B (2:40)

I'm realizing that there's a hilarious and inaccurate connection between the topic of today's episode and the fact that I no longer work at the ready.

A (2:49)

We're here to talk about Sam's performance.

B (2:52)

Uh, no, no, we're not unrelated, at least consciously topics, but, well, we'll dig into it. So Rodney, what, what are we talking about today? What is the pattern that we're actually digging into?

A (3:05)

Yeah, so it's that time of year, folks, where my friends in trad companies are going to dark caves and coffee shops to spend their on average 210 hours over the course of the year writing and delivering performance reviews. And even though this topic feels. Well, people really want to talk to me about it. They really want to talk to me at the networking about the performance management thing and why it's so borked. So here's the pattern. We all know it's broken. HR people, managers, leaders, work design, we all know is broken. And yet the interventions to improve it mostly make it worse. So we are going to talk today about what's wrong, what we would design if we were starting from scratch, and what role, if any, AI plays in the whole mix.

B (4:00)

Awesome.

A (4:01)

Sam, I think we should start with what's wrong with performance management. Why don't we just do it the way that we should do it? And also like when we talk about performance management, what are you talking about?

B (4:11)

So the version of performance management that I think we're kind of poking at here, we're somewhat creating a worst possible version to talk about, but we'll see if this worst possible version actually is. Is accurate to your organization. So generally once a year, maybe if you're in a particularly progressive organization, you do it twice a year.

A (4:33)

Yeah.

B (4:33)

And usually involves some sort of like self reflection or self evaluation that you have to do and send it to your manager. And then your manager goes through all of those and thinks about their perception of how you have done over the last period of time. And they have to write some sort of justification. They may have to rank you against your colleagues, against your teammates. If it's particularly Microsoft in the 90s, you're going to be yanked based on that. If you're at the bottom of it, managers probably have some sort of guidance from above about how many people can get what rating. And it's just A very convoluted and time consuming process where nobody generally feels particularly good about it at the end. Oh. Also usually connected to promotions and compensation. So even though we are talking about this as, you know, managing your performance, like getting better, what we're really talking about is like, how much money are you making and do you make more money next year?

A (5:34)

Yeah.

B (5:34)

What did I miss? What are the other bad parts of it?

A (5:38)

I put goal set. I put annual goal setting into the performance management basket so that at the end of the year we have something against which to check boxes.

B (5:50)

That's that real fun feeling of when you go to look back on the goals that you set a year ago and you're like, fuck. Like, all right. A. Either I didn't do these for good reason or I kind of forgot about them and there were still kind of good goals. Neither are great situations to find yourself in.

A (6:08)

Exactly. I mean, I think that you've already hit on what to me is a profound problem with performance management, which is it's trying to do, I'm going to say four jobs. And so it doesn't do any of them. It's trying to develop employees. It's trying to evaluate employees against expectations of their current role. It's trying to determine whether they should go into their next role and it's trying to justify or nudge a compensation decision.

B (6:45)

Yeah, yeah.

A (6:46)

You're never going to create a process that does all four of those things at the same time. It's not going to happen.

B (6:51)

Yeah, yeah.

A (6:53)

So, like, part of the problem is like that those things are all smooshed together and most people that I talk to are like, they cannot be unbundled. And I just feel like surely that's not true.

B (7:06)

And that's not even touching on the fact that all of those jobs to be done often poke aspects of our psychology that the systems that we generally have aren't designed to support. So the bias that we bring to evaluating folks, both recency bias and all the other flavors of bias that we have, just all of the challenges around, like having difficult conversations, psychological safety, like there's so much psychologically wrapped up in those for at least four jobs. And generally this system kind of just pretends that we're either really good at those things or those problems, those challenges that we have psychologically as humans don't exist.

A (7:50)

Yeah. Like those are, those are not important factors. And we should talk about the psychology of all of this because it's very profound. I think one of the reasons that it is so, so screwed up Is that. And, and you and I had an interesting conversation with a client about this once. Let's just take straight up performance management. Like, let's just take the annual review process. We'll just take promotions and comp and goal sett and we'll just take the rest of it out. Let's just talk about annual reviews for right now. It is unclear who they are for.

B (8:21)

Right.

A (8:22)

And as a result of that, they're designed as a compromise for everyone. So the structure of them really ultimately is for hr. Like ultimately, HR is deploying this process so that HR can say, we have a process for documenting performance so that later we can fire or promote or be deposed, come what may. That's just true. But then we pretend that no, it's really about the conversation. Like it's really about enabling and tooling the managers to have these very important and very robust conversations with employees. No, it's not. It is absolutely not. And then out of the third side of our mouths we go, it's really for employees. It's really for employees to get this very enlivening feedback so that they can orient to their growth edges and also feel celebrated and blah, blah. It's not that either. And so part of the problem is actually just the dissonance.

B (9:32)

Yeah.

A (9:32)

If everybody went into annual reviews just going, this is a box checking process for hr. We should get through it as quickly and painlessly as possible. We have to hit this rationalized bell curve in order to hit submit. Off you go. I would argue that's easier and better than the absolute pile of garbage that we layer on top of that truth.

B (9:54)

Yeah. Because at least it wouldn't take as long if we were.

A (9:57)

It wouldn't take as long and it would be honest.

B (9:59)

But the problem, I think I agree with you. The primary purpose in most cases is that paper trail so that we can fire you later. But in order for the information that we write down in that paper trail to have some semblance of validity, you need the, the facade of the other two to like prop it up.

A (10:20)

Right.

B (10:21)

And, and so you. So we all kind of have to like play the game. And the game takes forever and is demoralizing.

A (10:28)

Yeah. Like many more games, in fact.

B (10:32)

It takes forever.

A (10:33)

It's demoralizing. In fact, in one third of cases, the performance management process makes performance worse. Again, this is like a very like first principles conversation. If the first principle of performance management is mitigate legal risk, it is less likely to make performance worse. When we pretend that the first principle is to improve performance that's why it makes performance worse. Cuz people go well what the fuck was that? That doesn't help me at all. If what they knew is that their HR person is worried about the coming lawsuit when you get fired. And so some stuff has to happen that would just be an easier pill to swallow. Anyway, I digress. What else is wrong with this process besides that it's trying to do too many things and we lie about and are lied to about what it's trying to do?

B (11:18)

I mean I think the main thing that comes to mind for me is just how time consuming it is. Yeah, I don't know how many projects I've been a part of where it's not. We're not even there to do anything with performance management. We're just there at the time of year where it's happening. And you know, my collaborator on the client side is like oh sorry, I'm going to disappear for like two weeks because I'm about to get a bunch of nasty grams about how I'm late finishing all my evaluations. So I got to go pull some all nighters to go make those happen. And I think there's less and less of that now because that is the clearest use of AI in this moment. I've got some notes, I need to write some evaluations. AI, go do your. Go do your thing. And at least I've got a bunch of time back. Now I may not actually know what they say or I've really vetted them too super deeply, but at least I did the dance.

A (12:06)

I think that's right, Sam. And one of the reasons that it's so time consuming, I mean there are lots of reasons. The processes are usually bad and the tools are usually not great. But, but also what I see a lot, especially in big companies, they do this thing, HR does this thing where when I go to write myself evaluation there's like the description of the ratings or the numbers so that I'm supposed to fully understand that like what does meeting expectations mean? You know that stuff. And then it's like oh, and then there's like a competency framework for your level that you should reflect on. But then also you should reflect on the goals that you set at the beginning of this year. But then also there was some like company framework that was like business goals or finance, you should address those in your self evaluation as well. And then also at the team or functional level like we had okrs. So be sure that when you're evaluating yourself you speak to those by the time I have written a self evaluation that hits the points that I meant to hit by. It probably has very little to do with me. Like, I've now just.

B (13:13)

You've written some scripture that's fun.

A (13:15)

Like, I've taken an exam that's just like, how well can I string together words that address 17 different rubrics that I haven't thought about since January? And. Or maybe I'm seeing for the first time.

B (13:28)

This is so hard.

A (13:29)

And I know there are HR people listening right now that go, like, I would never. But like, I'm being hyperbolic, but only like 25 a little bit. You are doing a version of this in a lot of places.

B (13:39)

Yeah, no, totally. I was just gonna say this is sounding more and more like a board game designed to just like, make you frustrated. It's like, it's. It's like a psychological test, actually. Like, they give you a game that's unsolvable and like, how long would you bash your head against this in a room by yourself?

A (13:54)

Exactly. And like, I think everything that we're saying is valid, as I often do on our podcast. I think everything we say is valid and that's why we don't have any guests here to talk about this.

B (14:07)

But.

A (14:07)

But actually, even as I say that, I think that what people complain about, which is a lot of what you and I are complaining about right now, is actually the wrong stuff. What people should be complaining about is the psychological damage that is done through this poorly designed intervention and what it takes to overcome or like sort of piece relationships back together afterward. And A, and B, they should be complaining about how, especially as managers, they are accountable for driving their employees performance, but they are forced into a process that they have no control over and then often are also not resourced to do anything about. So it's like, you know, they have to go through this process and they have to force rank people and this person has to get a, not meeting expectations, and then that person cries and goes and gets drunk at the bar afterward because they weren't expecting to be the loser on the team. And then the next day. Now the manager's job is to like, fix it. Like to manage that person either out of this hole or out of the organization. And then that manager often isn't given like, resources or tools or money or capacity to do that. So it's like, if I were a trad manager right now, I mean, the time suck and the poor design is annoying and the compliance orientation is stupid, but I feel like what I would be more pissed off about is just basically being in an untenable situation where I'm meant to be doing this very uncomfortable work with very little authority or support to do anything about it.

B (15:59)

Yeah. Damn that. I agree. That is what we should be complaining about. That's. Yeah, that's good. Can I throw one more? That. That just inspired for me one more thing that we should be complaining about. Yeah. Which is like I'm gonna take like a slightly more people positive lens than we have been taking so far. We should be complaining about the missed opportunity for growth conversations. Yeah. And because I think people are hunger. Are hungry for.

A (16:23)

They are hunger getting useful.

B (16:25)

They are hunger. They are hunger for feedback.

A (16:27)

And.

B (16:27)

And people want to get better at the things that they do. They want to develop a sense of craft. And again, I'm talking about generally most people, not everyone, but most people want these things. And if this is our process where we are ostensibly doing that and you're getting a sense of where you are and how you can be better and it's designed for all of these other things and you're getting very little of that actual growth that you may be hungering for. Like that is sad and a missed opportunity.

A (16:59)

I agree with you. And this sort of leads me to where I wanted to go. Because what I wanted to ask you is if you were starting from scratch. It's day one. There is no performance management system. None of the things that we've talked about exist in the organization. What do you start with?

B (17:17)

I mean, I probably start with not calling it performance management.

A (17:20)

Uh huh. Cool.

B (17:21)

That name bums me out. I like what we have at the ready. It was summative feedback, which I think is descriptive and useful and good. I think there are probably other names out there that are also good, but the performance management, that management word and also performance. Both parts. Not a big fan of them. Especially when they go together like that. Other than the name though, I think I would get to your previous point about the four or more jobs that it generally is trying to do and get much clearer about what do we really want to design this for. The way we answer that question informs all the design that comes after.

A (18:01)

Yeah.

B (18:01)

What about you?

A (18:02)

I think that's right. Yes and yes. And the thing I think we should be designing for as the first principle is increasing performance.

B (18:12)

Yeah.

A (18:13)

First of all, part of why we need promotions and compensation is because people don't have enough developmental activity. So they need these externalities to know whether they're doing a good job. I think in a really robust feedback environment. Where people are being given opportunities to grow and stretch and then they're also finding out how that's going both from their perspective and in like a multidirectional way. There's less of a need for the explicit, like, now you're a senior, blah, blah, now you're extra. So much more. Like, people don't need that as much. I mean, I know everybody wants money, but, like, my aim has always been actually to just pay people enough so that we don't have to talk about money as the main thing. So if you take all of that out, if you just go, you know what, we're not going to worry about promotions right now. Because if people are doing a great job and they're growing, we want the kind of people who are not so fixated on titles anyway. And if we're going to just take comp out of the equation, to say comp is so fucked up anyway, this is just making it worse. We need a different way. Then I would design for increasing performance in individuals and teams. And if you start there, if you center the individual as the user, it completely changes the design of anything. Because now we have to put some control in the user's hands. No receiver of feedback on Earth ever in this life has been like, just put me in front of a firing squad, tell me every single thing you want to tell me. I will have no control over this process and no sort of handbrake on it, and I'll just be able to absorb and adopt everything that everyone says. If people tell you they can do that, they're lying to you and to themselves. We have to start from the user and say, what is a process that is designed to give people enough information, but also enough scaffolding and also enough safety that they can stretch? I would just start there.

B (20:21)

Yeah. Because starting there creates like 20 different paths to walk down or other things to figure out that you need to do in order to do that well. So I'll throw just one out there and we can probably hit a bunch more. If you're going to design a system that centers this on the individual or in the team and it's all about actually getting better performance, what you're not going to do is just have a big rigmarole once a year or even twice a year, you are going to be much more thoughtful about, like, what information does this person. Does this team need to get from their colleagues, from other teams that they collaborate with, maybe from clients. Clients or customers that will allow them to change what they are doing to Change behavior, to change their own internal os, to have a higher level of performance. We're not going to wait for one time a year to maybe get a boost in performance. Like that is. That is asinine.

A (21:19)

Right. And I think that we can't expect people to develop completely divorced or disaggregated from their context. So it's like if we're on some sort of rhythm, if we have some sort of cadence for feedback and for some kind of process, we also shouldn't expect that there's just like an up and to the right journey. Like some quarters we're dealing with like a family issue, or we're dealing with a death, or we're really fucking tired or like some, you know, some seasons we sprint and some seasons we retreat. And that's. That's nature. That is human nature. And no system that I've ever seen, including ours, I would say, frankly, really taps into the fact that work is alive and it grows and decays and sheds and grows like any other natural system. And we should have feedback that mirrors that, that says, you know what? For me, like, I've had quarters where in my summative, I've been like, I am not trying to do anything new. I am not trying to learn something. I am not trying to stretch. I am trying to maintain the work that I am holding and get a little bit of space and sanity back and figure out what is next for me. And that's it. So I just want your feedback on how that's gone.

B (22:49)

Yeah. And if what Rodney just said there sounds like a crazy thing coming out of your mouth in your organization, that's a sign that your current process is not designed, you know, with centering the. The individual. The fact that you could show up at the ready and have that sort of conversation, and we're all like, yep, that makes sense. Cool. Like, yeah, we're not. I'm not putting you at the bottom of my rank list. Like, that's, I think, is a good sign. The other thing that makes me think of is that because in traditional performance management systems, we do it this once or twice a year and there's a lot of process around it. The feedback is kind of all one size. It's just like. And it's generally quite big. And I think something that we've done really well at the ready. And I'm probably going to continue saying we for a while. I'm just going to roll.

A (23:37)

You're still. You're still part of the we, the royal we.

B (23:41)

Thank you. Thank you. There's different sizes of feedback that you want on different cadences. So the idea of ask cats anything I should start, stop or continue at this time, which I think. Did Alistair come up with that?

A (23:56)

I did.

B (23:57)

You came up with that?

A (23:57)

It's meant to be after this session because it was meant to be like a little hot wash after we facilitated something.

B (24:04)

So yeah, it's just this stupid little abbreviation which you throw into slack or you know, in a hot wash afterward, which is like really micro. Like give me some feedback on like how that thing just went. And that's one very bite sized version of feedback that is actually really useful if you care about getting better at something. Like, it's very present in that moment as compared to like the summative conversation that happens quarterly. Like, all right, here's what I've been trying to work on. Here's the things that I've been working on in kind of like a project level, a larger work stream sort of level, help me see myself better in terms of how that work is going, which tends to be juicier, bigger, more ambitious sorts of things. And there's probably even levels above that that is only useful on a yearly sort of basis. And I think a system that allows you to have these feedback conversations at different levels on different cadences is just so much richer than having to do it once a year.

A (25:01)

And the other thing that happens sort of naturally when you have those different altitudes is like ascats will be very micro and very individualized based on an event or a moment or a small piece of work. But then also a lot of us have weekly hot washes on our projects where we'll talk about how we as a team did and maybe how individually we perform. And then a lot of us have end of project retrospectives where we talk about again, how did the project perform overall and. And like the point you're making that's really useful is like there is not one size and just like I said before, it can't do like all of the four jobs I named. Well, there's also not a single feedback process that can do all the kinds of feedback well because there's like feedback on the work and then there's feedback on how we teamed and then there's feedback on the outcome and then there's feedback on my particular demonstration of skills. There's all different flavors of feedback and we probably shouldn't try to do them all at one time.

B (26:08)

Totally something connected to that that I think is kind of also a prerequisite for the type of feedback systems that we're talking about is work that we did at the ready a couple of years ago around like a definition or lots of definitions of what good even looks like, what expectations we have for each other. What are the expectations that you have as a member of this system and in a specific role? And that is really valuable work even outside of the kind of lens of a performance management system. But if you want to have a system that is about helping people get better, you have to bring it into an environment where we've already started to have conversations and codify what good and bad looks like for us in our context.

A (26:55)

Right? I think that's exactly right. And what Sam is not talking about is like a competency matrix. What we are talking about is like, what can you do? Like, if you are at this level, like, like, what are you expected to be able to do? So like at my level of transformer, I am expected to be able to create a significant shift with the C team of an enterprise organization. Yeah. Someone that is three levels below me in that matrix is not expected to have that impact now.

B (27:37)

Right. So it's not like they have that ability and they're just rated poorly. It's like not even relevant to you.

A (27:43)

It's not even relevant. That's not even what we would expect someone at that level to be able to do. And so I think too often we get into sort of like the competency game or even the skill game that's like, you know, is an excellent salesperson, is a top tier salesperson, is a standard setting. And it's like, what does that mean? Does that mean that you can sell a $50,000 project two times a year? Does that mean that you can sell $1 million? Like, do the hard work of specificity, like every other fucking thing we talk about on this show. Do the hard work of thinking about specifically what you would expect someone to be able to do. And then that becomes the expectation of the job. And then it becomes pretty easy because if I'm not doing the thing that I'm expected to do. Well, why, why is that? Like, is it because we're prioritizing other things right now? Is it because actually I can't. Now we can have a more nuanced conversation about what's going on than just like, is Sam good in this role? Who knows? That's dumb.

B (28:50)

Also, it doesn't mean that those definitions need to be written in stone. In fact, yes, definitely should not be because your first crack at them are not going to be perfect. Because an outcome of somebody not meeting an expectation of. Of a level of a role that we thought they should be able to. Yeah, it may be that they're just, you know, shouldn't be in that role yet, or we did not define very well what is actually possible there and we should be open to changing that.

A (29:18)

Yeah, I think that's right. The one other thing I want to say about how it could be better from scratch, and I don't know exactly how to do this, but I'm just going to talk about something that I've seen a lot in my own. My own, like, real management of people and even. Even at the ready where we don't have trad management. Being a steward with people who are less experienced than I am. I do think a lot about the, like, stress and like, performance curve. And Jack can link to this in the show notes. But essentially there's this way of thinking about this that's like if you have performance on the Y axis and pressure on the X axis, there is a bell curve. And the middle of that bell curve is basically performing and optimal. If we have low pressure and low performance, what we see is like, boredom and like complacency and like comfort. But if we have too much pressure and too much expectation of performance, we have strain and burnout. And I think about this a lot because we have to give people who are seeking our feedback, again, feedback in the context of where they are. When someone says to me, tell me how I did with that executive team. If I gave them feedback based on the bar I hold for myself, that would tip them into strain and out of learning because they have 15 years less of experience and a thousand fewer reps than I do. So me holding my bar and being like, well, if I had been doing it, I would have said X, Y and Z is not useful because they don't have access to my experience. In that moment, I have to think about what the performance expectation for that person as they are is. And I think really weak feedback is, like, overly positive. I see this at the ready. Everybody loves to give everybody a gold star. And I think then you end up with comfort and entitlement. And I think that's just as bad actually as strain and burnout. But I've seen the opposite end too, which is like, here is this incredibly high bar that has no basis in reality for where this person is and what they're capable of right now. And that's. That's bad too. So I think about that in terms of how we orient around the individual because if I was going to build it from scratch, I would build it for individuals to get better. And I would think about all of these many factors for how one does that.

B (32:02)

Yeah, listening to you talk, it's making me think about how obviously you can't design this in isolation from the rest of your organization and how the rest of your OS really is contingent upon. Or they mean they influence both directions, obviously. So what I'm thinking about is you need to have an OS where there's a high level of trust for the type of thing that we're talking about to happen. Because ultimately, at the end of the day, if a piece of feedback is coming from somebody who you think is bad at their job and you don't like and you don't trust them, it's not going to land in a way that changes your performance for the better. But getting feedback from somebody who you trust, who you respect, who you know does good work, that hits differently. And there's a lot of other stuff outside of the performance management system that determines whether or not you have an OS that is characterized by a high level of trust and respect for your colleagues.

A (32:57)

Yeah. So one of the very hot topics Meta just published something about this yesterday is AI performance management.

B (33:07)

I was wrong.

A (33:09)

I'm sure. There was also an article about the Metaverse yesterday. So AI Sam AI and PM AIXPM collaboration. How and how not to do it?

B (33:24)

How and how not to do it? Well, I mean, I think it's important to differentiate between are we talking about how to apply AI to the gnarly bad system that we spent the first part of this episode talking about? I mean, there, it's like copy and paste some notes into ChatGPT and tell it to write a firm but respectful feedback for whoever this is based on. I mean, it probably saves you some time, like, great, go for it. But I don't think that is the most audacious or principles aligned way to be thinking about how to use AI to develop a better performance management system.

A (34:05)

So if like everything else we talked about in our AI episode, if you assume that AI inserted into existing performance management processes that are both not valuable and in some cases detrimental to performance is only going to make them worse, although maybe faster at being bad, which is worse.

B (34:30)

That's worse. Doing bad faster is just more bad, you can get more bad, you can.

A (34:36)

Do more quickly but poorly. So like, let's assume that in most cases, if you just insert AI into a version of what exists now so that as the manager, it's writing my Reviews and as the employee, it's writing my evaluations. And then probably there's also like an agent who is like comparing those and spitting out like, just like. Let's put that use case to the side for a moment.

B (35:02)

Sure.

A (35:02)

The automation of shit to make it faster. Let's just, let's put that away for a second.

B (35:08)

Good.

A (35:09)

And let's talk about if we were building from scratch.

B (35:14)

Yeah.

A (35:14)

What would we use?

B (35:17)

So I actually had a lot of fun thinking about this in preparation for the episode. And I thought a way to answer that question is to think about the feedback process that I'm most familiar with, which is the summative feedback at the ready. And like, how do I wish? Because we designed that before AI really became a thing. So at least as of my last day, AI was not really like baked into our summative feedback process. But I see how I would love to have AI in a future version of this. So I'm going to do like a quick walkthrough of like the big things that I. That I see here. Does that sound good?

A (35:50)

Yep.

B (35:50)

Okay, so it starts. So it basically happens quarterly. And the idea is that you ping who you want to be part of the conversation of, you know, people to give you feedback, ideally people you've been working with over the last quarter who know your work well. It may be your project team, it may be other folks you've been collaborating with on internal stuff, but you convene those folks. You then do a self reflection kind of answering the. There's some evaluative questions, but you write your own thing first. Then you send it to everybody else for them to read and reflect on. And then there's a conversation that happens every time I have sat down to write my self reflection. I look at a blank piece of computer paper and I say, what the hell happened in the last quarter? Yep. And if I had a better brain and a better memory, that would be an easy exercise for me to do. But because I was always doing lots of things and my brain is bad, it was always a real pain in the ass and I had to go like, dig through emails and dig through Slack messages and like, what was I working on? And figure that stuff out and then write my thing. That is a perfect place where my little AI buddy should be able to give me the highlights and help me remember and present things to me in that moment. Like, hey, here are the things that happened even with no evaluation tied to them. Just to like get it in your head right away. That would have been super Helpful in the conversation with my colleagues. I would love to have a little AI agent there who is representing a point of view that none of the people in that conversation necessarily have. So obviously everybody kind of has their, like your clients not there, clients not there because everybody at the ready is nice. A real kind of critical person is.

A (37:30)

Not necessarily no devil's avocado in the room as my exact.

B (37:34)

And it would be, it would be fun even if I could like kind of set the temperature of that. That agent. And they also can just see across things more than any one person can. So that agent should be able to see across the various projects. I was working on all of the Slack channels, the way I was showing up in all of the different things. And I would imagine there's some interesting reflections that it could put to me that I could decide whether or not it was useful or accurate after that conversation. Having a sense making partner where, you know, that call just ends. I have a bunch of notes from what everybody said to me and I'm like trying to work through it and like make sense of it. I could do that with an AI partner to like help me make sense of what I just heard. And then finally the whole point of all of this, I want to take different action in the future. I want to learn things, I want to develop new skills. I want to not be triggered in a certain situation the way I was the last time that situation came up. And there's like real boring versions of what an AI could do, which is like, here's some resources based on the feedback conversations that's. That'll say that saves me a Google search. I'll take that. And I think there's some really cool things in the future where the AI is aware of the context in which this feedback is most relevant. And it can sense like, okay, this Slack message is going to piss you off. I know it. I know you. I know the context. And remember in your feedback, like, you wanted to do this thing differently. So like, remember that. And I think that's an interesting future to like kind of have a little conscience on your shoulder, a little angel on your shoulder being like, hey, remember you wanted to be different in this sort of situation. And that I think is actually exciting.

A (39:18)

I mean, I would take like the last part even, even one step further. And it's like, you know, AI is so good at building routines and just being able to see, to like nudge you. You know, you could run a workflow that's like, I really struggle in this meeting every Friday and So before this meeting occurs on my calendar, I want you to check in with me in the following way and ask me to do this exercise where I like ground myself and prepare to go into this thing that is always problematic for me totally when.

B (39:53)

And I would say, you know, you could kind of do that without AI if you are self aware enough. Like, all right, I'm going to make a reminder that's an before this meeting. The really cool version of that would be like, you don't even necessarily realize it's that meeting that's AI is like this is the place where this is the place to practice this. So like, heads up, bro. Like pay attention. Like, oh, okay. Like I don't need to be the one to like putting those lines together.

A (40:17)

Yeah, I mean I, I remember, I think it was when I interviewed Josh Burson and AI was still, you know, relatively new. I mean it's still new, but in terms of like there being a lot, a lot of use cases and just experimentation that we were reading about. Yeah. And I remember my hypothesis at the time was this is going to be an incredible tool for feedback because it removes so much of what makes feedback hard. It removes a personal relationship, it removes the feeling of being judged, it removes the feeling of being misunderstood, but then also of not wanting to over explain because you feel defensive. It's like you have so much control in this small bubble. I mean, I do more self work in partnership with AI than in any other way and I do a fuckload of self work. But like, you know, before my summative, I spent a long time using a recipe in granola to understand how I am showing up in meetings. Like where am I taking too much airtime, where am I over functioning, where am I not contributing? Like really like looking at where I spend my time and how I show up in those spaces compared to what I say I want to do, compared to the role that I say I want to hold, compared to the feedback that I got last quarter, all of which I fed it. Where are the disconnects between what I say I'm trying to do and what I'm doing? No human being is ever going to give you the kind of picture of that without bias that AI is. And the great thing is if you don't agree, you can just fucking ignore it and it doesn't really matter. But the truth is that like I've never experimented in this way with like pretty deep self work with AI and not been like, oh fuck, it's right. That's right. Like what it's telling me is true when I feed it my calendar and it tells me you are in these places that you should not be because you said you're not going to do this anymore. It's right. It's just like it is a mirror in a way that no one who wants to maintain a relationship with me is going to be a mirror.

B (42:43)

Yeah. This is a place where I've already seen myself like change my mind. Because I think early days when I was first starting to experiment with AI, I was like, you stupid robot. You can't understand my human. You can't give me feedback. I want my feedback from a flesh and blood human. Kind of forgetting the fact that I've gotten a lot of really shitty feedback from people who were really mailing it in and like, not like didn't really know what was going on at all.

A (43:09)

Yeah.

B (43:09)

And yeah, and now I have had the exact same experience that you have had. And it's not to say that I'm mindlessly accepting every single thing a chat bot throws at me, but if you are, if you're seeing something that is giving you and you have to think about it, that changes you interacting with that feedback. Sometimes you decide, yeah, I don't think it actually is, is right, but sometimes it hits it just right and you're like, ooh. When you need to hear something, even if you aren't really sure what you need to hear, often when you see it or read it or hear it, like you can tell, like it hits a thing that is, you know, exactly. And I think I have found the tools that I have talked to. It sometimes hits stuff that is useful to hear and then it's up to me to decide what I want to.

A (43:52)

Do with it 100%. And the other thing that I use AI for a lot in the spirit of getting feedback, which probably a lot of listeners are going to tune out right now, but I don't care. When we tend to do human to human feedback, first of all, we're basing it in language or writing. And second of all, we tend to be focused on behavior, on work product, on the experience that we are having of another person. When I am getting feedback from AI, often what I'm feeding it is like my own experience of something and the emotional texture of an interaction I had or the physical sensation that I am experiencing during a hard week. And being able to get feedback that feels so much more holistic, that's like, yeah, here's what happened, but also here's what it reminded me of from my past that I don't necessarily want to talk about in my summative feedback. And here's how it made like me feel in my shoulders or in my gut that again, I'm not necessarily going to like, ask my colleagues to like hold and process for me. I've learned so much about myself and about how I show up and the difference between my intentions and perception. And like, I've learned so much because I can openly and freely and in a very unconstrained way share so much more information with this mirror than I would ever share with another person. And I think not doing that just feels like such a missed opportunity.

B (45:38)

Yeah, no, I totally, totally agree. Is there anything else kind of. I know I kind of used the ready summative feedback process as the scaffolding for this, but anything like, putting that aside, other things that you think will be interesting for folks to experiment with with AI and performance management or things you see maybe coming around the corner?

A (46:02)

The one thing that we didn't talk about that I think is almost certainly coming is part of why performance management processes get so reduced down to this manager employee one to one thing is just because it is less complicated than multidirectional feedback. And I even hear about places where there is peer or360 content generated, but then the manager filters it and knits it into a narrative and delivers it. Which to me is the terrible idea that no one should do. But I feel like the manager manager employee thing is so, so profoundly jacked for so many reasons, not least of which is because there are very few roles in any organization where we can produce anything alone.

B (46:53)

Yeah.

A (46:53)

And I think the opportunity that we're not using AI for is how are we performing as a team and what is the contribution level of members and what type of contribution are they making? And who is doing too much, who is kind of kicking back and who is carrying the emotional labor that is unseen and who is feeling a different level of responsibility than other people and who is on the verge of burning out. Like these are things that AI will be able to do with and for our closest working groups that no traditional performance system is ever going to do.

B (47:34)

Yeah, it's too invisible to the way we currently do it. And I think that opens up a lot of doors for folks who maybe feel like the contributions they make, especially in a teaming environment, aren't acknowledged by the way we currently do things. But you have, you know, I think I could totally see a future where, you know, the fact that you contribute to every Google Doc and you're answering questions in Slack and you're just like these, like the quality of the interactions that you have with folks, which I think can be hard to quantify just without any sort of tool. I think AI today could probably start to like, point at interesting things there and surface the types of folks that you want on a team that maybe are quieter and don't shine in a traditional performance management system.

A (48:20)

And to me it all boils down to the fact that AI can be used in the flow of work. It can be inside of Slack or inside of meetings or inside of documents. And these other abstractions force us to look at the body of work and then make conclusions to write down somewhere else. And so that's just inherently really limiting. And so I'm actually pretty excited to see how AI in the flow of creating value in an organization can start to really shift this.

B (49:02)

Cool. I love that. My last kind of, you know, sort of sci fi moment that I'm going to be looking for in this realm is that I think in most cases getting a bad piece of feedback or getting feedback that you're not doing well in a role is kind of can be a bit of a death sentence in an organization because you're there to do a specific role. And if you're not doing the role or some main part of the role, then you're kind of at risk. And in really large organizations with tens of thousands or hundreds of thousands of people, I find it really hard to believe that if you're not thriving in this role, that means you need to get out of this organization. I have a hypothesis, a hunch that there may be other things that you could be contributing to this organization that is actually very needed by the organization that is not knowable kind of at the human level. And, and if you have an AI that is really aware of needs in the, in the organization and what you're actually good at and you know, some ideas that we talked about around, you know, the fact that we all have a portfolio of roles, not just monolithic job descriptions, suddenly you've got a really interesting opportunity for some mixing and matching to put people in context where they will thrive.

A (50:19)

Yeah.

B (50:20)

And that, that level of kind of degrees of freedom to like what we do with the feedback that we get, I think could be really cool.

A (50:26)

I agree. That's awesome. Do we do it, Sam?

B (50:32)

Of course we did it.

A (50:33)

I think we did.

B (50:34)

Five out of five meets expectations.

A (50:37)

I think we significantly exceeded expectations.

B (50:41)

I am giving myself a bonus.

A (50:43)

Nice.

B (50:44)

All right, so we're always looking for new topics for the show. And if you have an organizational pattern that you're having trouble changing, shoot us a note@podcasttheready.com this show is engineered by.

A (50:56)

Taylor Marvin and produced by our recently returned from vacation friend Jack Van Amber, who we miss very much at work with. The Ready is created by the ready where we help organizations around the world change the way they work. Thank you so much for listening.