somebody, one single owner, right? old The one single owner. Sure. of a business Amazon, outcome and transparency. You know, all up and down the chain, you know, inside out the organization. And then they'll tell you if these things are not aligned, it's a culture problem. So like it's not a goal setting process, culture problem and good. As long as you have good leaders in your organization at all levels, they will catch the gaming in the system and they will put an end to it. feel gave just like TED a talk. Yeah, right. I Good leaders catch the gaming. I mean, just that five words or whatever it is. Good leaders. Do they really catch the gaming or in some cases several perhaps they initiate the gaming and people are just either tacitly or explicitly following those leaders. Yeah. Right. like them. like Mike [UH] whoever Be Mike is. The against points are so good in this one that I can't like. I want Be to throw. I had a couple of more that I stenciled down against points Brian you say that you don't like output based roadmaps because they're a bunch of made updates or whatever. And you know we're all just like struggling to hit the milestone of the you know the little the little diamond on the chart or whatever like we're all hit that and then we hit it. And it doesn't matter because there was a made up milestone in the first place. So was like it because made up maybe maybe the leader got their bonus because [UH] the team hit that milestone. I'm just saying we shipped what was on the plan because it was on the plan. Nobody ever knew any, any deeper than that. You know, it was like, yeah, people, people are, people have been playing these games for a long time and like get over yourself. You know, people play games. It's a people, right? It's a, it's a one off people problem, Brian. That's, that's the argument. Yeah, yeah, yeah, yeah. to, welcome to my TED talk. Oh, welcome to my Thanks for coming anything even if so we shouldn't do even if the system in place creates more dishonesty, we shouldn't do it [UH] That's what we're saying. easiest facade to hide behind here. You know, you know, the other one, the only other one that's really great is the, the, the credential parade. Oh, the credential parade that's happening here as well. They'll care. This is another thing that bugs me. That's why know. I waswas didn't you morphing this podcast into the Brian hates OKR [UH] Yes. No, we kind of figured that from the beginning, the credential parade to say you should be using. Outcome based OKRs And you don't need this extra column in this stakeholder thing. And you'd like who are you to try to improve it because Google used these as is and they did a great job and they didn't start any genocides arguably that we know proponents of OKRs of Look,[UM] Yeah. are very, you know, driven. Here's how you do OKRs. You know, one outcome, one result or whatever, one measurement or all of those things. the credential parade though goes a bit beyond just OKRs. we've seen this many times, haven't we? Yeah. Right. Well, that's, that's what the people will be doing. they will basically pitch it as OK. Ours are they're[UM] because Google use like a syllable them and because Google is successful. Therefore, I don't need to listen to anything you're saying. They must be great as in like. But where is my brain here is the opposite of likelike look, even Google uses terrible systems know, people forget Google are in a specific business space. You are not necessarily in that genocides know, you space. Yeah. So, so and different Also, you you know, this copy pasting yeah, treat trade with caution. You know what, I, you know what, I think, I think I changed my stance. Oh, it's just that I think I changed my stance. I think I agree with the, I agree with the steel man here is like[UH] I'm, I'm firmly on the camp of like Google used Awesome. it. Great. dishonesty lives in That every organization. just tells me that So you're saying you're saying you're saying that Google used it. Everybody should use it. You know, aren't they a stellar company to emulate? Right So OK, culture then outcome culture just adds a specific accelerant. It makes truth a competitive disadvantage within the system especially if you're striving to you greater get know share etcetera a etcetera that everybody wants to do. you've changed your stance but I would like to hear from people that have tried it tried the chase and not succeeded yeah, I mean the Facebook situation like Facebook just market like they they don't want to get their service blocked in the country. That's their OK are to be unblocked in every country. Right? And they're going to do literally at any cost. Literally anything cost. at any You that incentive structure makes honesty like irrelevant. it's it's incompatible with that goal. And you keep our service running in this country at any cost. here about it's just it. other one in a people problem. That one. And it came up real early in the podcast too. And I didn't say anything. I bit my tongue because I knew it was coming back. the system doesn't inherently evil anti-government sentiment people or people need with compromised morals or whatever you want to say. They don't need to find and hire those specific people the system just needs good people with bad incentives. And the system will do its work. Yeah, yeah, yeah. It Yep. And it will transmography. preserves our bottom line and as to it. That's not even a word. Maybe it is So why not? Right? people into what And it's more, it's more people to they push ads to. Yes, just matter of a time. that's, that is, It's a matter of time. that is the real OK. Exactly. Yeah, this All right. There is is often disguised, always get more I might add, as the people. Increase the target audience. culture, the Acme way, whatever your[UH][UH] terrible company name is. hate Right. This is our way. And they sell that because you and see, look at all I these people that have been quote just unquote successful in our culture. So come on board and you know, get assimilated, indoctrinated or whatever We'll teach you our way basically. Right[UM] So yeah, absolutely. So there's a fix to this. Y going like it in'all not are cybersecurity, to there's a concept of red teaming where you dedicate a team to do nothing other than try to penetrate systems, try to try to breach defenses, stuff like that. So it's to assign someone to play the role of the red team and say hey if people find themselves [UH] morally compromised and decide to gain these things, let me find ways to gain these things But[UH] that would require you to take somebody off of like, you know, working the assembly line for like two point two seconds. Yeah. Just to break down your bad ideas. And I feel like that's where this is going to break down. So the find a red team reviewer. Have that be their role again. It's not a job because this is something that like could could easily be perceived by people as being like[UH] problematic. You know, oh, they are always challenging our, you know, like the old like we got rid because QA of they kept this is more like the ethical hackers, right? For your own company, right? But but a role. But you know just a role. Not, not, not a permanent job. Sure. Yeah And then you know, if the red team, if the person doing red teaming can figure out how to game your stat in 10 minutes, obviously like it's, it's not good. Needs to be resigned. You need to think about, you need to go and think about your life is what I'm saying. Go sit in the corner, Mark writing bugs. So Zuckerberg. Think about your life and then[UH] you know, and celebrate when you[UM] find these problems and change them. You know, publicize that then celebrate you've changed them. You've got to instill the culture though to accept these problems in the first place and then have a way to remediate them. Then you can celebrate. What if we punish honestly? Do you think we can find our way into a Facebook if we do that? Yeah, that's fine. That's fine. Anyway, what do you think? Has your company ever celebrated a[UH] downfall the of a you knew was masking metric or goal that a real problem? Let us know in the comments [UH] anonymously is what I'm saying. Can you make an anonymous comments? I don't know if you can. And this episode, if it helped you [UM] think about your outcome based goals a little bit differently. That was that's the only thing that I had in mind when [UH] when we were talking about planning this agenda is, you know, maybe like a little, a little, little second look. You know, a few little tactics that you can maybe try to say, I know, don't someone. are we like hurting people with this goal? You know, our users basically because we talk a lot on the podcast too about like rep damage. When you put out features that don't work as like if you're not even considering the types of users that use your app when you're designing new features and deciding what, you know, what new things to bring out. Like, what does that say about you as a company? But Brian, I'm going to be in a different job in six months to a year. Listen, I know I get it. I it. you get How do sleep at night? That's what I'd like to. I get it. I get it. today we talked about why outcome based goals can become a permission slip for evil. Feel like I should, I should put evil in different color. And we talked about the blind spot in all the outcome based road maps. And we talked about a stakeholder outcome mapping that I feel you can you can go back and use that pretty immediately. The stakeholder mapping. You know, outcome whether it gains traction or not you know, I don't know I've not seen anybody use it. So interested. First of all, I'm interested if people are even know, I'd be very understanding their customers deep enough to do something like that. And that's a whole different problem. It is. It is. Yeah, it is. product outcome [UH] trap [UH] part one by the way so if you're if you're thinking OK, I'll one, just, I'll just, by I Sorry, I'll just add some the better guardrails way. need and and do this [UM] do select size. You to so that's a taught me Brian. Cool. But part two, we're going to challenge you a little. It's going to, it's going to complicate this because we're going to cover in part damage two why the based goals is often invisible until it's too late. And why organizations from the outcome systematically destroy the people that try to warn them. Boy, if if there is something, if there's something that organizations hate cannot stand. It is a whistleblower. And Deming already figured all this stuff out by the way. Years ago, decades ago. Yeah, he figured all this stuff out and tried to try to warn everyone. And then as we know with the Deming work, everyone you know, tech CEOs all across the world, said who. Right. in certain OKR parts of the world. Yes, that is true. kumba Noir. That's what I'm saying. Oh, all right. So that's wrap. I a guess let us know if you found this useful [UH] if you're going to use the [UH] the tool we suggested. Yeah. Stakeholder mapping tool. And [UH] let's also know about other topics that you're interested in. Also, when you're doing that, like and subscribe.