The Digital Project Manager

Are You Tracking the Wrong Metrics? How AI Can De-Risk Delivery

• Galen Low and Kelsey Alpaio

Share episode

How do you know if you’re tracking the right project metrics—or if you’ve been chasing the wrong numbers all along? In this episode, Galen sits down with Lior Gerson, Co-founder & CEO of TargetBoard.ai, to unpack how AI is reshaping KPI management and why aligning metrics with business strategy is the real game-changer.

Together, they explore how project leaders can move beyond legacy metrics like velocity and utilization to focus on measures that truly drive impact. From tackling cultural mismatches around KPIs, to building real-time forecasting models, to leveraging AI for smarter, faster reporting—this is a conversation about cutting through the noise and making metrics meaningful.

Resources from this episode:

Galen Low:

How can an organization or a team determine whether they're tracking the right metrics and what can they do if they realize they've been measuring the wrong thing all along?

Lior Gerson:

You need to track whatever is going to help you get to where you wanna go. Everything else is noise.

Galen Low:

Walk us through an example, either of a metric that was maybe the wrong metric, and that process of how they changed towards measuring the right thing.

Lior Gerson:

One of our customers, he wanted to improve like how they forecast what they're gonna deliver every quarter. He was able to take our platform and automate it and it took him like a day. But now whenever he wants, he has like that in front of him, like, what am I gonna deliver this month? Do you think there is a stage where like an AI agent becomes a project manager or replace the project manager completely?

Galen Low:

I think for some people's definition of what a project manager does, yes. Welcome to The Digital Project Manager podcast — the show that helps delivery leaders work smarter, deliver faster, and lead better in the age of AI. I'm Galen, and every week we dive into real-world strategies, new tools, proven frameworks, and the occasional war story from the project front lines. Whether you're steering massive transformation projects, wrangling AI workflows, or just trying to keep the chaos under control, you're in the right place. Let's get into it. Today we are talking about how traditional views of measuring project success are being challenged, and why AI-powered KPI management tools might be your only fighting chance to keep pace with the growing list of variables that could make or break your project's outcomes. With me today is Lior Gerson, Co-founder and CEO of TargetBoard.ai. Lior started his career at mySupermarket, which was the largest price comparison website in the UK servicing tens of millions of users every month. There he was responsible for multi-million dollar projects collaborating with giants such as Walmart, WPP, and Dentzu. And then he moved on to found Vroom.com as CTO, an online car dealership that went public for $2.5B. After that, he led product for Placer.ai, a location analytics unicorn processing petabytes of data. And of course today he leads TargetBoard, an AI-powered KPI management tool. Lior, thank you so much for joining us today.

Lior Gerson:

Thanks, Galen, pleasure of being here.

Galen Low:

I'm really excited to talk to you because you are someone who's like incredibly deep in the world of KPIs and making data informed decisions towards success. Honestly, I hope this conversation goes places that I hadn't planned, but here's the roadmap that I've sketched out for us today. So, to start us off, I wanted to get one big burning question outta the way, like the hot question that everyone wants to know. But after that, I'd like to zoom out from that and talk about three things. First, I wanted to talk about culture mismatch around KPIs and how to translate metrics across different levels of an organization. Then I'd like to talk about how project KPIs are changing and where you see them going. Lastly, I thought maybe we could zero in on some practical ways that businesses and project teams can get started on leveraging AI enhanced KPI tracking and how that might impact their workflows for better or for worse. How does that sound to you?

Lior Gerson:

Sounds good. I think it's gonna a fascinating discussion.

Galen Low:

Awesome. Me too. Me too. All right. Here's the hot question. Lately, there's been a lot of talk about the fact that a lot of projects and actually entire organizations might be actually measuring and tracking the wrong thing. So for projects, I've seen a lot of discourse challenging everything from the formerly sacred Iron Triangle of scope, schedule, and budget, all the way to Agile metrics like velocity. And then organizationally, I've seen folks take aim at things like revenue per FDE, gross margin and utilization. The big thing that's being talked about in my world, at least the project management world, is how project leaders shouldn't just measure project health, but also project impact. So against that backdrop, my hot question is how can an organization or a team determine whether they're tracking the right metrics and what can they do if they realize they've been measuring the wrong thing all along?

Lior Gerson:

Yeah, that's a great question. The project initiative that you're using are set to be aligned with a business strategy, right, with a product strategy. Wherever the business wants to go. But when you put it in the box of project management, all you really wanna achieve is saying, okay, I wanna reach my milestones on time, on budget, with, you know, as little friction and as much, you know, accuracy and predictability as possible, right? So all the magic that you need to track are the metrics that are gonna help you get to that state, right? It doesn't be anything. It can be, you know, how fast tests are moving, where things are blocked, vacation days, how people are using AI. You can track anything as long as it keeps you in that box of what you're actually trying to achieve. But if you're not tracking the things that are gonna help you reach that you're tracking the wrong things.

Galen Low:

I like that perspective because I think we tend to gravitate towards the right things to measure. I even said it in the question like, what is the right thing to measure? You're saying actually. Whatever the thing is that helps move you forward is the right thing to measure. Even if it's way different than what that other team is doing. Even if it's way different than what you did at that other organization. Like it should be aligned to the mission.

Lior Gerson:

You need to track whatever is going to help you get to where you wanna go. Everything else is, you know, is noise.

Galen Low:

That's fair. That's fair. So I mean, for folks who are like, oh wow, Lior, that's, you're speaking my language, we are definitely not tracking the right things. We need to start. How do you kind of map out. What the right things are to push you forward? Is there a sort of best practice or a process?

Lior Gerson:

You need to understand your business, right? You need to understand your project. You understand what's moving the needle? What are the movers and shakers? What's gonna make an impact? Is it people? Is it resources? Is it complexity? How do you sort of unpack everything? Say, okay, these are the components. This is what needs to happen now, how to attract that. It's actually happening and not sort of falling between the cracks. Is public. It like you're gonna have like lots of tools. Like data is gonna be all over the place. It's not gonna be updated. And you're gonna say, oh well people are not updating their system so it's garbage and garbage out and I can't do it. You know, and I'm spending all this time instead showing any returns. There's like a million ways and a million reasons why it's not gonna work, but we need it to work, so, so you know, it's on you.

Galen Low:

Okay. That makes sense to me. I'm wondering, would you be able to just maybe walk us through an example, either of a metric that was maybe the wrong metric and that process of how they changed towards measuring the right thing, or even if there was just some unconventional metric that somebody that you worked with was tracking that actually pushed forward towards the goals, but actually is a pretty non-traditional thing to measure?

Lior Gerson:

Let me talk about like one of our customers, they're called Versa. P. There are a PTF delivering product operations. It's called bu. He is a brilliant guy. We love him, we love working with him, and he wanted to improve, like how they forecast what they're gonna deliver every quarter, right? So plan was actual improving the planning accuracy, improving forecasting. They were running very fast and things are changing all the time, right? So priorities change, capacities change, people move between teams. So it becomes really complicated. And what he came up with was a formula that helps take the previous performance of every contributor and every team and then add that on to whatever, like the current capacity for that team or that group is gonna be. And instead of, you know, having to sort of crunch all that data in Excel, you know, which takes a lot of time. You have to bring in you know, data about vacation days and, you know, tasks and priorities and lots of stuff and doing it in Excel. So he was willing, you know, to take our platform and automate it. And it took him like, like a day. But now, whenever he wants, he has like that in front of him. Okay, what am I gonna deliver with this month? You know, so like the CEO or the CPO comes in and say, Hey, you know, I need to shift priority. Something else comes, you know, is most t he has that what's actually going to change and how it's actually impacting all the initiatives that they're trying to push. I thought it was brilliant taking data, you know. Putting it in the right context to tell the story that he needs to tell.

Galen Low:

I was just at an Agile conference. And I mentioned velocity earlier, right? And everyone's yeah, it's like a poor measure of impact. And sometimes it's just a poor measure because you know, in a lot of agile teams you're meant to have a dedicated team. You know, they're resource full-time. And people I talk to, they're like, that never happens. People are, you know, to your point, they're going on vacation, they're getting pulled away to other projects. So you look at this velocity metric and it's going you know, up and down and you're trying to plan ahead and you're like, what is going on? It's it's not a very good metric. And I would say most people probably wouldn't. They either wouldn't think to have an aggregate metric that is based on dynamic data or they wouldn't know how to build it. Right. But I think that makes a lot of sense. It's almost like. This notion of like real time resourcing where it's okay, well this person's gonna be away. We should expect this dip in this metric about our output, about our performance, about our impact. You know, whatever it is. It's actually grabbing data from other places, not just trying to have the one measure, the one metric to rule them all. I mean, it does function as that, but it's an aggregate metric. Actually, I think that's really cool.

Lior Gerson:

It's interesting you bring this up, right? But think about like how data propagates across the business. That person goes away. It takes time until people know their way, right? If manager's gonna know their way. But until that propagates to the right person who has like the understanding of how it's gonna impact the project timelines, and you know, what, you know, secondary or tertiary effect is gonna have, it's gonna take days or weeks until that's time you're not getting back, right? You add more resources, you're just gonna extend the timeline. It's not like it's gonna make it shorter. You have that learning curve. So the faster you're able to get that information across your organization, getting it to the right people, it's massive impact. It's compounding interests.

Galen Low:

That lag is like so real. In our community, there's always a lot of conversation about well, I wish someone had told me that my lead dev was gonna be going on holiday for three weeks. And then even that lag, like I see it in my role all the time where it's like you get the like a DP, like vacation request and you go in there and there's a process and a flow and bottlenecking that happens before anyone even knows that person's vacation is approved. A little less what like date that is. I like that kind of streams it in.

Lior Gerson:

Vacation is easy. You know, think about, you know, you're working remote, you're on Zoom, and somebody's gone. Who's gone, you know, let go. Nobody told you to be let go.

Galen Low:

Yeah. Okay. I see.

Lior Gerson:

You know, two weeks in, he is not answering my calls. I'm like, what's up? Oh, he's not here anymore.

Galen Low:

He's not here. Yeah. We're hearing a lot about that too, in the community. Right? It's like production's in force without the communication around it. Yeah. And then what. Can you this is a bit of a side quest, but can you at a high level, walk me through what that integration looks like? Because I'm picturing myself opening Excel and being like, okay, I need to like Zapier or make, or do some kind of transformation of data and have it like pasted into a sheet and I have to write a formula and all that. Would I still have to do that in TargetBoard or is integration a little bit more hands free? Hands free, might not be the right word, but less labor intensive.

Lior Gerson:

So TargetBoard.ai works in a very different way from any other system that exists there. We connect to your source data source system and we automatically build all the metrics and all the KPIs that are relevant for whatever you need to do. Right. And then we use AI to generate them. But it doesn't matter if you're looking, you know, at metrics for how your capacity changes, how your velocity changes, how the quality changes. If you're looking at one project, multiple projects, doesn't matter how you work, don't have to change anything about how you work. We automatically learn how you work your systems. It doesn't matter if you're using Monday or if you're using Jira or anything else, right? And then all your metrics, all your reports, all your insights are automatically generated and you can then customize. It's like DIY customization, or you can do it with a team. You'll get the exact metrics and the exact way of tracking everything that you want. It's zero effort, like it's like day one, you're up and running fully confident that the data you need is there and it's accurate and you don't need to worry about excels anymore. By the way, if you do use Excels for like manual tracking and that's like your source of truth, that's fine. We'll connect with Excel. We'll pull it in also. But you don't need to update anymore, right? You don't need to explain how things are. How did you measure this? What is this based on? Like all these questions, they're answered in the platform. Like you share it like with your executive with your C level. They'll have the answers of how things are measured, where they're coming from, where it's going, what's moving the needle. It's not you know, you don't have a situation of, you know, you're sitting in that meeting and you're getting asked a hard question and you don't have the answer. You have to say let me get back to you and let me get back to you. It is like the worst thing to ever say. It's I don't know. I'm sorry I didn't come prepared. I don't know. I didn't study for this test, so it's, although it was like one click and now you like the follow up, is there.

Galen Low:

What I like as well is like the human error component. Like I do track a bunch of stuff in a spreadsheet. There was an incident where I was tracking page views and Google Analytics for, and someone was putting in users, so we had this discrepancy of we thought we were measuring the same thing, but actually we had two different data sources. I was messing it up. But anyways, that notion of somebody might have a different interpretation of where that data comes from to measure that metric versus, I like the idea that like, okay, it's pulling directly from a source of truth. Oh, and by the way, like it generated some of the metrics using AI based on, you know, what the requirements are. To be like, you should probably measure this. And that gets sort of defined and integrated because I've been in a lot of conversations where we can't quite decide on what the right metric of success is, partly because we don't know what we can measure, right? We're pulling from our deck of cards that only has four options and we're like, okay, which one of these four? Whereas it could be a lot of different things To your earlier point.

Lior Gerson:

It's so common. Listen I work at a company and when I came in doing list was tracking like a hundred thousand visits per month. Be the big company. I said, well, where's this coming from? It's, it looks weird. And I was leading product there and turns out that the website was actually getting 10,000 visits, like 10%. That other 90% were like key automation bots. And you know, the business doesn't know. The business doesn't know, also means that when you find this, you have to explain it. So the CMO has to go and say, Hey, by the way, you know how I reported my CAC and everything, you know, for the last year, by the way, I like 90% off.

Galen Low:

Not a fun conversation. Yeah. I wonder if we could use that actually to zoom out a little bit, because I find that metrics are just a thing, right? Like the fact of the matter is that like even if a team or department is measuring the right thing, like there's often like this translation that needs to happen between different levels of an organization or different departments. For example, a scrum team might be tracking customer satisfaction like CSAT. Maybe they're tracking it release over release, but like maybe the leadership is tracking I don't know, like recurring revenue and like retention and churn. Can a tool like TargetBoard help translate KPIs for one audience in a way that makes them meaningful for other levels of an organization? Or is this just one of those things that like a tool on its own just can't fix?

Lior Gerson:

I wanna take a step back on this question. If you think about metrics in general, the traditional way of doing BI in tracking metrics means that you have to build every metric. You have to do like a session and plan every one of your metrics. So, so there's an actual cost on anything you wanna track if you have to develop it. Right. And then when you think about programs like OKRs, qbr, they're sort of very top down. They tell you exactly what you want, need to track, and it gives you like a tunnel vision of this is what I'm gonna track. I might have a lot of other things that are gonna be really important for me because they have to keep the border flow. But I'm not gonna report 'em them, and maybe I'm not getting even track them because I don't have the resources to do that. So with TargetBoard, we track everything and we get it very easy for you to pick and choose. We even recommend what to pick and choose to focus on. Right? And then you can send alerts and get insights and notifications, all that to make sure that thing you care about on track are there. We also make it really easy to explain, like we actually have features saying, how do I explain this up? How do I explain this down? It's like AI generated, okay, take this master, help me explain this to my manager. Help me explain it to my developer or to my sales rep to understand why it's important for me and how I wanted to think about that. So absolutely like getting the whole semantic layer protecting you to make sure that everybody like, like we get this feature request every day. Can I rename my metric? No, you can't because somebody's already has this metric. They always have a name for it. You can't just go ahead and give it another name. And by the way, if that name is not representative, what you're ex tracking, you can't give it that name. So we have all these features to help you sort of protect. Make sure that you're actually tracking what you're saying. You're tracking, and you're actually able to explain what it is that you are tracking, why that is.

Galen Low:

I think that's such a cool feature because I don't know, I've worked with a lot of dashboarding software over the years. Not like deep, but you know, I have exposure to it. And that's usually the gap, right? To your point, it's okay, well that's all fine and good. What is this dashboard telling me? Can you explain this? And then the data storytelling becomes the gap because not everyone is great at it. Some people are great at it, and maybe that's the superhero skill. But I like the idea that AI can help you educate someone about why this metric is important, where it comes from, what it means if it's wavering, and that it's not necessarily I'm sure it could deliver that answer directly from the tool. But I like the, your framing of arming a human to be able to explain their metric well to someone who is not them, to someone who is in a different department, has different specializations, has different knowledge, and still can frame that narrative of what this all means so that we're not just looking at a bunch of numbers and going, wow, that changed 45% week over week. That's interesting. All right, let's move on. Right.

Lior Gerson:

That's what happens.

Galen Low:

Everyone looks at it, they're like shrugging. They're like, wow, that's a big number. And then we like move on. But no one actually understood what.

Lior Gerson:

Never read the reports. You'll have scheduling reports to have a dashboard held. People don't really read the reports unless they have a very specific, they want the reports to be there. Right. They want to know that somebody is tracking that, that is there in case they need it, but they very rarely read the reports like, you know, it's the outlier. By the way, and when we send reports, we also have an AI agent that sort of analyzes them and summarizes them for the business user. So they get okay, so you just gave me this big report on what I to care about.

Galen Low:

Yeah. You act please summarize this for me. I like that it's inbuilt. I'm jumping ahead a little bit, but I remember you telling me in the green room. That there is like this conversational aspect to it as well, like not just the summary of here's what's important about, you know, this report, but I was complaining about, well Google Analytics for my favorite villain, which looks like it's set up to have a conversation to be like, you know, especially in the world of web, I'd be like, how many unique visitors did I get to this page last week compared to this same time, you know, last year? Maybe there's a way to do it. It's never really worked for me. Or even please explain this metric to me. Is set a feature that you have developed are developing just a conversational sort of insights.

Lior Gerson:

So I think, you know, a lot of people are trying to crack it. I think what you're experiencing in the, in GA4, you know, you'll see the same thing in mixed panel table below and so, and the problem is that LLMs are really bad with data. They seem to be getting better, but you know, if you ask them, you know, how many you upload a CSV and you'll tell 'em, Hey, can you tell me how, what's my email open rate? They won't be able to give you the answer unless you really specify like each column, the CSV, what they did, so, and we think we solved that. So we haven't added that to the platform. You are able to ask questions. Get reliable, trustworthy information that the AI also checks to verify that it's not hallucinating. And we think that's sort of, you know, it's gonna be a killer feature because people are going to want to process and consume data in multiple ways. It's not gonna be only conversational or only the dashboard, but you have different ways and different needs to interact with that. And I think that they compliment each other very well. It's a little bit like, you know, retail 10 years ago where I said, am I you investing mobile or not? So, so everybody went into, did omnichannel and that's sort of, I think where data's going also. And it's working pretty cool, but not all of you are using it. I think like when you think about people being able to actually phrase in text the question that you want answered in data is not that straightforward. And because you have the waiting cycle you have you know, you're gonna have ask 10 questions by, you know, to get like the one answer you're gonna, you could have gone on a dashboard like with two clicks.

Galen Low:

Yeah. It's actually might not be shorter.

Lior Gerson:

How many deals did I close last month? Oh, by the way, I actually care about one region. So can you down to, and you know, can you show it to me by it's a lot of writing, only waiting, responsive where, you know on web, it just clicks.

Galen Low:

And I like that idea that data is different, right? And I have experienced that with quite simple data, like time sheet data into ChatGPT, and it's just it just starts hallucinating or, you know, like grabbing the wrong thing. And I'm like, oh, you know, like our brains are so wired to think in tabular data. Not necessarily these LLMs. I like this idea that A, yes, please don't just take that report necessarily and shove it into ChatGPT and expect it to kind of be good at the data analysis. That's why I like that you're building your own feature. BI like that it checks itself, right? Maybe that's sort of standard fair for a data scientist, but you know, I like the idea that it's like there is another process. It's gonna be like before I deliver this very important information about, you know, performance and metrics and impact. Is it right before I just sort of blurt it out?

Lior Gerson:

Trust, trust is everything. That's the risk, you know, with MCP servers. So like you're asking for data from like whatever system, you don't actually know what the AI is gonna do with it and like how accurate like the response is gonna, what we do with our MCP server by the way, and that it verifies that the data and by, that's across all your systems. So pull data from all your systems, combine it together and verify if that demonstrates and the response that you're getting is a hundred percent accurate.

Galen Low:

Yeah, it actually your MCP server. Model context, protocol, that server is the backbone. It's like the thing that's vetting all the data coming into it, not just the one player on the field coming in and coming up. That's very cool. I wanted to dig into the future of metrics. Just get your POV because the sense I get from the circles I travel in is that things are changing around metrics. You can't just have a static scoreboard anymore with the usual KPIs. But I was hoping if I could get your BOV and look a bit into the future, what are some unconventional metrics that you see emerging that might be setting the tone for what KPIs might look like in the future? Or what are your predictions about how organizations will track KPIs like 3, 5, 10 years from now?

Lior Gerson:

So I think at a micro level you'll see much more like AI KPIs and we're already starting to see that, right? You know, how many agents am I running? How effective they are, what are they doing, you know, to the business KPIs. But at the business level, at the high level, I think you care less about how it's done, right? So I think not a lot's gonna change. You know, if you're tracking ARR, if you're tracking churn, you're tracking delivery times, the top metrics are gonna stay the same. So that's the language of your business. What happens underneath the hood is gonna change. And they're also, I think you know. So, so if you take it one step down saying, okay, like we need to deliver A, B, and C, but now we're tracking velocity or we're tracking quality, then the velocity and the quality are the second level of those metrics. If you're taking one, one level below, then you're starting to say, okay, so my velocity, what is it made up out of? Right. Okay. So it's in development work. It's you know, things that are blocked here and there, things that are stuck in review. Things that, you know, had their scope changed in and they had to circle back and, you know, I had to open the context and, oh, by the way, I have all these AIs that are doing stuff and some of them are working, so they're not working. And now I have to spend time and resources and figuring out like, why, you know, GPT-5 is not giving me the same results that I was expecting to before. So you know that it's that level of. Like the nitty gritty details of, you know, what's moving the needle behind the scenes.

Galen Low:

I love that sort of layering because I think you're right. I mean, you know, you're if, as long as your business model's the same, some of those top, I think you described earlier as like a top down metric, right? The OKR tunnel vision, sort of top of the chain might not change that much. I do love that idea that like beneath it could be changing a lot. And that's where, you know, when we started this conversation, you were like, you could just measure what's gonna help you get to your goal. If your goal is to have more people using AI to make better decisions and that's gonna move you towards your goal, then you know, track that at that lower level cycle time. And actually the thing you were mentioning earlier about, people don't always know how to ask for what they want. You had it in a data context. But I'd say, you know, in general for prompting and I was tagged into a post on LinkedIn by Jim Highsmith who was a co-author of the Agile Manifesto, and he had written this story you know, about these folks, kind of like thinking through new metrics. And I don't know if I'm coming from a place of ignorance on this if people are measuring this, but one of the things was prompt quality index. Right. Measuring the quality of people's prompts, because to your point, just because people are using AI doesn't mean it's actually making them more efficient. They could be in circular conversations with a robot and you know, they would still show up as using AI in their job, but maybe not necessarily having like high quality prompts. I dunno how you'd go about measuring that, by the way. But I like that notion that, yeah, what matters to measure at the lower levels will change and can change quite a bit. At the top levels might not change that much in terms of, you know, like business success and like growth and revenue metrics. But I guess maybe for me the question is isn't it a lot like you, you had mentioned it. I think I know what you can, I think I know what your answer might be, but there's a lot of overhead every time you like change a metric 'cause then you need to figure out a way. To connect it to a source and educate people on what it means and then get people in the cadence of reporting on it and understanding how it's pushing things forward. Is that going to be in the future or even now, a sort of like heavy operational overhead that businesses and teams should plan for to be like, yeah, we need like a data person on everything because when we change our metrics because of ChatGPT 5 releasing or whatever, like we need someone who's gonna be. They're ready to wire it up, you know, please wire up this new metric so that we can keep going.

Lior Gerson:

That's the stake today. That's what's happening now. It's been happening for years. Right. You have an analyst for, or analyst team for every department, and that's what they do, right? They just make sure that the metrics are working and it's. It's not even custom metrics, right? They're just reinventing the wheel. Everybody's using pretty much the same metrics right there, but, you know, and you know, big companies, they love replatforming. They love replatforming. It's all they do all day is replatforming, like migrating from one system to another system does the same thing. So, but each time they're replatform it creates this whole big data project of oh, so now we have to update all our reports, right? So there's this huge cycle of job security there. And part of what we do is saying, oh yeah, you actually don't need all that. Like you can replatform. All your reporting, all your metrics are gonna stay the same. If you have to update it, you know, you, you can automatically set it up to notify everybody who needs to know that the metric has been updated. And you don't have to do anything to update it. Does it on its own. I think the world today is broken because people have too many systems or too many data that they can't really use. And I think where it's going is that tools like TargetBoard are going to make it much easier to use that data and to get to that data. And you don't need all those men in the middle. A lot of time don't actually understand your business. Like we, we had a conversation with the chief strategy officer of you know, a ma major company. I don't wanna say the thing. And he said, well, I have a hundred BI and LS people on my team, and every time he has a question, he goes, every time I have a new question, it takes me at least six months to get an answer that is there. The infrastructure is there, but that it has to go through so many people for them to understand what I'm actually asking. It. That's all off. We can give it to you today, by the way. He told me. Yeah, but I can't condemn the data I guys to take it. That's a different problem. It's my problem, but but if you, where I'm going.

Galen Low:

Yeah. No, absolutely. And I've seen that happen. I've seen it happen, especially in financial services. I've seen it in like CX and the customer experience craft, like they're trying to use data to create, you know, an intelligent customer experience based on stuff that's being gathered across many departments. And yet to your point, they're gathering tons of data. No one knows what it is or what to do with it or how to turn it into something that could improve customer experience. It's just there. And like sometimes people are just scoreboard watching and then eventually they get disappointed and they replatform. Right? It's so.

Lior Gerson:

It takes me back to what you said before you know, what are people tracking? So a lot people tracking like costs, right? And cloud costs. Now people are tracking, you know, problems cost, like how much was they spending and how to optimize prompts. Do they spend less money on their l lms? And you have all this data that is just sitting there in companies and it has a cost, it has a storage cost and it tracking cost and amazing. And nobody knows why. And you have all these data, nobody's using that. So it's you know, these like the people who collect stuff and you know, in the garage and don't use them. So company become hoarders endless systems. They don't use them for anything.

Galen Low:

Because you know what we were taught to think it was cheap, right? That storage is cheap. This is just a few numbers. Yeah, exactly. It was cheap and now it's not anymore. And now we're thinking about Yeah, like cost per prompt and you know, these sort of transactional calls. And also the storage, because it's more complex now, and also we've been accruing it for decades. It's not doing anything with it. I can relate to that. I'm a pack rat as well. I try not to delete pretty much anything, so thank you cloud for making that a reality that I could do that. But also to your point, you know, even with some of that stuff, like I'm not doing anything with it. I just, it comforts me that it's there.

Lior Gerson:

Maybe a little bit. Maybe in 10 years you'll need it.

Galen Low:

Yep. Yeah. Exactly. Yeah, that's exactly right. I I had it here to ask you sort of about TargetBoard and your positioning as an AI enhanced KPI tool. You've already kinda walked us through some of the AI driven features. There's the sort of AI driven suggestions of metrics. There is the sort of insights summaries, I guess, from a data perspective. Yeah. I mean, is there any other like AI feature that you wanna talk through, or maybe even the other side of it is like. What made you decide to double down on AI? Like it's in the name, well, it's at least in the domain name, right? AI, like clearly was core to the DNA of the solution that you've built, what are you most proud of about it and where do you see it going?

Lior Gerson:

So I think it's a complicated question, right? I'm proud about a lot of things. I think we've been able to leverage AI from very early on to really augment what we're trying to build an experience. We're trying to create. There's pros and cons of being an AI first company, and I think we're AI second company, but AI is very prominent to everything we do. And every feature we build, we say, okay, how does AI make it better? Not how do we do this with AI, but why are we trying to build and how does AI make this better than you could have before? So lemme give you an example. So we have all your metrics, right? And for any metric, doesn't matter how you slice and dice it, you can create notifications, automations, and alerts and so on and on, right? So you have a deal that's second, a pipeline. You have a button in production, you have a support ticket that's, you know, for a very specific QF customer and you're not meeting your SLA. Create an alert, send it to wherever it needs to go. But having people think about the alerts that they want to create is a hassle, right? People need to think about need, having to think about what they want to create. How, so, for example, what we released this week is that the AI, when you're looking at a metric, it will suggest the automation that you can create. To improve that metric. We already have like recommendations, like strategies for how to improve metric. Like I actually recommend the automation that you can create, like one quick click of a button, create it for you. And this was relatively way harder to do pre AI and now it's much easier. Right? We're moving faster today than I've ever moved any other company.

Galen Low:

That's fair and also insane. Based on what I know about your background, it is a good use of it. I mean, I think like the, what I find is that there's a natural sort of cultural inclusion of AI in the world of data because, you know, coming from like the benefits of machine learning. And frankly, just dealing with swaths and swaths of data. Right? You know, like even in your previous roles where like your tools are like, they're processing petabytes of data. We don't even have a like conceptual, tangible model for what that even looks like in our brains. So it makes sense that okay, we're gonna need like compute power. But what I really like is that sort of and I like it actually in the LLMs too right now. Right? Where it's suggest to me what to do next. Whether it's right or wrong, and then I will learn what you think is the right or wrong next step. And then I will continue to suggest it because, you know, to this day I, you know, I see a lot of folks and especially when I'm outta my element, like I do a thing and I'm like, what should I do next? I'm not sure. But if something, you know, helpful clippy right? Is going, Hey, you writing a letter, do you wanna put a letterhead on? It actually is a great way to I dunno, just kind of cut through discomfort or doubt or hesitation or just like knowledge gaps of actually I wasn't sure what to do next, but actually a little alert would be great. Thank you very much. A little notification and then to your point, at least I'm assuming it also sets up that notification. It's great, I'm on it, I've done it, it's done. Every Tuesday you're gonna get this notification about this metric or when it's out of tolerance or, you know.

Lior Gerson:

It's a very different approach than, you know, what data science was like, you know, three years ago. So taking an example, you know, from Gettacar, you know, we had a team of data scientists that, you know, did autofinance optimization which every time you get a credit application which lender did you send it to? That was, you know, a team that worked on it nonstop trying to maximize, you know, the single digit percentages of how do you get more applications approved. And that was like a data science project. Now it's every feature you build saying, okay, how do I make this better? And it's not that it's not simple, but if you do it right, you're able to get like just amazing results and very quickly.

Galen Low:

I love how fast these things are changing. Right. Maybe just for fun, do you have a question that you want to ask me?

Lior Gerson:

What are the most impactful trends that you're seeing as project management today?

Galen Low:

That's a good one. I think the biggest one is actually the one I started out with, which is this notion that project managers are not just, you know, iron Triangle people trying to deliver projects on time, on budget, you know, within scope. This sort of like this revelation that actually we can be more strategic, but we need to know more. We need to know more, we need to do more. And we were already doing a lot and. Now we feel like we have to be sort of strategists or like delivery strategists. There's the onus is on us to understand the business more. And then there's this mysterious question of cool I'm accountable in some way for impact, but you know, I'm not necessarily involved in measuring that impact. In a lot of cases, we lead a project and then we move on, it launches and we, you know, we go away onto the next thing, off into the sunset. And now there's this big mindset shift of okay, well actually we are value delivery specialists. We actually are meant to create the value, even if that means the plan was wrong and we had to change it, even if it means the scope was wrong, we had to change it along the way. Even if it means battling personalities in the boardrooms and you know, or you know, in the, you know, in the dev pit, there's a much more strategic component to it. And I think we're still trying to figure it out, but then I think it dovetails very nicely with like our adoption of AI. Where, you know, right now it's helping us be more efficient, shorten our workflows, create a bit more speed, but fundamentally help out with some of the stuff that we got bogged down with. You know, a lot of project managers are like, I'm too busy to then also be a strategic partner for my stakeholders. And now there's this sort of rise that where it's okay. What are you doing? Well, okay. Yeah, like all my status reports oh, I've gotta update the dashboard. I've gotta go into Excel and update that spreadsheet. It's cool. You don't have to do that anymore because we can set up agents for that. We can, you know, look to AI for solutions, and then the next thing is okay, but now what do I do with that time? How can I then learn the skills to be a bit more of a leader, a strategic leader in the delivery of value? I think that's an interesting trend in my world that. I think intersects with a lot of this because in a perfect world, we spend more of our time asking ourselves the question, are we doing the right thing to achieve the mission and have the impact? And that's almost exactly where you started about how do we know that we're measuring the right thing? You're measuring the right thing if it's helping you move forward. And I think that's kind of like an interesting aspect of project management and project leadership today.

Lior Gerson:

Very cool. I'm gonna ask you another question. So, I know what I think, but I wanna hear what you think. We've been talking about AI and you know, talking about, you know, how AI can help project managers automate and become more efficient and so on. Do you think there is a stage where like an AI agent becomes a project manager or replaces the project manager like completely?

Galen Low:

I think for some people's definition of what a project manager does, yes. Because there is the sort of level of, oh, project managers are just administrators. They're taking notes, doing meeting minutes, they're following up with action items. They are measuring, you know, an estimate against, you know, an actual, how long it actually took to do all that stuff. Yeah, definitely. I think is almost easily replaced by AI. I think the parts that are not are like the human to human decision making. Like I think for a while at least some of the big decisions, right? Replatforming decisions, risks around, you know, implementation, I think will be conversations between humans. That's why I really like your data storytelling feature because it's not oh, I, it'll be like a little circle with a question mark in it, and like the executive can just click it and it goes, this metric is this. Instead, you're arming the person who owns that metric to tell the story about it. And I think that is the like decision making layer in projects and in business that I think will continue to stay the same, like humans making their case you know, battling things out, convincing one another, persuading one another to make good decisions and maybe think a little bit outside the box and think about, you know, things that maybe haven't been done yet. And those sort of innovative thoughts. And I think that comes from humans interacting with one another. Not to say that AI won't or isn't innovative, like I think it can create ideas that, you know, most of its users haven't thought of yet. But fundamentally I think it's still remixing and I think fundamentally we still treat it as a helper and not someone who's gonna make a decision for us. That's changing. But I think for now, I think.

Lior Gerson:

I agree with that. It also what makes a difference between a good project manager and a bad one and you know, was telling people, oh, you know, we can't afford a project manager, you. Do that on the side, like you manage this project and then like it doesn't work. It doesn't worry because you know it's a profession. Can't do it on the side.

Galen Low:

Yeah, I like that. I like the idea that it's like, it might not replace all project managers, but it will probably replace the bad ones.

Lior Gerson:

It's true for everything.

Galen Low:

Lior, thanks so much for spending the time with me today. This has been so much fun. I love nerding out with you. Before you go though, where can people learn more about you?

Lior Gerson:

LinkedIn, I get back to almost anybody who reaches out. Happy to speak to anybody.

Galen Low:

That's awesome. I commend you. My LinkedIn inbox is such a mess. Thanks again.

Lior Gerson:

Bye Galen, it's been a pleasure.

Galen Low:

That's it for today's episode of The Digital Project Manager Podcast. If you enjoyed this conversation, make sure to subscribe wherever you're listening. And if you want even more tactical insights, case studies and playbooks, head over to thedigitalprojectmanager.com. Until next time, thanks for listening.