Slik kan KI hjelpe prosjektledere med å gi tilbakemeldinger

Read The Transcript:

We’re trying out transcribing our podcasts using a software program. Please forgive any typos as the bot isn’t correct 100% of the time.

Galen Low: Hey folks, thanks for tuning in! My name is Galen Low with the Digital Project Manager. We are a community of digital professionals on a mission to help each other get skilled, get confident, and get connected so that we can amplify the value of project management in a digital world. If you wanna hear more about that, head on over to thedigitalprojectmanager.com/membership.

Okay, today we are talking about delivering feedback as a project manager, and whether Generative AI can make this process any more impactful, or at least just a bit less painful. Joining me today is Projjal Ghatak, CEO and Founder of OnLoop, a Gen AI company focused on making managers and teams as effective as possible in a hybrid working world.

Projjal, thanks for joining me today.

Projjal Ghatak: Thanks Galen, thank you for having me.

Galen Low: I was really interested when I heard about just your company and your product overall. It's something we've been talking about in the community, just like how to drive performance in a project team, how to deliver feedback well, how to get people just performing, but also, just like feeling good about the work that they're doing. So it's like super timely. I'm super glad to be diving in.

I thought maybe before we get into the thick of it, can we just talk a bit about your background? I found it really interesting, and I know there's a lot of different facets. Could you just tell our listeners a bit about your journey leading up to when you founded OnLoop in 2020?

Projjal Ghatak: Sure. So I started my career in management consulting in Southeast Asia, actually. Accenture Strategy had just built out a new practice in Southeast Asia, so I joined them right after undergrad and spend the first three, four years of my life in consulting project teams. And I'm sure we'll go back to that as we talk about sort of food feedback in a project management capacity.

I then went to Stanford business school for my MBA and in many ways, did not want to go down a traditional career after business school. So I really went to business school to figure out who I was and what I wanted to do and Stanford as a business school spends a lot of time on helping you understand who you are as a person and who you are as a leader to help you really discover what makes you tick and what you really want to work on as an individual.

And that's different for every person. I spent my time at business school with a couple of early stage companies in Latin America. So I was in Guatemala and Miami and an early stage tech company in Columbia with an early stage retail company. And I think being close to the action and being on the ground, I think reinforced that being an operator was the direction I wanted to take.

And post business school, I ended up actually joining a family office owned by an Indian family that had a bunch of metals and mining assets around the world. And I was responsible for the turnaround of a lot of the North American assets, so I spent a lot of time in Sault Ste. Marie, Ontario, which you might be familiar with, and you know, coal mines in Kentucky and West Virginia.

Both sort of run financial turnaround as well as strategic turnaround of old school businesses. But, I think working in a family business and metals and mining wasn't my life's calling forever. And so in 2016, I actually moved back to Singapore and decided to go back to tech. And I dabbled with the idea of starting a company at the time, I actually had an idea around a recruiting product. And I think the thread around people and talent has been a thread my full career from a business lens, but realizing that sort of ultimately businesses are about people and the way you achieve outcomes is through those people.

And so we end up spending a lot of time and attention on accounting software and CRMs, but throughout my career, and I joined Uber when I came back to Singapore and spent three and a half years at Uber in a variety of roles. When I joined the company, we were 9,000 people. When I left, we were 27,000 people.

And I've done sort of country, regional, and global leadership roles. And so, a whole lot about org and what teams need to look like for high performance. And I felt like I had to do everything on my own without a ton of support from quality products and quality approaches. And HR just felt very behind vis a vis other aspects of work.

And we had actually seen the software evolution sort of create Zoom, Trello, Asana, Notion, products like Riverside for podcasters. I don't think someone would come up with Riverside if the people they interviewed were the IT department of podcast hosts versus podcast hosts themselves. And frankly, everything around talent was somehow built for a functional organization and not for the end user.

And I hated performance management with a passion. Every L&D tool we were given was hardly ever used by people, and so it all felt broken in many ways. And so I was quite naive going into it, feeling like nobody had thought about solving these things, therefore they hadn't been solved, but that's clearly not the case.

There was a few structural challenges to work through when you bring something new into the world, but, my time at Uber taught me to be a fighter and sort of realize that if you want to bring something powerful to the world, sometimes you have to go against the grain. You eventually fall with the grain once you get scale, but for a while, you might have to go against it.

And so six months after leaving Uber, I founded OnLoop and we're three and a half years in and still loving it, despite all the ups and downs of being an entrepreneur and expect to keep doing this for a long time.

Galen Low: I love that mix. And honestly, I think what you said about business school, I don't think a lot of people who haven't gone, well, a) to Stanford Business School, but business school in general, I don't think they realize how much it is about being a leader and not just like the inner workings of a business.

You know, you and I are both Accenture alumni. So I know that kind of, builds a certain muscle there that would do well at Uber where you're like, okay, maybe I don't have that much support. I'm kind of on my own. I might need to push against the grain, but I have the tool set to succeed. And then yeah, see what's broken and start something new.

Projjal Ghatak: Totally. You know, on my third day at Accenture, as a 21 year old analyst, the manager quit and I was told, you now manage this project. That would have not happened at McKinsey Bay and BCG, and maybe for good reason, but yeah, that teaches you a certain skill set that just makes you better, and I'm very grateful for it.

Galen Low: I wonder if you could tell our listeners a bit about OnLoop. We were chatting about it in the green room. I'm super interested in just what made you start it when you started it. It's 2020, you've come out of your tenure at Uber, you see things that are wrong and you start OnLoop. Tell us how you help folks and where you want it to go.

Projjal Ghatak: Yeah, so we fluked it with timing. It wasn't that the pandemic inspired us to start. I'd written down exactly all the problem statements I wanted to start back in November, mid 2019, right after I'd left Uber. But the pandemic sort of really threw the working world into a flux. And being able to manage a team became significantly harder through hybrid and remote.

We think those challenges are very much still a challenge and a big reason why a lot of CEOs want people back in the office because the only way people know how to manage is through in-person, informal ad hoc feedback loops and not be intentional in the way you need to in a hybrid setting.

And so for us, we started with, hey, performance management and feedback seems to be broken. Let's go try and fix that. And we were lucky that we were able to raise $2 million right off the bat to work on a product that solved that problem. But as we stayed problem-focused, we realized that the problem was bigger.

And feedback, for feedback's sake, becomes a nice-to-have, but doesn't become a must-have. Like, everyone's known feedback matters for a very long time, but hasn't really made a dent in the universe, and you've got to ask yourself why. And that's really, if it's not in the context of overall performance, and that has goals and people's overall energy as layers that sit below feedback and then ongoing learning as a layer that sits above feedback.

We realize that a manager has to manage that entire vertical stack and that sort of give birth to our collaborative team development or CTD framework, which we sort of you know, more and more compared to Agile or OKRs or an approach to how a manager should be running the team. And then we have a software product that accompanies that makes it much easier to build the right habits around goals and feedback.

And that's where Gen AI is very powerful in taking people's raw unstructured observations, converting that to structured feedback and extracting behaviors, values, competencies from that. So I feel like we got very, very lucky because we started working with Gen AI back in 2021. So we were private beta of GPT-3 back when GPT-3 wasn't all that cool and you had to apply to use their software.

And, that sort of double turbine off a new technology that allows us to solve an age old problem in a brand new way combined that hybrid work made a manager that much harder was really a confluence that really helped us in bringing OnLoop to market.

Galen Low: I love that, and I love everything you explained because I think you're exactly at the nail on the head. Feedback, kinda like managing people, is one of those things that, I don't know, it's just a word we throw around and we expect people to be able to just figure it out, right? Oh, give feedback, right? But, you're like, okay, actually, there's a framework, there are different layers, there's a stack that we need to, pay attention to, and fundamentally, we are groups of humans, doing high pressure work together.

We've known feedback is important. We do have unstructured conversations. Everyone kinda sucks at feedback. But it's still a core piece of the puzzle, and I love that that's kind of like what you've built the tool around.

Projjal Ghatak: One analogy I'll use is that, you know, if the only way to work out was a moldy gym, everybody would hate exercising too, right?

So if we've made a practice awful, people are going to run away from it. And the same thing applies to planning, and OKRs, and therefore how people think about goals, right? And feedback and goals are the two most important components of high performance, in the same way exercise and diet is. But we've fed people, processed fat food, and giving them dirty gyms, they're not going to work out.

Right? So I don't blame the people for hating feedback or finding it awkward because I just don't think we've empowered people the right way. And billions of dollars of training has been completely ineffective in teaching people the right things. And therefore, yes, some level of knowledge is important, but enabling people in the day to day practice of it becomes even more important.

Galen Low: I think that's fair. I wanted to dig into it a little bit. Maybe this is its own podcast on its own, but this notion that we kind of aren't good at giving feedback. We know it's important, right? I like the fitness analogy. In my head, I've got, like, this recycling analogy. Remember, like, when recycling was really hard, we knew we needed to do it to save the Earth, and yet we weren't doing it because it was difficult, you know?

And, like, okay, well, like whatever, bring your batteries to this depot and, use this bag for this thing.

Projjal Ghatak: Actually, I have a different perspective on it.

Galen Low: Okay, tell me, tell me.

Projjal Ghatak: People don't do it because you don't see the impact. So if everybody lost a kilogram after every run they made, they would be running every day.

But people don't see micro progress. And when you don't see micro progress, you feel like it doesn't matter. But if I give every household a target, and every time they put a can in recycling, that number popped up by one, and they could see what it could get to, I guarantee you that will change how much people recycle.

But when there's a drop in the ocean, people don't understand the impact, and therefore they're like Meh, it won't make a difference anyway, why do it?

Galen Low: Humans are bad at seeing gradual change. It's one of those things, one of those things. Listen, obviously feedback, it's a great thing to create that loop to get that personal growth, professional growth, team growth, and then overall that sort of business growth. But I thought maybe we could take this for a spin and put this into a project management lens.

Because the thing about being a project manager is that we're often responsible for giving feedback to our teams, but rarely are we their direct manager. And that can make things super awkward. And I do think we're people leaders, especially in a productized organization where most of it is projects.

We are seeing the work happen right in front of our faces, are well positioned to deliver feedback and yet, it's really easy to get the, Dude, you're not my boss, kind of response to delivering feedback. So I was just thinking about OnLoop, and I was thinking of the tool. I'm like, How can project managers use Generative AI to help give and track feedback in a project context?

Because there is this sort of performance management framework. There's been some really good tools coming out in that market that are less stiff, right? Less procedural. But I haven't really seen anything that specifically talks about, Okay, what if you're just managing by influence? You're a project manager, you are not that person's manager.

You know, the like, HRIS isn't gonna ping you at the end of the year and be like, Please, tell us whether this person is good or not. You're just working with them every day. Yeah let's go there. Can a tool like OnLoop help a project manager?

Projjal Ghatak: So I think there are a few layers to unpack there, right?

And you hit the nail on one, which is around managing by influence versus managing by control. And I think even if someone is someone's boss, at the end of the day, there's only that much you can manage by control and so managing by influence becomes the way to go.

I want to bust a few myths because feedback in itself is a very loaded word. And actually, nobody ever said that feedback means constructive feedback versus reinforcement feedback. And nobody also ever said feedback has to be about someone else versus about yourself. But when we think about the word feedback, we automatically assume it's typically constructive feedback and it is about someone else.

But there are actually four types of feedback, self feedback and others feedback, and then within self and others, what is celebrate or improve feedback. I think that's point number one. And for people to develop successfully, they need all four. And often as a person, if you want to deliver feedback to someone, you will be a lot more effective in putting yourself in a Socratic coaching approach and asking questions versus delivering an insight.

So if someone had a really bad presentation, going to that person and saying you had a really bad presentation is not going to improve behavior and make the person better. But going back and say, Hey, Michael, how do you think that went? And seeing sort of Michael's self assessment and then breaking it down.

Okay, so what do you think went really well? And what do you think could have got better? And it takes a little bit more patience. But what it does is more likely or not the person themselves would have realized this. Right? And in that case, when they don't, it might be fair to say, Hey, Michael, I don't think this was your intent, or you may not have realized this, but when you keep using filler words in conversations, I can see that people start getting distracted, and that can make you less impactful in presentations.

And that's called the situation behavior impact model of giving feedback whereby I'm being very specific on when the person did something was a saying, that was a shit presentation, because that's useless feedback. Nobody can do anything with that, and nothing is 100% shit or 100% great.

And it's not binary either. But, there's a woman named Kim Scott, who I've met, who wrote this book called Radical Candor, which has become the Bible on feedback. In many ways, it's an amazing read, if anybody wants to learn or understand feedback. But she talks a lot about challenging directly, but also caring deeply and being kind.

And good feedback is both direct and kind, and often people remember the direct part of it, but forget the kind part of it, and then that's what you call obnoxious aggression. And in many ways, people have weaponized radical candor as a term to be obnoxiously aggressive, and generally people know when you care and when you don't.

Now, the other distinction I'll make is between what is task feedback around, Hey, this document you wrote, like we could have form examples. Which is not something people should think a whole lot about. And that should be happening all the time. Doesn't even have to be recorded. Doesn't have to be the OnLoop app.

There's micro feedback that should just happen in the day to day. But I think behavioral feedback goes back to thinking about an observation about a situation, and making it easy to jot those down. And in our product, we make that a voice capture that someone can just record an observation and then we'll restructure that as feedback for you and show you how to write that feedback out.

But the key is in having the right frequency of observation so that if you're having a feedback conversation, you're actually not nitpicking on one thing that you remember, but you're actually bringing up things that are more patterned. But, I've spent a lot of time at Accenture as a consulting team lead and had people who I wasn't the career counselor, but I was responsible for their delivery on that particular project. At Uber, I was brought in, in Singapore and told that I had to control all the regional budget. Whereas before I came in, the budgets were controlled by the country GMs and they didn't report to me and were actually more senior to me.

But they were very willing, once they got to know me and how I operated, to be like, Yeah, we're actually better off managing this budget regionally versus keeping control. I think people generally have to believe that you have good intent and that b) you have value to add. And typically, in my experience, that usually has no correlation with reporting lines.

Because then people see it as guidance coming from the other person versus criticism, and they actually want more of it. And in fact, younger generations actually crave feedback and don't receive enough. And so it's interesting that people often feel like they don't want to give constructive feedback, but high performers really crave constructive feedback.

And you will often lose high performers if you don't give them constructive feedback, because that's the only way they know how to grow. And people actually love receiving improved feedback of their high performer. On the other end, people love giving people reinforcement or celebratory feedback. That's never a problem.

And often that might be a good place to start. So practice the situation be your impact model to show someone what they're doing really well, and that will then give you ways to then deliver constructive feedback following the same model. And actually telling someone to do a good job is equally useless as saying you're giving a shitty presentation because neither of them have specificity that can let the other person actually understand what they're doing well or not doing well.

Sorry, I rambled on for a while, but there's a lot of angles to it. And I think it's important to parse through and we can go in any direction that you think would be helpful for your audience.

Galen Low: Honestly, I think that is really useful because, when I think about your definitions, especially as a sort of task feedback versus behavioral feedback, I think instinctively project managers are like, oh, task feedback, right? Like, okay, about this moment in time, that presentation. And in some ways, that behavioral feedback, especially when you're talking about coming from a place of caring, it's a bit of a area of ambiguity.

It's like, should I care enough about this person on my project team? I'm not their manager, but you know, like I work with them close enough. I noticed patterns. I think I can make them even better. My project's going to benefit. They're going to benefit. I'm going to benefit. The organizations involved are going to benefit. Like, should this be my problem? And I think a lot of project managers get to that point because we do care, we're just not sure if we're allowed to. Do you know what I mean?

And then that's what I love about what you said is that like it doesn't usually matter what the reporting lines are in a world of constructive feedback, in a world of reinforcing feedback. That isn't actually what matters and I think that's where we get stuck as project managers is we're like, Oh, no, the line doesn't go here. There's a dotted line or whatever. And then we resist it, right? We don't deliver the feedback. We work with teams that are, maybe could be performing a little better.

Maybe we see a lot of potential in these humans that we're collaborating with. But we're just going to be hands off because, I don't know. We're not their career counselor, right? And I think that does damage because in a way just to underscore your point, it shouldn't have to matter what the reporting lines are.

It's impact, it's value, it's them being able to see that change, right? Like you're recycling thing. It's like, okay, cool. I'm getting feedback. I'm craving feedback because I want to do really well. I know I'm not perfect. And so, give it to me, but also give it to me in a way that is useful. Hey, usable, not like, good job pat on the back and not, bad job, stick on the back. Any kind of thing, like, it's more than that.

Projjal Ghatak: Two other things that sprung up for me as you were talking, because we've run into these situations too. Now, there are certain situations where legally it is an issue if you give it in writing.

So we have a situation where we were working with an agency and one of the big tech companies is their customer and the only customer. So if you work in this agency, you're basically spending all your time embedded in teams at this large tech company. And so obviously they don't receive feedback, right?

And so there's no way you can also know how to promote these people because that loop is entirely outside the organization. It's not even in the organization, but because there is a independent contractor relationship there, the big tech company does not allow the person to give any feedback in writing.

Because then that can break the sort of wall between being an employee versus being a contractor. And in those cases, giving things in writing are a little bit more challenging, but that doesn't stop you from giving someone development feedback and having a conversation about it. I think the other thing I'll say, and this goes broadly for things in life, I think ultimately some people are playing finite games and some people are playing infinite games.

And you can think about that as short term, long term. You can think about that as transactional relationship. And I think the more you play a long term game in life, it becomes easier to care. Because if I treat you badly one day, I have no way to know where you're going to show up in my life again. And if I didn't give you good feedback, or I treated you badly, or I didn't focus on your growth, and conversely, if I actually really helped someone out, that's going to come back and pay you in very interesting ways down the line.

And I think when people are brought together in project settings, they might see that as a finite game, but it's like, well, this is temporary. I may not ever see this person ever again. And I think if you take that approach, it is that much harder to muster up the energy or the inspiration. You give feedback and develop and care, but I would urge people to think about things that are in finite games versus finite games.

Galen Low: I agree with that. And that's the other thing that humans are bad at, is like long term thinking, and also, I think, at the center of all of this the definition of a project is that it's temporary. So of course it's imbued with all of this, like, short term thinking, and yet, you just explained to me a, you know, a rather common, actually, like, contractual relationship.

Where it's okay, we have a vendor, we're bringing them in, they're gonna sit at our desks, they're part of our team. But there's this contractual relationship, we gotta do things differently, we can't write down the feedback. And then there's these questions of the people on the floor, they're like, Are they on our team, or are they just the contractors?

Do I give feedback, do I just shut up? Should I think long term? Because they've been here for a year, they're gonna be here for another five years. Project ends, and then another project starts, it's the same people maybe I should care. And I like that idea that maybe you should just care at the outset, because of the other thing you said, which is like, yeah, you could treat them badly, and you'll never know when they're gonna, appear back in your life, for better or for worse. There's a kindness to that. There's a caring and a compassion to work.

Projjal Ghatak: And I think it's high ROI, right? Like, the way I look at it is that the ability of impact I can have as an individual is minuscule. And if I want to make any large change happen in the world, that is purely going to happen by driving a coalition of people in that direction.

And good people have no door of opportunities. And if you have to get good people to give time and resources for your cause, they need to believe there's something in it for them. And, I'm lucky to now build an organization and work with people. And, we just recently brought on a consultant to help with marketing, was an old friend from Accenture.

And, we never thought we would ever work together again, but here we are. And, I think in my life, it's really paid a lot of dividend to care. And I think, I'm not the most soft, fuzzy person. And so I look at things more from a pretty hardcore impact perspective. But caring has hard ROI. It's not soft ROI.

Galen Low: Wow. Okay. It's also a strong message for the folks who look at some of this stuff. Feedback and performance management, they're like, ah, it's just fluffy. Show me the results. I'll look at the numbers. Then I'll tell you if you get promoted or not. But actually you know, it is all part of it.

Projjal Ghatak: Yeah. And project management is knowledge work and knowledge works hard to define in dollars and cents, and thinking about someone's capability behaviorally is the hard skills, right? Like it's like the soft skills are the hot skills. And so really being able to parse that out and really being able to articulate it well is I think hugely important, especially in things like product management or project management, where you can't touch and feel something and said, I delivered that. It is sort of a blue at a coordination and a multifaceted, knowledge worker role.

Galen Low: Absolutely. No, I love that. I wonder if we could circle back around to the Generative AI piece, because you said a couple of things in there and you piqued my interest of like how your tool shapes feedback. You were saying that in some cases, contractually you might not be able to like deliver formal feedback, maybe not written feedback.

But if I'm reading between the lines, there is still a use case there to have OnLoop help me craft feedback, even if it's informal, not written down, not going into a performance review. I wondered if you could step me through that. You mentioned like an informal voice note and just be like, Oh yeah, blah, blah, blah.

Here's my feedback about this person. And then having that structured, taking that unstructured idea, structuring it and helping deliver it. Can you step me through what that looks like?

Projjal Ghatak: You know, I've never done this on a podcast before, but I want to try and share screen.

Galen Low: Yeah, yeah, it'll, it'll work. We have a few people who will be listening just on audio. So I'll do descriptive.

Projjal Ghatak: Perfect. Okay. So I want to play this out, right?

"Today in the UB presentation, Alice Stewart, didn't do a very good job in managing the room. It was clear that the group was unengaged, but the energy level was super low. There was quite a bit of prep and so, the materials were correct and we had done a lot of work, but in the room, we really didn't land it really well. So it sort of came out flat."

And that's the kind of feedback people usually have, right? So they were muddled up observation of what happened. And frankly, if you go deliver this to a person, it's unlikely the person's going to do something about it. But that's how you feel and how people word it. So what technology is able to do, and obviously, we are tailor made around how to think about feedback. You can use, ChatGPT and try and do this by prompting yourselves, we just make it seamless.

So what this does now is that this splits this up into two very specific pieces of what we call captures, right? And the first one's a celebrate capture where it says, Alice, your presentation for the UB presentation was commendable. The materials were correct and well prepared.

This shows your dedication, attention to detail, which are crucial for successful presentations. And then it tags attention to detail as the sort of behavior that the person exhibited and did well in that situation. That's a competency map that we have. Companies can have their own. That can be customized to an organizational context.

And then conversely, it writes another improved capture which says, Alice, it's important to engage the audience during presentations. This not only keeps the audience interested, but also ensures the message is well received. Try to work on managing the room better and keeping the energy level high during presentations.

And so greater enthusiasm and manage stakeholder expectations as those behavioral tags. And this then now comes back to you. You can edit it, add stuff to, it doesn't get sent to Alice or goes anywhere. We essentially plays this back to you and shows you how to take a unstructured observation, write it down the right way, break it up into what went well and what didn't go well, so that that is very clear.

And then it's very clear about what behaviorally the person is doing well or can do better. And converts it into, as you said, feedback that is more likely to have an impact on changing behavior. And so we call it AI and Answer, and it's got people really excited. And, frankly, it reduces the friction down, because A) it takes less time, B) we can be natural about it, and then C) is that able to convert into a form format that is likely going to have more impact and is direct and kind.

Galen Low: You know what I love the best about that is like, I've heard that conversation before. I know that was like a personal voice note, but you know, like you come out of the room and you see your other colleague who wasn't there and you're like, Oh man, it didn't go so well.

Like they didn't engage the room. It kind of fell flat, blah, blah, blah. But the other interesting thing is that, like from that conversation, which I've probably heard a million times in conversation with people, I think both of us would have only not you and I, but the people talking would only take the constructive feedback part of it and they wouldn't have noticed to celebrate.

But you're absolutely right, in that voice note, they were like, we did so much prep and it still fell flat. Oh, you did prep? Like, that's good! Yeah, like, you're doing something really, really good, because I've also been on the other side of that, where it fell flat, and we did no prep, and yeah that, that wasn't good.

And I'm sure there's something good in that as well, and we're not really wired to see things. I think of it as like this happened, it was good. This happened, it was bad. Kudos is like, Hey, it was great when you help that customer solve their problem. Awesome. And then my constructive feedback would be like, Oh yeah, like we could have probably put an agenda on that meeting before we stepped into it because we did not achieve our goal.

And I'm like these moments of good, bad, good, bad, good, bad. I think that's the other thing that humans, our brains are wired to be like, good, bad, good. Yeah. Like food that's going to be nutritious or food that's going to kill me. Like, okay. Yeah. Good, bad. But there's in between.

Projjal Ghatak: And those observations are all important, right? And that's why everything OnLoop is private till you share it. Like if you're a project managing five projects, 20 people, there's no way to keep track of all of it. There's no way to see the patterns to it. And so we really recommend people make the observations as they go. They don't all need to be shared real time.

There's going to be captured more in the moment of where it happens so you have the raw input. And actually, when we first built the product, we didn't take a lot of unstructured input because we didn't want, we didn't know what to do with it. So we would actually make people write the capture and put in the tags.

Now, we can do that work for you. So we don't mind taking in full conversations and full one on ones, and we'll extract the feedback out of it. And then over time, you get to see a pattern on what behaviors come up more often, worse is not.

Galen Low: That's super interesting. I want to come back to that in a little bit.

But before we go there, I wanted to get your perspective on how this could feed back into a more formal sort of performance management review process because we've been talking about project managers who aren't necessarily the managers of their project team. We talked about this contractual relationship where sometimes you're a vendor but you're embedded, right?

You're at the desk sitting there, but they're not the people giving you feedback, like how would you recommend that, you take some of this feedback and how could it get funneled into something that's a more of a formal performance review or into a performance management system?

Projjal Ghatak: Yeah, so that's what we got first well known for because we were able to automate the back end of that process too.

So in the product, everything becomes captured, right? And so if I'm someone's career counselor, right, and my career counselor has worked on five projects over the course of the year. Over time, I then request captures from people that person worked with. So either I as a manager can do that, or the individual can do that.

And then at a press of a button, we generate something called a prism summary that essentially compiles all of that into a written report, taking key clusters of behaviors and writing that out, and then showing which behaviors came up most often and not. Right? And we usually need about six to eight observations to write that one of each type.

Now, if you're a high performer, you can ask for a lot of observations in our org. We've seen people have 80, 90 in the course of a quarter that then compiles the report out. And we generate one for the individual as a self review, right? In terms of their self feedback and then from others that they've collected.

And then that combines with manager feedback and what managers are best. And managers can request feedback that the individual doesn't see if they choose to, and then that compiles the manager. And we create those two Prism summaries as an outcome that we give to people. And then they run performance conversations using those files.

And so, as a result, we do it end to end, but we actually did the back end of that first and then realized that unless you build a habit around the front end of it and the captures, then it doesn't matter what comes out at the end. But that's the first piece we started prototyping with GPT-3 and realize that we could take these morsels of captures and create a coherent narrative.

And frankly, people get promoted when their managers are good at making those narratives. And they don't get promoted when their managers can't. And so much of individual outcomes at work is not based on your performance, but based on how good your manager is at representing you in calibrations, or in reviews, and that's just not correct.

And we talk a lot about gender bias and pay parity, and all of that is because there is inherent bias in practices, and the straight white male is going to be the loudest voice in the room, typically, and defend all his colleagues. And then we wonder why there are straight white men in all leadership positions who bring their people because they scream the loudest around things. And when we don't have the right visibility factors to judge people fairly and consistently, the loudest voice wins. I get angry when I hear people put lipstick on things like DEI when the core practices is what drives it.

And unless we change those core practices, you're not going to drive outcomes. And when we create a level playing field for all genders, all sexual orientations, all cultures to have the same eloquence and take away that eloquence bias, we can have fairer outcomes across the board or else we'll keep talking about it and nothing will happen.

Sorry, I'm in a philosophical mode today, so you're getting more philosophy than you signed up for.

Galen Low: No, that's good, because honestly, that is the implication here, right? It's that we've got this technology, we've got these tools. The sky's the limit right now. It's people that's the thing, right? It's not the technology that's holding us back the hardest.

What's holding us back the hardest is ingrained old habits, bias, lack of empathy, across the corporate world. A lack of understanding of, or being able to listen to anything but the loudest voice in the room. And if I'm connecting the dots, I was going to ask you, I'm like, how does this help?

But I think the way I see it is that, I use the word eloquence, right? Eloquence bias. And can this and other Generative AI help us get to the point where you know, as managers, I know I'm flipping the lens a little bit, but as managers, we could like put that story together, right? You're, you have your prism analysis, it's feedback that comes in through the person that I am whatever "reviewing", but giving me the structure so that I can go and advocate for that person.

Because maybe in the past I've been like, okay, well, this person did a good job. They hit their KPIs and someone's like, yeah, but you know, they're not quite, fitting in or whatever. They have this area of improvement. I think we should just keep them where they are. Same title, same pay for another three or four years, you know? And you're like, Oh, okay.

Projjal Ghatak: Yeah, and some people's work is not visible, right? Nobody cares about the piping in the house until the pipe bursts and then you don't care about anything else, right? And, but you're not thinking about the pipes. The pipes are not sexy, but the pipes are important. And if someone's amazing at maintaining pipes, you should treat the people well.

They may not present to the CEO every second week, but they still do an amazing job. But that doesn't get visibility because that eloquence bias is a term I came up with. And so thank you for adopting it. But like, I think so much of sort of injustices in the workplace boil down to eloquence. And eloquence comes from confidence, and confidence comes from certain race, gender, sexuality traits, that from childhood tells you, you'll get what you deserve, and you've always got your way.

And those things get ingrained very, very early on. And my founder journey's taught a lot too, in terms of my insecurities and my fears, and I feel like I'm a pretty bro-y dude, but you know, so I don't get a lot of the empathy from people around it. But I still doubt that when I'm standing next to a six foot two white male, I do feel small. And that affects confidence when you're walking into a room and all of that plays on how you show up.

Galen Low: I can relate as well. It's funny, I have this question. My question was like, Are we gonna get better at delivering feedback because of Gen AI, or are we just gonna, get worse? You know what I mean?

Because they're just feeding stuff into the AI. AI's gonna do it for me. But I think it's worth a flip, like, in other words, OnLoop and just Generative AI in general.

Projjal Ghatak: Yep, so there's an article on the 27th of November in The Economist that says, "It shows that AI generated feedback improved employee performance more than feedback from human managers."

And so it's already been researched. The big part is how do you keep the human in the loop? Which is why what we do in our product is we give the person what we call the author back the replay of what it is so that they know what good is like and what good looks like. Because often if you don't know what good looks like, and in the moment you don't have any access to it, you can't get better.

So I'll give you another example. I've got so much better at parallel parking now that we have cameras in the car. Because now I know, like, I can go much closer to the curb and it grows that sensory sense of mine because I'm seeing the camera. So now even without a camera, I'm better because that has now shown me how much I can go through and where if I'm close to the car from an eye perspective, I'm actually not going to hit the car.

And so in the same way, we, if we deliver it back to the person over time, the person will keep absorbing bits and be like, Oh, this is what good looks like. And very likely next time they may be able to have that conversation themselves without the AI assistance, which is our dream, right? Because ultimately I'm in the business of making humans better, not selling technology.

Technology is an aid to make the people better. If they can do it themselves, fantastic. But the technology helps create that loop and show you what good is. And that's why the way we've designed the product is that it comes back to you, doesn't get sent to the subject or goes into an ether where you don't know what happens. It's come back to you, and that's why it's a loop.

Galen Low: I love that. And it's a good theme on Generative AI in general. A lot of people are thinking of it as do this for me. But actually the right use case for now is, teach me.

Projjal Ghatak: Copilot.

Galen Low: Copilot, indeed. I love that. I wonder if we could swing back to the privacy thing, because you mentioned this.

You mentioned that you thought you had to build this backend framework. You thought you had to, have your customers define their own sort of taxonomy and framework for performance. But then you realize, actually, the more data we have, the more we can process this.

And then even that voice note, right? We're all getting a little more used to it now, but as a resistor, I'm a bit resistant. I'm not an early adopter. I'm like, ah, I don't want this getting recorded me, like, mumbling my honest thoughts about a meeting and about somebody into the cloud or into wherever to get processed by a machine.

I'm like, that seems like, you know, covering the phone to talk about someone behind their back, instead of using a mute button. You know what I mean? It just doesn't seem like it's going to go well. How do you inspire trust in people or how can, folks using your tool inspire trust so that people don't feel like they have to hold back.

And give the cleansed version of their feedback to the tool that will cleanse their feedback so that they can get it delivered back and then clean it up even further and have this pristine feedback that is just not honest because we weren't honest with the tools.

Projjal Ghatak: Yeah, no, it's a great question.

Actually, one of the things I see a lot is that there are a billion knowledge workers in the world and one thing is true for every one of them. They don't want to be performance managed, which is why we came up with the term collaborative team development and the CDD framework around driving high performance for people.

And a lot of trust comes when people believe it's for you and it adds value to you versus you being told to do it. And so, we tend to sell into businesses, not into HR for that reason in some ways. Because if the HR team tells you to do something, you're likely to believe that's part of the record and part of the compliance and the paperwork.

Versus if I give you a developmental tool for your aid that is given for your benefit versus the organization's benefit, that creates a little bit more trust. I think to how things are designed, creates trust. And so if a product feels welcoming and feels warm, people want to spend more time in it. And so OnLoop looks and views are consumer app, not an enterprise web app, which is why it's primarily mobile, primarily bite sized, it's primarily colorful.

So we spend a lot of time thinking about design because design creates trust as well, but yeah, there's a lot of baggage and, I sort of carry the baggage of shitty performance management with me, trying to change what I'm trying to do in the same way that Uber carried the baggage of taxi mafias and taxi medallions around the world to drive the change that they do.

And the reality is, if you want to drive any change, you have to deal with the baggage because there's a reason why it hasn't happened yet. The first person who thought of it, like other people have, and you've got to fight, then you've got to fight through it and fight through the baggage and get more early adopters who then are happy and then they give others more confidence to try and as a natural curve there as well.

And it's a journey that it's not going to happen overnight. And which is why people say, when you start a company, get ready for 10 years, because it will take you at least a decade to get it anywhere close to where you want to get to. I still have a long way to live, so not going anywhere.

Galen Low: I'm glad I caught you in a philosophical mood today. There is some incredible insights there. You can make a couple t-shirts about eloquence bias and yeah, they'll be available at the store. No, there's no store.

Projjal Ghatak: I think it is the number one cause of every workplace injustice that exists. And there are others, but it's very, very widespread.

Galen Low: I love that that's part of the overall mission. And actually, I appreciate you, carrying the baggage, but also showing us, as people leaders, that yeah, we kinda need to carry that baggage too. We need to go a bit against the grain to make a change, but it's got to be because we care. It's got to be because we want things to get better.

And we need a way to be able to see that impact progressively, not, you know, it's a big change, but we can't turn away from it just because we don't see the results stay over a day.

Projjal Ghatak: Totally.

Galen Low: Projjal, thanks so much for hanging out with me today. This has been so much fun.

Projjal Ghatak: Thank you, Galen. I had a really good time. Thank you for having me.

Galen Low: Alright folks, there you have it. As always, if you'd like to join the conversation with over a thousand like-minded project management champions, come join our collective. Head over to thedigitalprojectmanager.com/membership to learn more. And if you like what you heard today, please subscribe and stay in touch on thedigitalprojectmanager.com.

Until next time, thanks for listening.