Building Influence Without Authority: Lessons from 10 years at McKinsey

I'm good, good. I'm dialing in from Melbourne, so it is evening, 7 PM, but we went to daylight saving. On the weekend, so it is nice and light.

Is this— I, I never keep track of it. Have you gained or lost an hour now?

Spring forward, we lost an hour.

Unfortunate. I'll

take it.

It does, it does drive me crazy. I was trying to schedule a call with someone, um, in the US for a few weeks from now, and it's in that halfway where their clocks won't have changed but ours will have, or the other way around. And so we kept sending each other times and then the calendar invite just wasn't matching. I was like, I swear it's 11 AM Eastern, why is he sending it for— I don't remember if it was an hour later or earlier. Anyway, drives me crazy. So, but that's not what we're here to talk about today. I see we've got lots of folks joining from Spain and London, but also Oxford, Lagos, the UAE. Amazing. Hi everyone, welcome. Um, this is the second kind of session that we've run through the Data and AI Product Management community and here on LinkedIn. Um, for me, the, the whole focus of these sessions is to basically give folks that are either working as data product managers or who have roles that are very similar, those kind of technical roles like data scientist, data engineer, data analyst, analytics manager, but who have to do a lot of the human interaction, talking to people, figuring out what we should do, engaging with stakeholders. And for me, there's probably no topic and skill more important when it comes to that stakeholder engagement than being able to influence folks when you don't have authority. And so that's why Claire is joining us today. That's why she's staying up till late from Melbourne, Australia. I'll give just a very quick glance. I'll let you introduce yourself properly, Claire. But Claire spent 10 years at McKinsey and Quantum Black. And I think if there's one thing the good folks at McKinsey are are kind of known for is building influence both internally and within their clients. So I'm super excited for this conversation and to learn from you, Claire. Welcome.

Wonderful. Well, thanks for having me, Nick, and hello everyone. So a little bit about me. I have been working in data for probably 15+ years. I started out with a maths degree and an electrical engineering degree. And my mum was like, why are you doing maths, Claire? What's that ever going to lead to? Well, I was very lucky because, um, I think once I finished my degree and started working, uh, data science and data just took off. And I was, um, I studied my undergrad in Melbourne but then went to Cambridge and did a master's there and was lucky enough after that to land in McKinsey. And they said, look, we are looking for people who can work with data. We haven't started that group yet, but just come and join us anyway. And so I started there doing a lot of marketing and sales type work. Um, and then, uh, I ended up spending 10 years there because of the variety of work that sort of went on. And, um, through that time worked on a lot of different data, uh, projects. Um, you know, they were just doing churn models for telcos. That was the big thing when I started. You weren't even allowed to use R or Python. Uh, it was all built with sort of proprietary, um, software. And then after I had my first child, it was amazing just to see the switch in terms of data science and Python really coming to the fore and being able to work in that as well. I then joined Quantum Black, which was acquired by McKinsey. There were 40 people when I joined, and by the time I left, there were about 400 people. So it was really good to sort of be build a, um, uh, sort of grow the, the practice there and the way of working as well and sort of work out, you know, what do our data engineers do, what do the data scientists do, and what does everyone else in the middle do. Um, I left that, uh, about 5 years ago and came back to Australia and, um, have led, uh, teams in mainly in the retail and consumer goods space in Australia here. So sort of teams of 30 or 40 people, um, building out their data and analytics function. And then this year went out on my own consulting. So I do work around governance, AI strategy, data strategy, and data and AI adoption as well. So excited to be here. And I think— sorry, just reading some of the comments— to share what I know.

Amazing. Claire, thank you so much for joining us and I'm super excited for today. I also, I guess I didn't put 1+ 1 together and realizing just how early in Quantum Black's journey you joined and that you've kind of seen it go from, I guess, just a concept from how you put it to the machine that it is now. That's awesome.

Yeah, it was great fun.

Yeah. And I think I'm also curious to ask you a few things about the stuff that you're doing now with clients, but maybe I'll leave that for closer to the end. after we've kind of tackled the main topic. So main topic, right? We talk about it a lot, at least for me, like in all the product management circles that I'm in, influence without authority is this— I wouldn't say buzzword because it's not used in the same way that a lot of buzzwords are, which is it's inserted without substance and meaning behind it, but it's really common. Very often the challenges that that we face either as product managers or as data leaders is that we have a mandate, but we can't just order people to go and execute against that mandate. It's not the same as being a GM with P&L ownership and just saying, hey, we're gonna be doing this, so everyone's gonna follow suit. I guess, tell us, tell us, Tell us how you saw this play out.

Yeah, yeah. So at McKinsey, I think it's probably easier with influence at McKinsey because on one hand, people have paid you to be there. And if people aren't helping you, they get pretty— they get told pretty quickly to help you. However, I do think there's— influence is really important in terms of just the relationships you build with people and how you can help them. And I think One of the sort of big lessons for me is you build influence by helping other people and by understanding what is it that they need, what are they worried about, what do they need to get ahead, and how can you sort of make this quite a good relationship together. So, it's not them just giving you things, but also what can they learn from you, how do you help show them in a good light to their bosses as well. And I think I remember one time when I was fairly junior at McKinsey, just seeing sort of this play out in helping a team. And I don't remember what this meeting was about, but I distinctly remember it. And it was quite a big one. And just really working out how did I sort of work with a couple of the people to really stand out of how well they were doing and help them have enough material to share with their wider group. And then the people that wouldn't give me the time of day and then they tried to show what they were doing, it was very different with how people came across. And so I learned from that, how do I try and help people and show people how I can help them more so that I can build influence with them and they can build influence with their bosses as well. I think then moving into corporate, I realized that influence is even more important when you're leading a data or AI team. And that's because you're always working across different groups. You need to try and get someone from marketing to help you. You need someone from operations. You're not their boss. And one of my first roles outside of consulting was with a brand called Wesfarmers, which is a big conglomerate in Australia, and they are huge on influence. You couldn't go to the CEO and get them to tell their direct reports what to do. He'd be like, that's your job. It's your job to go make it happen. And so, you really had to lean on that influence and work out what is it that one group wants and how can you help them and how do you bring them along the journey with you. And so, I think What I learned in terms of how to build it is, one, be really proactive. So, go and find out what people are worried about, what people are interested in doing, and build those relationships. Be open and transparent with what you're doing. I think one of the things I remember in corporate, someone telling me, well, Claire, I'd like you to every 2 weeks make sure that your team presents to everyone what they're doing and they can have a say on what you're doing. And I was like, oh, that's a bit unfair. Why is it that I have to be in front of everyone? But actually, I learned that it is the best thing to do as a data team and for 2 reasons. One, you look really good because you get to share all the stuff you're doing. But 2, you get an opportunity to build this partnership with other people where they can say, oh, did my team ask for that? Actually, I want something else. How do we get the priorities right together? I think by also just showing that transparency, you show this openness and that you're willing to change and collaborate. I think that for a data team, for an AI team that's trying to get other people to do things, is really useful. Then I think The last part about building influence is just go and meet people. Your peers are probably the most helpful people in the organization. I see a lot of teams always trying to speak to the more senior people and how do I get seen with that and how do I have a coffee with the more senior person? Your peers and people even lower in the hierarchy, they are really going to help you make things happen. And I think the more senior you get, the more you realize that your peers' bosses are actually asking them what they think. And going above your peers doesn't really get you any friends or any benefits. It's much, much better to go sort of with your peer to get something to happen rather than going above them. So yeah, lots, lots to do from that side.

I love it. So Work to make people look good, because then that's how they'll appreciate you and you'll have influence over them. Be transparent about what you're doing. And maybe I'd reword it also to like overcommunicate. Maybe don't assume people know because someone in their team was informed. And then number 3, don't be a snob.

Yes. And I think I'd almost put the overcommunicate and both, but 2 ways. 2-way communication. So give people the opportunity to sort of have a say in what you're doing as well, which can feel really hard at first. I've had teams be like, why would I ever do that? But it is really, really useful.

It's interesting. And one thing I've noticed with data and tech teams in general, I don't know, we tend as a group to be a little bit more introverted. And we see all of these things as, oh, it's people playing politics and there's something dirty about it. And I'm like, well, no. I mean, unless if you're literally trying to like steal credit about something or backstabbing someone, you know, lying behind their back, what you're doing is you're enabling the flow of information in your organization. You're making sure that the message, like, we are working on this, or we were told we should work on this, but actually it might not be the right solution. actually makes it to the right people in the organization. And I wanted to ask you, I wanted to ask you a question, but first I'll mention something for everyone listening. Guys, the point of the session is for it to be interactive. So I've got a bunch of questions and if you don't put any in the comments, I will happily ask all of mine. But we would love to get your questions. They might be something super general or if you want, you know, and give us a specific situation you're dealing with or had to deal with and have us comment on it. You're more than welcome to. And now, because there's a 2-minute lag, I still get to ask my question before we get our viewers to send all of theirs and have them flow in. So you mentioned at the start that when you were at McKinsey, it was a little bit easier because there was this top-down mandate in your clients. So it meant that if someone in the client company wasn't cooperating with your consultants, you could basically escalate and get those teams. to work with you. And I think even when you're in an internal role, sometimes that's the case. Like maybe if you're head of data, there's some people you have direct kind of authority over. I wanna ask you, how do you balance? Because for me, anytime you have to escalate or pull rank, you are straining that relationship as well. So how do you balance that, you know, using your authority or your sponsor's authority versus the more subtle kind of influence? Where, like, at what point do you draw the line and you say, you know what, I've tried the 3 things you just mentioned, they're not good enough. I need to pull rank. I need to escalate. I need to go and do something a little bit more aggressive about the situation.

I think it is, I'll say this here, big one. I think it depends. I think you've gotta sort of give it at least 3 tries. those and try different things as well. So I think of a couple of things. I remember in one role, there was just one group that was not going to play ball. And no matter what we did, we weren't getting anywhere. But then sort of what I worked out was it was someone senior on their side. who was putting the blocking down. And actually, the way was to go and talk to her and have the conversation so that she could, you know, unblock it for us. But it wasn't— but it was after I had had multiple conversations with my peers and with people around and just trying to work out what's the dynamic that's going on. So, I think sort of trying 3 times and just sort of standing back and saying, what's going on. I think also getting— sometimes we can be really in our own worlds and just take things the wrong way. Maybe someone's having a bad day for something totally different and it's just not helping. And so I think sometimes having a chat with your boss, with another person on your team, just to get their sense of what's going on can be really helpful as well. Because, yeah, we sometimes sort of get in our own minds. But I think if you have tried, at least you can sort of tell your boss that I did and now I need your help.

For me, the bit about us being in our own heads is especially kind of resonant because I've had it so many times where at first I was like, I was seeing the world very black and white, maybe because I was new to the team or the new team was new to me. And I was like, why are they acting in this way? It makes no sense. And every time I confront them, either I get kind of bullshit responses back, or they stonewall me. And then, you know, as you spend more time in the organization, or like you say, you go around and ask for not the gossip, but like get the context of what is going on, you realize that, oh, they're being very difficult because they've just been hit by a round of layoffs and they're really strained for capacity, or because there's some other project that you're not even aware of that's that there's tensions about. I think we've gotten— yeah, we've gotten quite a few questions. Let's see, I'm gonna spotlight one. So how do you influence people in leadership and talk about real examples? Do you want to take this or start with it first?

Yes, ah, definitely. The first thing is to understand what they care about and what you want them to do as well. So, I think there's 2 parts, what they need and what you want them to do. And I think working on what they need and telling them that you are working on that and showing them and bringing them on that story will help them then get behind you. And I give an example of a consumer goods company I worked with, which we were getting nowhere with the whole leadership team. We were trying to get data governance, data owners in, and they were like, don't get this, don't know. Every meeting was stonewalled until we actually went one by one and just tried to work out, well, what's going on in their part of the business and what is it that they really want and how can we deliver at least one thing to one of them to make them help us? And who is that one person that once they help, they will influence their peers? And we did that. We managed to get one executive-level dashboard that they cared about, that they could understand, and that we took them on the journey, we trained them, we showed them, we made them look good so that they would also be like, okay, I see this team is working on it. They are communicating with me. They are actually making progress. And then he could back us as well. So, I think it is really— if it's a whole leadership team, you're like, how do I get this? Work out who's going to be the influencer across that group and then work out what they want and how can you show that you can give them something before you're asking something from them. I think that sort of exchange is really valuable to sort of take on board, that realize that, you know, especially if you're new in role, you've got to prove yourself a little bit to get that backing from someone. And so how do you sort of work out what they want and give that and help them get it as well?

I love that. And for me, this very strategic approach to not just, oh, we've got requests and we will go out and kind of fulfill them or projects we wanna do, but actually strategically analyze who are my stakeholders, like who are the different players in the leadership team and which ones am I more likely to get that initial goodwill with, or maybe that they have more influence that they can then exert is the sort of thing that when I was in consulting, it was just like the standard way, you know, you, and that's why you, you'd have the meeting before the client meeting and then the meeting to align about that meeting and what we're gonna do and say. And people joke about it, but There's a reason why it's done and it works. And I think when folks are in internal roles, I see it happen less, that whole like, let's map out our stakeholders, not just list them, not just understand like the lay of the land, but more meaningfully think about, you know, we talk about data strategy a lot, but I feel like when most people talk about data strategy, it usually means, do I go with Snowflake or Databricks? Less so about— this aspect of, okay, how do I make sure that my work has maximum leverage in the organization?

I see one comment from Dario there, which sort of comes out this, how do you deal with executives that have been in the company for a long time and have a lot of influence? I've been in this situation a few times. I think you've got an advantage there because there's other people that have been in the organization for a long time, and that executive will have a reputation. They'll either have a reputation of not getting on board, and then, you know, you've probably got an alliance with other executives that you can work with, or you've got a— or they've got a reputation of being someone that's really open and really willing to help, and then it's about having a conversation with them. So, and then the third way, if they've been there for a long time and you can't get time with them, is to work out who on your team knows them. So I was working at a company where someone that reported to me knew one of the execs really well because he had worked on the team with him for a few years. I was like, oh, can you please go and have that conversation and sort of start to turn him around and understand what he wants? And sort of that works quite well as well because you don't have to do everything. It's also working out how does your team influence the people that you need to influence as well.

I really like that. And I think we've got another question I want to spotlight that kind of is the step before some of the stuff that we've been talking about, which is, How do you understand people's motivations? I guess other than you meet them and they tell you, what's the not so obvious side of that?

Well, so I would definitely say go and meet them and ask them what's keeping you up at night, because that, that you at least find out, you know, what they're, what they're worried about. I think the hard part in data is they think, oh, I'm speaking to a data person, so I'm going to tell them what I think about data is keeping me up at night. So that's one. I think then it's also speaking to other people in their business unit who work with them or other people at their level just to understand what's going on. There was one company I reported to the CFO. That was brilliant because he knew what his peers were worried about and he could sort of have a view as well. So I think, yeah, asking a few people's opinions. The other thing is if you work in data, You've got access to some of the data. So maybe have a look at some of the key metrics, like what are the sales going— doing, what is happening with customer. And I think once you have some of those metrics and data to hand and can talk to it, you're actually in quite a good place when you go and chat with people.

And on the data part, I remember you also posted a couple of days ago this grid showing for each executive kind of typically what are the high-level kind of metrics and goals that they have, which for me is always great as a starting point. Like, you will know that if you're— if someone's the CIO, they're probably not mostly concerned about how we're gonna drive revenue up, right? They'll have things that are in their own domain and responsibility that they care about, like cybersecurity risks and adhering to SLAs. Partly because those things might enable revenue growth or cost reduction and profitability and whatever else, but it's not their main thing. And for me, if someone, you know, doesn't have a kind of business background, hasn't gone to business school, even starting there for me is great to know that, okay, vaguely speaking, these are the things marketing cares about. Try and use the language, like use their language to describe the impacts that you're describing. Think about it from their end. And then of course, listen to them like you say, be like, hey, what's keeping you up at night? What's the one thing that if you had a magic wand you could get rid of or change? Yeah, to better understand them.

Yeah, definitely, Nick. Yep. And I think, yeah, sometimes having a little think of what should that role worry about is also good to sort of have that, again, proactive step of How do you come to them with something interesting?

Oh yeah, and that's where having the data becomes super useful. All right, I've got— let me spotlight another question I've seen that we kind of skipped initially. How do you sell yourself as providing influence in a data team when wanting to transition to a more senior role or leadership role in a data-centric position? Now, if I've understood this right, Andrew kind of sees this as I want to both provide influence in the data team, but also I want to transition to a more senior role. And it's kind of framed as, do I go for A or for B? If I've read the question right, and Andrew, correct me if I haven't, I would say like one leads to the other. It's— I know. Yeah. Do you interpret it differently?

Well, I think you're on a good path if you're trying to influence someone and you want to go to a more senior role, because I think showing that— and so if you're in a data team and want to move out of a data team to a more senior role, it is a little bit around how do you find the projects where you can really partner with the business and finding those ones where you can say, oh, I am the data person, but I'm also now going to put the story around it and share it. And you put together— so if you're doing some analysis, you're putting together the document that says, hey, this is what we found. This is what the business should do. Let's take the next— these are the next steps we suggest. And so I think there you're sort of showing how you've got that story part and you can move into the other part of the business. And then I think it's also having the discussions with the part of the business that you want to move into and also seeing if there's any projects. I can highly recommend doing secondments within your business or project trades. It is the most risk-free way to try something new. I did it at McKinsey a couple of times. It was fantastic. And so now whenever I have people that are like, oh, I'd like to try that. I'm like, oh, how do we get a project where you can go and go off to that or a 6-month secondment to try something different? I hope that helped in the direction of the question. But I think if you're trying to influence someone and you're actually making the effort to have those conversations and play a slightly bigger role than you are, then I think you're on a good track.

I agree. Like, for me, it's about exerting leadership, right? Leadership doesn't mean I'm at the top and I have the bigger rank. It's like influencing without authority. It's a behavior that you can demonstrate even as a new grad. And for me, that is a very strong signal if I'm the person deciding on your promotion or to hire you into a different team, that you are ready to get to that next level. All right, how about this one? Spicy scenario. So leadership has one intuition, the data you give them says something else, and they're skeptic. How do you build credibility in that scenario?

I'd check your data. If your data does not match someone's long-term intuition, then firstly check your data that it's right. And then secondly, understand what has changed between that intuition and what's actually going on. And what are the different sort of data points you could bring to it? A lot of the time, we used to say that 80% of the time you're just going to find out what you expect, but there's probably going to be 20% insights that are like, oh, actually, we didn't know about that. Let's use that. Let's use— that's something new. I think— and then the other sort of factor is if people can't see it in real life, then I'd say something's wrong with the data. And then I think here the great example is around left-handed people dying early. That's the maths that's been done wrong, not that being left-handed makes— What? It should be done in a specific way, not the fact that left-handed people die earlier, even though it sounds true. So, yeah, so I'd, one, check that it's correct, two, just work out why and what's the story that comes with that. And then I think having that one-on-one conversation with parts of the leadership around that intuition. And understanding why they have that intuition and why the data shows something differently. Because I think there are also places where that happens.

For me, it's also about, like, are we talking about actual intuition that's been built after years of experience, or is it more the anecdotal evidence that leadership may have been hearing for the last few years? Like, as I read this question, I thought about this project that we did with the container terminal. And we're trying to understand like, what are all the bottlenecks? What are the things slowing down operations? Because it's a super complicated system that things aren't linear. Like you change one thing here and suddenly 6 other things might get affected. And basically there was this intuition that the port could have been going faster. But the problem was there was too much time that the trucks were spending driving from one end of the port to the other to drop off containers. And that, that was a big reason why delays were happening. And what we realized is that is a reason why delays were happening, but it was much smaller than some of the other things that weren't being acknowledged. And the thing was, it's a very visible thing. You can physically see the truck having to drive a kilometer and a half to make that trip. It was harder to see things like, oh, the cranes are having to do a lot of shuffling, or there's these other smaller delays that maybe aren't as big individually, but add up. And so for me, it's also about being able to frame it as almost like collectively we thought whatever the intuition was, was the case. And then we looked at the data and it looks a bit different. Like I found that as a framing device to help, but that's only when you have done all your checks and you know that the data actually is disagreeing with that. collective wisdom or intuition that might exist. It's not, ha, you were wrong and I'm right because the data says so.

Well, I think that anecdotes is quite interesting. At McKinsey for years, we struggled with this and we, because people would be like, oh, but I don't do that. He's like, yeah, the people at McKinsey, it's not a representative sample, guys. We can't go back by what we think. So it's quite, Interesting. I would say most customer, um, chief customer officers or marketing teams, they're the ones who actually have quite good intuition about their customers as well. And if I look, I think Vanessa has a question around, um, the use of industry benchmarks to support your reasons for a data trend. Just sort of linking this one, I think that's very good. I think generally the marketing people are like, oh no, that's not really— that's different to what our customers are like. So, you've got to sort of pick your battles there. If it makes someone not look good, they're probably not wanting to see the industry trend, but— or will try and pick holes. But a lot of the times, I think a lot of execs do want to sort of know how they're sitting against their industry. And they probably do know and they probably do get the data in little snapshots as well. So I think it is a pretty good thing to lean on if you can get it.

I've realized I've been scrolling up the messages to see if we accidentally missed one. I think there's only one we've skipped, which is what happens when you're the one playing nice and you're playing by the rules And someone is playing politics, not just overcommunicating by trying to commandeer the efforts, steal credits, all those, all those fun things. I was going back to the start of our conversation about how do you toe the line between authority and exercising direct authority versus influence and trying to influence your stakeholders in more indirect ways.

I think There's a few ways to do this. One, I would very definitely sort of stand back and say, is someone playing this or are you overthinking it? So just getting a view. And I think speaking to your manager is quite a good way to get that sort of sense check. Because I think, they also will know what personalities are. People at work see things. It's not as though this thing happens and that the managers or that the leaders don't realize that's what someone's doing. So, like, there is a lot about that. I think the other 2 ways that I look at this is one, standing back and saying, what is best for the business? You know, I'm here to make this business grow. What should I be doing and how do I sort of keep that as my top of mind? And then the third one is saying, what am I doing and how do I communicate what I'm doing positively? And that goes back to being proactive. You need to sort of share what you are doing to get the credit. And people will jump in and share it Much more quickly than you, more often. Big up what they do, yes, but they are sharing it. And so if you're not doing that, then I would say start doing it. And you know, you'll start doing it by, hey, end of week email, had a great week, just wanted to let you know these 3 things happened. I can tell you, your boss wants to have that. They want the good news stories. They do not just want you coming to them when it's problem. So if you do have something good, I would say get into that habit of writing it on a Friday afternoon. This is some great things, just wanted to share it with you. And then when you do have a real win that is like, oh, this was great, getting it out to a bigger audience, sharing it with the stakeholders that you now have your monthly showcase with, but really finding those moments of it's part of your rhythm to share what's going on so it doesn't feel unnatural, but also you are getting your news out there. If you've got— what's the Facebook at work called? Oh, there's Viva Engage on Teams and the other one. But if you've got those, start posting your stuff there. People might not read it, people may read it, but I think then you get used to sort of talking about what you're doing and communicating that, and it'll be harder For people to commandeer the effort.

I love it. And it's very interesting to me because you've not given the same answer as before when we talked about toeing the balance between authority and influence, because it's a different setting, but it kind of parallels, you know, just like how to build up influence early on, it helps to make people look good and do things to help them out. Similarly, it helps to just focus on what's gonna be most valuable for the business. And then that takes care of itself in the same way that it helps to be kind of transparent and overcommunicate and overcommunicate in both directions. It helps to reinforce the positives that you're doing and to share your big wins widely. I really like that. And I'm repeating them just because I really wanna make sure that people are kind of noting them down for what they are. And it doesn't matter if you're in a, leadership role, or if you're a data analyst, for me, this advice kind of crosses through different seniority ranks and points of responsibility, because everyone is going to have a boss that needs to be made aware of the wins. Everyone's going to have peers that need to be made aware of the wins. And the last— I wanted to add one more comment on what you said about, and when you've had a big win, share it widely. Going back to what do your stakeholders care about, share it widely in terms that they care about. So it's not, oh hey, we built this super complex model and it's 99% accurate and all the feature engineering was super hard to do, but we managed it. Marketing doesn't care about that. They're gonna care about being able to have more qualified leads automatically filtered through, right? That's the benefit of your 99% accurate model.

Yeah, I agree with that. You'll get a lot, maybe a lot more sort of people mentioning it and commenting if you have something that they can sort of play back to you that they understand.

Exactly. So where was it? We had one more question come through. In your experience, what's the most effective way to turn analytical insight into organizational belief to move people from understanding the data to actually acting on it?

Eugene, great question. I think it can't be done in one step. There are definitely a few, a few parts to this. And I'm sure— so I, I worked with one consumer goods company where there needed to actually be the organizational belief to use data, but, but then also of how well the company was going as well. And part of that was working at different levels. You can't just go top-down, hey, you now need to believe this. But you do need to work out who are those more junior people in different parts of the business that can help you. And because you'll find people that are on the same mission as you to change that sort of organizational belief. So how do you get them to work with you? They've got little projects. Again, this influence of what little projects are they doing? How are you sort of having some of these filter up to more senior people, as well as then starting some of the conversations at the top level as well and working out who are the allies around that? So, I think if you've got one analytical insight, it's working out who's going to be that sponsor. in the exec team and who, or maybe you don't quite have the exec team, but who a bit more senior can help you, but also who more junior backs you up as well and willing to do that. And I think I really flip between the need for executive leadership on things and mid-level leadership. I used to be very much like, well, actually you can get groundswell and go, I think trying to balance both gets you there faster. I don't think— don't just concentrate on the more senior people, but work out who's your network on the ground that can help build that belief as well.

I'm curious what you think about this. For me, I've gotten slightly tired when I hear about insights because my problem with data teams just producing insights is that often the last mile of delivery is the hardest and most critical part to solve. And so I'm like, sure, there's insights in a report, there's insights you can derive from a dashboard, but why do we need to have that step of, oh, and then my business stakeholder needs to take the insight I've given them, reinterpret it themselves, turn it into action. Even if I've got a recommended action, there's still a kind of translation step there. And I think maybe this is a bias because I've worked mostly with data science teams where we were more building systems than producing insights. And so we're like, well, you don't need to do something differently. We're using a model that's gonna prioritize your Salesforce leads. So you just do what you used to do before, which is go down the list in your CRM, except now the way it's prioritized is gonna be more intelligent. Now I'm not saying it should always be like that, and there's a higher cost to build a system than provide insights. But I don't know, I want to get your thoughts on how reductionist I'm being with this little ramble. I'm not saying insights are never useful, but I worry that too many data teams think their job always stops there and never goes to the next stage.

Yeah, yeah. So I think insights help immensely. in terms of just bringing the story. And I think the job of the— the challenge with the data team is that you need to give yourself time to say, I've done it. I've got to do the analysis. I've got to think of what that insight is. I now need to think outside of what my domain is and work out what is that action that they should take. And I remember this very early on in my McKinsey career. People would be like, yes, yes, Claire, you did that really quickly. But now you need to think about what to do with this. And I think getting into that mindset of, yes, I've got this insight now, what's that next step that people need to take, is incredibly useful. And yes, there will be a translation. You might get it wrong 50% of the time or 90% of the time at first, and then 50% of the time you get the— you sort of understand how that part of the business works and what that action is going to be. But I think it's really valuable because the people that you're translating that data for, they've finally got data to sort of back up what they want to start doing. And now you can start to know what they're going to do. And that's going to help you in your career as well. So I think the insights really just brings this to life and helps bridge the sort of the gap between the data team and what's happening in the business. I think for data scientists, they're very lucky that they've got data analysts who are doing this because, yes, you've got this model, but for all these people to trust that model, actually some really simple analysis on what's going on and what might change can help sort of get people on board to change the way it's being scored or being prioritized.

100%. I totally agree. And one thing I've noticed throughout my career is that as I've gotten more senior, partly because I don't do as much of the analysis myself anymore, I've— the number of like insights and bits of analysis I've done has grown smaller, but the impact of the individual actions I've taken, of even like a simple insight about, okay, did a report on our cloud spend because no one's been doing that at the segmentation I'm doing. And I've noticed that we are massively overspending in this one country, probably cuz there's some mistake and we've, we're burning through our credits. It's like that didn't require complex analysis. That didn't require me to do tons of deep dives and follow up with a bunch of stakeholders. But the impact was huge because I was there thinking, Why is my cloud bill so high? Not, oh, there's all this data. What can I do with it? It's exactly like you said, you might not get it right a lot of the time at the beginning, but over time you sharpen that muscle. And also you realize that not every lead is worth pursuing and there's gonna be some stuff that's much more promising than others.

Yeah. And before you know it, you've got more intuition than a lot of people. Because you're sort of looking at it all the time.

You've become those people, leadership with their intuitions.

And I see John's comment here, Build Business is seeing ML thrown as they don't trust or understand. I think some of this analysis is about building that trust with simple analysis so that when you've got the machine learning, People, they might not understand it fully, but they can sort of appreciate what might happen, you know, for the— if they start offering the customers different ads, or this might happen with the customers because they have a better intuition of what's going on.

And actually, this comment just reminded me of a question that I think we skipped earlier that I wanted to spotlight. which is what happens when you've got a business leader with exceptional domain expertise but not so much technical knowledge and they want to stay hands-on?

I prefer not to stay hands-on. So if I try and just re— play back my understanding of this question. We have a business leader who, who really knows their domain, um, and has good intuition, and then, um, you want to do something technical with them. Let's say implement a new churn model and, and how you're going to address those customers, but they don't want to, to hand that over to the, the data team. Uh, I think here it's showing through experiments is probably your first part. How do you do something small to show them the balance? Sorry, show them the benefits. And then being able to prove sort of the dollar terms or the customer terms so that they can see what's happening. I think you do have to gain trust. And you've got to gain trust on, hey, the numbers are right. This is going to take those metrics in the right direction and that it's going to work all day, every day. I think sometimes being in a central data team, we're like, oh, well, it didn't quite work today, that pipeline failed. When you're on the front line and you need it all the time, I can understand why leaders get worried about changing things in that way. So I think it's starting small and slowly growing, and it will take time. Or the other way, a little bit more of a sneaky way, is work out who have you worked with at their level and did something really good for them would make this person jealous and have a bit of FOMO. And that works a treat. Like, if you can get one person happy, all of a sudden there's FOMO at that same level. So I'd also take that. And it's slightly underhanded approach as well.

I love it. I mean, I, for me, it's not even that underhanded, right? It's back to what you said at the very beginning of if I've got a bunch of people that I need to win their trust, let me find who I can make look good. And then everyone looks at them and they're like, well, I want some of that too. For me, it, I also wanna kind of quote back the thing you said earlier that sometimes in the data team were a little bit in our head too much. And so as I was reading the, like, Janek scenario, I was like, okay, should the data team in this scenario be the ones leading with scalable long-term solutions? Like, for me, a business leader with tons of domain expertise is also a super valuable stakeholder to kind of be working with and help guide what we should be doing, because I don't know the business best. They do. And so now, of course, if their decisions end up causing us to have tons of technical debt or missed opportunities, it's a, it's a thing we want to work to address. But also, I'm like, hell yeah, I want to work with a, with one of those high expertise. And I'm almost like, you didn't say it, but I'm kind of picturing someone that's quite intense as I was reading through this paragraph. For me, those are often like, they're difficult stakeholders to work with, but the kind of difficulty breeds high reward as well and high impact. Instead of a stakeholder, it's like, oh sure, I don't care, you guys do whatever. But then 9 months from now, I'm gonna yell at you because there's no impact coming off your work because I wasn't involved and I told you, you guys are empowered to do whatever. I don't know, that's my—

It's hard to win sometimes.

All right, let's see if we can go through the last couple questions, uh, before, before it's time to let you, let you go and wind down for the night. So from Jack, we're— we've got a question about the kind of other side of overcommunication, like what's the What are the limits you should have? How can data teams balance stakeholder engagement and adding business value to avoid building as a service desk and building for vanity metrics?

Yeah, okay. So Jack, when I first saw this question, I was like, I doubt you're ever going to be overcommunicating. But I did work with a team once where actually all the stakeholders were like, we are not going to another meeting with that group. It's a waste of everyone's time. So I'd say listen to work out whether you're overcommunicating. If you are sharing what you're doing and sharing output, I think you're probably not overcommunicating. The challenge is getting people to engage with your communication, I think. And just given that they've got a whole lot of other things going on. And so it would be trying to find that small group of people that has a bit of a diversity across the business that you want to engage more heavily with in sort of a one-to-one and layer on your more broadcast engagement. You say here, how can data teams best balance stakeholder engagement and adding business value? I would say having good stakeholder engagement will increase your business value because people will understand what you're doing. And I think that stakeholder engagement is— think of it at a few different levels, which are already integrated into your workflow. So you're going to have, if you are a service desk, if you have that sort of one front door, approach with a data team, which I think sometimes works really well. It is actually on a one-to-one, once you have that ticket, how you're having the conversations around what are you going to use this for and what's that process that you are making it personal, not just messaging you. I think then there's the next layer out of sharing to a leadership team and a wider stakeholder group, what's on your plan and your roadmap this month, for this quarter, and what does everyone else need to put in? What does other teams want to get in there? There, I think that gives your stakeholders an opportunity to get their value that they want and get their things on there. Then there's a third level of just showing people that, hey, we are a data team that is here and these are some of the things we do. Even showing that, people will be like, oh, I didn't know about that. I wanted that metric last month. Next time I know that. And you'll be surprised. It's little things that people take on. So I think about layering different conversations here and saying that stakeholder engagement is adding business value, even though it's not doing the coding.

Well, it is adding value for sure. Claire, thanks so much. This has been awesome. I'm just conscious of time. We've got a couple minutes left. So before everyone goes, tell us where can they find you? Maybe remind us, cuz not everyone who's dialed in now was here at the start. Like what do you do and what can you help people with?

I'm based in Melbourne, so you can find me in Melbourne. You can find me on LinkedIn, shoot me a note. At the moment, I run my own consulting. I work on data and AI strategy, data governance, responsible AI, and also capability building in the data and AI space too. So yeah, I'm happy to share anything that I know to people. Just get in touch.

Awesome. And if you're not following Claire on LinkedIn, definitely do because she shares so much good stuff and that's how That's how we met and that's how this webinar today has happened. And before everyone drops off, one last kind of thing for me. I've got 2 more speakers coming up, one next week and one a couple of weeks after. If you scan the form there, it's literally a super short survey. It tells you who the upcoming speakers are and they've each given me 3 potential topics. they'd love to talk about. And I was like, all 3 of these or all 6 of these sound awesome. And I didn't wanna pick, I wanted people to pick. So scan the QR code, I'll send it out through Luma if that's how you found this webinar as well. And yeah, this was super fun. And also I, I was surprised by how many people dialed in either early in the morning, UK and Europe time, and also quite late Melbourne time. So yeah, I'll, I'll leave the screen up for the next couple minutes. But Claire, again, thank you so much for joining. Hope you have a good evening and looking forward to seeing your other stuff. Thanks for having me, Nick.

Thank you. And thank you everyone for the awesome questions too. Have a wonderful day.

Thanks everyone. Thanks, Claire. All right.