Greg Freeman:
And I think from my side, just to give some early thoughts on this: it's really interesting to work with someone like Ylan, who is such a thought leader and expert in the more technical sides of data, and is definitely considered in the United States a thought leader within the AI space, and is really at the forefront of that conversation.
But also has such a high value put on the people and culture side of data and AI. We all know that there's a perception that a lot of highly technical, highly proficient data professionals think about data as a technical subject, and something that really is about technology and skills and code and all those types of things. Whereas from the first day I ever met Ylan, it was really all about, I understand that we've got to do all the technology and process and AI pieces. But these won't work if we don't bring the people on the journey, and that was a breath of fresh air, personally, when we first started having those conversations.
And that belief system has made the program at Blue Cross Blue Shield, North Dakota so successful as a result. So that's my first point of emphasis.
Ylan, would you like to introduce yourself? And then we'll get going with the first questions, and we'll have a nice flowing conversation as we go through the next kind of 45 minutes to an hour.
Ylan Kazi:
Sure, great to be here, Greg. I’m the Chief Data and AI Officer for Blue Cross Blue Shield of North Dakota. Our company is based in Fargo, North Dakota, and we serve the state market of around 350,000 members. I lead our enterprise data and analytics solutions team, our data governance steering committee, and I'm one of the co-chairs for our AI Steering Committee. So yeah, great to be here, and look forward to the conversation.
Greg Freeman:
Thank you very much. So, insurance—health insurance—typically considered, I think with the experience we've had, insurance firms are a little bit ahead of the curve, actually, in a really positive way. I think pricing and large consumer data sets have been at the core of businesses like yours for a while, so we do see that insurance is fairly forward-thinking in this space. But I guess my first question is:
What did things look like at Blue Cross Blue Shield North Dakota when you first realised you needed a better approach to data literacy and culture? What was the ground zero experience for you, your data colleagues, and the wider business colleagues that made you say, right, this is the right time, it looks right to get going now?
Ylan Kazi:
Yeah, so when I joined the organisation almost three years ago now, the senior leadership team had really come to the conclusion that they wanted to take their data and analytics to the next level. My role was newly created, a few of the roles on my leadership team were newly created. And I think what was a really helpful start was they had already had that realisation. I didn't have to come in and help build a lot of that consensus and collaboration.
They already had their minds focused on wanting to move that to the next level. So that was a great start with just the culture and the leadership of the organisation. As I started, as I kind of did my listening tour, and really got to know the team and the organisation, I saw a few key things. And I’ll preface these with: we see these in a lot of organisations, and even as I talk with my peers as they join organisations, these are very common.
One of them was just a huge focus on the volume of deliverables, whether those were reports, dashboards, I would say some insights, some analytics. It was just very reporting-heavy. So our team would take in different data requests, fulfill them, send them back to the stakeholder. I think in many cases, it almost felt transactional as well.
That was a big challenge, because yes, you can measure volume, but you really want to move and start measuring value instead, and focus on what will actually drive the business forward. So I think that was a really big one.
I heard a lot about "data issues," and I think as data leaders, when you hear that, it can mean one of probably 50 different things. I made sure to really find out what type of data issues were there. In some cases, there were data quality issues. We knew we needed to improve overall data quality. But what surprised me is that in a lot of cases, the dreaded data issues were a lack of understanding of how to interpret data and analytics.
There were also challenges with our stakeholders being able to ask us what they actually needed. They would ask us, can you send me the data? Can you send me this report? But in many cases, they didn't need that. They needed the insights, or they needed recommendations on where to take action. And there was a very large degree of variability with data literacy within the organisation.
I would say the last piece of it was, joining the team, there were a lot of fire drills every day, a lot of things that we needed to do last minute. We really created a better structure in helping our stakeholders plan their work and then also prioritise it accordingly. That helped take the team from just fulfilling things day in and day out, to being more strategic, and being able to allocate more of that time towards the more strategic objectives that would not just help our business, but also help the members that we serve.
Greg Freeman:
Got you. And I'm gonna kind of pick up on a couple of those things, I think, based on what we see both with Blue Cross Blue Shield North Dakota, but also elsewhere. I think you're probably speaking to—or you’re singing from the same hymn sheet—that a lot of the people listening and watching will be singing from, I think.
And it's really interesting that you've noticed there what we would describe actually as the data and AI literacy problem, which is that you've got a business side who don't necessarily know what they want or why they want it, but also, more importantly, how to want it, and how to ask for the right things. And then you've got a data team who are acting as a customer service function for reports and dashboards, and not really able to communicate back why that would or wouldn't be valuable, and whether that is or isn't the right solution.
So, just before I ask you about the first point you made, which is around leadership and leadership buy-in—what have you seen as your programs evolved in terms of closing that two-sided knowledge gap between the parts of the organisation, and closing that two-sided communication gap between data and the business? What have you seen as some of the upsides of that, and how's that changed and evolved, and what pains has it fixed since you've really started to invest in data and AI literacy on both sides?
Ylan Kazi:
Yeah, I would say a few things come to mind. One of them is around just efficiency in general. I’ll give a more tangible example. In the past, our team would be asked to provide a report. That report would be fulfilled—maybe it took a few days or a few weeks—but that would go back to the business leader or the stakeholder team. And it wouldn't necessarily hit the mark the first time. So then we would have to take in additional feedback, redo it, add things, remove things, send it back. It was a pretty drawn-out process in many cases.
A part of that was our data quality, a part of that was being able to build better relationships, but the other part of that was our stakeholders being able to ask for exactly what it is that they needed, and focusing on what were the business problems they were trying to solve.
As we went through our data literacy journey—and we’ve really continued on it—we get much better questions that are asked by our business stakeholders, and there's a much greater focus on what are the business problems that they are trying to solve, and then how does our team help them with the appropriate analytics solution. So we've seen just a great improvement overall in how that works.
I would say the other piece, and this one's maybe a little bit more challenging to quantify: the calibre of questions that get asked in meetings has really levelled up. Where it used to be more anecdotal evidence, or maybe not as many data-driven decisions, now it’s—I’ve sat in some of these meetings where someone will ask, okay, what's the statistical significance of this?
And that kind of—the first time I heard that, it took me out of left field, because I wasn't used to hearing that. But I think there's a benefit to being able to use a standardised language, and focusing first on what does the data show us before we start to move our strategy forward, and before we start to make many of these decisions.
Greg Freeman:
Yeah, that's absolutely massive. We had Helen Blaikie, who's the Chief Data Officer at Aston University in the UK. Shout out to Helen—she’s just been nominated for the DataIQ Award for Best Data Academy. She said the thing that really took her aback, in terms of seeing the benefits in real time, was that all of a sudden, she wasn’t the person bringing data to meetings. Her senior leaders were actually providing the data themselves and presenting it in an effective way.
Those are the kind of tangible things that you can almost feel and taste on a day-to-day basis. It sounds like that’s one for you as well, where you’re all of a sudden hearing the right language, hearing the right questions—all those types of things.
So, with that in mind—and you obviously get exposed to a lot of leadership conversations—how did you go about, at first, driving some of that appetite you already had from the leaders, but actually moving them to a point where they were supportive of a programme like this? Where they were buying into more than just “technology is a silver bullet,” which is what we always hear—“we’ll just put more money into technology.”
What did you do to transition them from “yep, we’ve invested in the technology and the right data professionals” to “now we need to invest in literacy, culture, and wider business support”?
Ylan Kazi:
Yeah, we had a lot of foundational building that we had to do when it came to our team, our data infrastructure, and the way that we work with stakeholders. There were quite a few different parts that we had to focus on. I think the way that we were able to not only emphasise the importance of data literacy, but also show how it would help our business, was by presenting it as part of that broader vision.
What we didn’t do was just come to the table and say, we need data literacy—let’s fund it, let’s move it forward. It was about showing that it was a very important foundational component of being able to evolve as an organisation: how we use data, how we better serve our members.
So really having that overarching vision around it. I would say the other piece was just being very open and honest about the things that we did well, and areas for improvement, not just from our business stakeholders, but also internally within our data team.
I found that by coming forward first and modelling that behaviour of “here’s what we do well as a data team, here are the areas of opportunity,” that made our business leaders more open in sharing what their strengths and opportunities were. We could have a better, richer dialogue, rather than only focusing on the amazing things we were doing and not acknowledging areas that needed attention.
Ensuring that it really was a judgement-free zone, and that we were all working better together, helping each other rather than pointing out where teams were underperforming or had massive opportunities. That culture within our executive leadership team was very strong, very open, and very honest.
Greg Freeman:
This one might be a little bit left field, based on what you've just said, but I think what you’ve just described there—that almost “data professionals having to leave their ego at the door” and say this isn’t just a business-side problem, it’s everybody’s problem and we’re part of it too—how did you... because you’ve got a decent-sized data team of people who are all their own individuals.
What types of conversations did you have with your own team about how they should enable and engage with the organisation, to leave that ego at the door, understand that they’re part of it? What did that relationship look like with your own data professionals to support those conversations and that progress?
Ylan Kazi:
With our data professionals, it was really asking the question: what is in our control? So when something doesn’t go according to plan, or when we would have a deliverable that completely missed the mark, what elements of that are in our control, and how do we do better next time?
It’s really focusing on how do we clean up our own shop first, before we start looking externally and working with some of the other business teams.
That focus on what’s within our control. I think the other piece was just being very open about what data teams do well, where they traditionally struggle—and vice versa: what business teams do well, and where they traditionally struggle.
One of the things that both data and business teams are guilty of is oversimplifying each other’s functions. A good example of that is, when we’re working with business stakeholders, they might say, “Oh, just take it to the data team.” So you think of this data team as this amorphous blob, when in actuality, we all know there are data engineers, data analysts, data scientists—many different types of data roles that do very different things.
But we as data teams, also oversimplify business teams. We’ll say “the business,” but we all know that Marketing is very different from Sales, or Finance, or Operations.
I think just talking about these things openly—things that we normally just keep under the surface—nobody can hide anymore. We’re just very open about the bias we may bring when working with these teams. That made people a lot more comfortable. At the end of the day, we could almost laugh about it—we know where our biases are, we know where we’re going to struggle.
How do we focus on managing around those, and really focus on the relationship at hand? I found that was very important.
Greg Freeman:
Absolutely love that. You know from our perspective on education and what we teach in our classes, that this common language—and part of that common language for businesspeople—is understanding the difference between a data engineer and a data analyst. Not because they’re ever going to become one of those roles themselves, but so they know who to take their problem to.
Because if they've got business questions to answer and they take them to a data engineer, the response is going to be something like, “Why are you here? I don’t understand—have you got a dataset you need? I can help.” That, obviously, creates a level of friction straight away. Not understanding what each other does.
But I think to acknowledge, with your own people, that there’s that level of nuance and complexity on the other side—and therefore they can’t just treat everyone the same—that level of humility is really powerful. So I appreciate you sharing that.
With your programme—and I think one of the brilliant things about the Blue Cross Blue Shield North Dakota programme has been its horizontal nature—of course, you started in the early days with a keen focus on your key strategic cohort in marketing, sales, and customer, but wonderfully, now it’s incredibly horizontal across the organisation.
This ties in with a question that’s been asked in the chat by one of the guests, and also something we were going to speak about anyway.
What was the thinking, as your programme evolved, behind including people from all areas of Blue Cross? And making sure that data, AI, and literacy were treated as horizontal solutions—not just focusing on small groups of people?
Something we commonly see is, “Oh, if we get the execs on board and make them data literate, we’ll solve every problem.” But that’s not the case. It’s a frontline problem, in truth, in our experience. The masses—the silent majority—are your biggest challenge.
So, what was it that made you evolve the programme to cover all bases and bring people in from all different teams and departments?
Ylan Kazi:
With our senior leadership team and our board, it was really essential to have their support—absolutely critical. If you don’t have senior leadership backing, you’re really not going to get very far.
Our CEO has been an incredible champion of all things data. A few years ago, data was our top organisational priority. So there was already a strong focus across the business to prioritise data initiatives. That was a great place to start.
As we progressed, though, we knew we had to bring middle management on board, and in particular, our frontline employees. Without that critical mass, many of them may have just waited it out, thinking, “Is this just the new flavour of the year? Will it be something else next time?” That can lead to people choosing not to get engaged or involved.
A big part of how we accelerated engagement was by identifying those individuals within teams who were excited—those initial champions who had been part of the early training sessions. We asked: how could we build a core initiative team from this group?
That gave us a really strong foundation. We included people from every division in the business, and that helped not only from a change management perspective, but also in ensuring it wasn’t just me or my data leaders reinforcing and driving things forward.
Instead, we had a broad cross-section of individuals across the organisation involved. That was extremely helpful.
I think we also had a real “aha” moment during our onsite session. We had probably invited a few dozen people—and a lot more actually turned up, which was fantastic to see. What really stood out was people voicing where they were struggling, where their fears were.
We showed them that it’s not a bad thing to ask a really basic question. Now that you’ve asked it, here’s how we can help you. That shift—from being hesitant or fearful to saying, “I’m actually learning”—was critical. Because to learn, you have to make mistakes. You have to ask what might seem like simple questions.
But asking those questions will help you in your role, and it will help the company and the members we serve.
Greg Freeman:
Yeah, I think that’s such an interesting word you’ve used there—“fear.” I think people shy away from the idea that their business partners might actually be scared of data and might not want to talk about it.
For me, as someone who spends all their time in this space, it’s one of the biggest blind spots among data professionals right now—that some people may not want to be involved in data. They may not understand why they’d use it. They may be fearful of the things underpinning data, like maths, statistics, and all those sorts of things.
As you go through these programmes—and I’m going to come to one of the questions that’s been asked anonymously (I don’t know who “anonymous attendee” is, probably someone not signed in to Zoom)—as you're upskilling people across the organisation and they’re gaining greater awareness of their own fears, or their own concerns and need for safe spaces...
How is that changing the way the data team works with them? What adaptations have your data professionals made, now that they’re more aware of this emotional and psychological reality?
So many across the world assume that the conversation should flow perfectly and that nobody is going to be afraid of data. How has that changed the way your team works with colleagues, and how have you seen that benefit the wider business?
Ylan Kazi:
I'll give two different perspectives on this. The first is from the data team side. I think it's been very helpful as our stakeholders have improved their data literacy. A big part of that is that when our team is bringing in data and conducting analytics, there’s sometimes a perception of, “Well, you’re using all this technology—surely it’s just automated?” Almost as if we press a button and the data comes in and some kind of magic happens. I’m definitely oversimplifying, but...
I think there can still be that mentality of, “How hard can it actually be?” And I think, by understanding more about data literacy, and even just the core data within their own divisions, there’s a newfound respect. People start to realise, “Oh, actually, this is really complex.”
There’s a lot of work that goes into making data usable and effective. So that’s the view from the data side.
The second perspective is from the business side. Our business stakeholders and leaders now better understand their role when it comes to things like data stewardship. In many organisations that are earlier in their data maturity, they tend to punt all decision-making about data—about standards and so on—over to the data team.
But the problem with that is, in many cases, the data team aren’t the subject matter experts in those specific data elements. That’s where we need much more business involvement—to help us understand what the standards should be, how data is actually being used, and how we should prioritise it.
So for our business teams, this shift gave them a better appreciation for the role they play in the process. And it’s more of a partnership. The data team doesn’t own the data—in fact, nobody owns it. But we are stewards of it. And in many cases, our business teams are probably the better stewards, particularly when it comes to defining all the business language and meaning behind it.
Greg Freeman:
Yeah, I think that’s such an important point. So many people are concerned about governance—probably even more so now, with AI becoming more prevalent and widely used.
Something we definitely see is that the more data literate and AI literate the workforce becomes, the more engaged they are with stewardship and ownership programmes. Right now, we’re typically asking people who don’t really understand data to be responsible for it, to referee it, and to care about it. That’s difficult.
But the more they come on the journey, the more literate they become, the more they understand that, firstly, like you said earlier, this is hard. It’s not an easy thing to do—if it were, everyone would be doing it, and clearly, they’re not. Secondly, they realise it needs to be properly governed in order to deliver the value they want from it.
We’ve had a couple of questions come in from Shireen at the University of Leeds—so, a big traditional university here in the UK. The theme across both her questions is really about: how are you communicating the need to your senior team—and even to senior data and IT colleagues—to shift away from thinking “technology is the answer,” and instead recognising that this is a people problem as well?
And are there any metrics or success measures you’ve been tracking that have helped reinforce the message that it’s no longer just about the technology—it’s about how we bring the people along too? Has anything come to the fore in that space?
Ylan Kazi:
Yes, on the idea of “technology is the magic bullet,” I think every organisation, including ours, has brought in technology in the past that didn’t pan out the way we hoped.
Those kinds of examples are probably the most powerful when trying to demonstrate that it’s not just about the technology—people need to be able to understand it and use it effectively in their roles. Otherwise, all you end up with is a very expensive solution.
So, one of the ways we’ve handled that is by looking at past technology implementations or partnerships where things didn’t land as expected—and identifying that a lack of internal buy-in was often the root cause. Conversely, where we’ve seen success with technology, there has always been a large cultural component that helped drive that success. From a tech standpoint, that’s been a key focus for us.
In terms of tracking improvements year over year, we’ve been using benchmarking. There are various firms—like Gartner and others—that offer standardised benchmarking tools. These allow us to track how we’re improving and evolving year after year, and give us an industry-standard metric we can refer to when talking to senior leaders and the wider organisation.
That benchmarking is really important. I’ve been asked many times, “Where are we on our data journey? How are we doing?” I can give my perspective based on experience—but that’s not data-driven. I want to model data-driven behaviour, and having benchmarks to reference helps me do that.
Greg Freeman:
Yeah, I love that. And Kedra has also asked a question about metrics, but I’m going to link it back to something we talked about earlier in the conversation.
What’s been so exciting about partnering with you on this programme is that, as you mentioned earlier, it was so clearly tied to what Blue Cross Blue Shield North Dakota calls its “ruthless priority” for the year—the one thing that matters most for driving the business forward.
Previously, that ruthless priority was data. Now it’s customer focus. And the connection between those two has been really powerful—how data helps drive customer outcomes and success.
So the question is really a combination of: why do you think so many data strategies elsewhere fail to deliver against organisational priorities? And, following Kedra’s question, what kind of metrics have you used to show the impact of your data literacy programme on organisational goals?
That’s probably where the customer ruthless priority comes into play. So: how have you avoided failure, because of how well the programme is linked to strategy? And how have you evidenced its impact—particularly with your colleagues in customer, where it's been so crucial?
Ylan Kazi:
What I’ve found with data strategies in general—and I’ll admit, I’ve been guilty of this earlier in my career—is that there’s often a tendency to create a strategy in quite an academic way.
What I mean by that is, you sit down and think, “Right, what outcomes are we trying to achieve?” And what you end up with are highly aspirational outcomes. That’s not necessarily a bad thing—but when every goal is a moonshot, it becomes almost impossible to hit them all. So the expectations are often set far too high.
The other issue is that people look at best-in-class technologies and try to design their strategy around those. But for many organisations, that’s not realistic. For us, for example, we’re not a technology company—so we shouldn’t have the same tech stack as Google or Amazon.
We’re also in healthcare, which is a highly regulated sector. You can’t just take what Amazon, Google, or Apple are doing and assume it will work in your organisation.
So when I approach data strategy now, I start with the art of the possible. But I then layer in a second lens: what are the realities of the organisation? Where are we today? What does the culture look like—and where do we want it to go?
That cultural aspect is probably the most important part of getting your data strategy right. Because no matter what you design, culture will make or break it. If you’re not designing your strategy based on the culture you have—or the culture you aspire to—it’s unlikely to go far.
In terms of metrics, I’d say it’s an evolving space. With data strategy and data literacy, many of the skills are foundational in nature, which makes them hard to measure initially.
A good analogy is: what’s the ROI of electricity in your organisation? People would laugh at that—but if the power went out for the entire day, the economic impact would be massive. You only notice it when it doesn’t work.
So foundationally, some of these literacy efforts simply enable people to do their jobs better. But as the programme matures, you start identifying use cases where, for example, something that used to take four months from idea to implementation now takes only two. That’s because you have higher quality data, more data-literate employees, and more efficient processes.
That before-and-after story has been hugely important for us—showing what things looked like a year ago, and what they look like today. And, crucially, showing how the time saved can now be reinvested in other use cases that previously weren’t even possible due to capacity constraints.
Greg Freeman:
Yeah, and I think another thing I’ve noticed—not only in my experience with Blue Cross Blue Shield North Dakota, but more broadly—is just how much the feedback loop from the business improves. You start to gain access to problems that you would never have known about otherwise.
We’re seeing some really strong examples of that across our client base right now. We haven’t done this with Blue Cross just yet, but we have with other clients—rolling out our on-demand platform, our self-serve learning platform.
To address one of the questions in the chat about your rollout—I won’t answer the whole thing—but obviously your programme has been heavily focused on live learning: live online classrooms, some in-person sessions for executives, that sort of thing.
But on our on-demand platform, we now have a use case capture solution built in. We’re receiving thousands and thousands of use cases from across client organisations that feed back into the problems we can help solve.
And I think the more data literate and AI literate a business becomes, the more that feedback loop strengthens—people start to see their business problems as data problems.
They may not be able to solve them themselves—that’s often the job of the data professional—but the fact that they now bring them to you is a huge win for the programme. Even during those executive workshops I ran with your senior colleagues, that shift started to happen in real time. Someone would say, “Actually, we’ve got this issue—I should probably speak to Ylan’s team about it.” That’s fantastic to see.
I’m going to combine another audience question with something we’d already discussed.
So, you’re deeply engaged with AI—and who isn’t in our world these days, right? But we see a clear link between data confidence, data literacy, AI literacy, and AI readiness.
One audience member has asked: there's been a lot of talk about AI—what opportunities or challenges is it creating, and how are people adapting to it?
We’d spoken previously about AI readiness—particularly through the lens of people and cultural enablement. How do you see AI readiness as part of the value of data literacy, AI literacy, and data culture more broadly?
Ylan Kazi:
The way I see it, data literacy is a foundational component of AI literacy and general AI understanding.
Within our organisation, we’re embedding that into our core culture. It’s about having the mindset that change will keep happening—and probably even faster than any of us can predict. That is now the new normal.
We need to accept that in a year or two, even our roles will have changed again. And that’s not a bad thing.
That shift in mindset is key. Another part is that data literacy helps our employees ask the right questions. Take generative AI—it’s huge right now—but a basic question like, “How does it actually do what it does?” sounds simple, but the answer is quite complex.
There are highly technical explanations, but those aren't necessarily useful for most employees. Instead, we simplify things—for example, explaining that you shouldn’t go into Copilot or ChatGPT expecting ground truth. That’s not what it's for.
What it's actually doing is using probability to generate each word in a response. Understanding that helps employees use the tool more appropriately.
So it’s also about ensuring they’re not using generative AI for critical decision-making. Instead, they’re applying their own judgement. Generative AI can inform, but it should never make business decisions without a human in the loop.
I think this approach will help organisations strategically. It will also support them from a risk and compliance perspective. When your people are highly data literate and AI literate, they’re better equipped to protect sensitive data.
They’ll make better decisions—and in doing so, they’ll help reduce the overall risk for the organisation.
Greg Freeman:
100%. I think—well, it’s tricky when you run a business literally called Data Literacy Academy—but I always say we’re a very data-literate organisation by design. We work a lot with our own employees to ensure that’s the case.
The benefit that gives me, as CEO, is a level of assurance around risk. Because 99 times out of 100, our team will make the right decision about how to handle our information—because they are data literate and now AI literate.
That’s perhaps more the stick side of things, but for example, I know they’re not going to go putting any of our IP into an open version of ChatGPT.
We’ve got internal versions of ChatGPT and Gemini that they can use, which makes doing things the right way much easier—and that’s a key enabler. But also, they understand the risks. They know they shouldn’t do it—so they don’t. And I think that’s a major part of AI readiness. It stems from having a higher-than-average level of data and AI literacy across the business.
Really interesting that you’re seeing that too.
Now, I’ll end with this question and then circle back to one from the chat. If anyone has any last-minute questions, do pop them in now, as I’ll move into the full Q&A in just a moment.
Do you have any advice for other data leaders who are trying to do something similar—who want to succeed alongside their technology, AI, and process and automation strategies?
Ylan Kazi:
Yes. Somewhat tongue-in-cheek—but you need to have a high pain tolerance. I’ll start there.
You also need a good amount of grit. I don’t want anyone to think that our journey was smooth sailing. We hit plenty of hurdles and plateaus, and had to course-correct along the way. That’s just part of the process. Whatever your plan looks like on paper, it probably won’t look exactly the same in reality.
I think that’s important to recognise. Another key point—especially for senior data leaders—is that you probably got to where you are because you’re technically strong, highly data-savvy.
But the focus quickly shifts. Success becomes much more about leadership—being able to inspire people and build strong relationships. That’s what keeps the momentum going.
Even today, I rarely go into systems anymore. Would I like to? Absolutely—it’s what got me into this field in the first place. But I find my time is far better spent on building the right culture, fostering the right relationships, and demonstrating how data can drive business value.
Where I’ve seen other data leaders fall short is when they stay too technically focused—especially when speaking to senior leadership. If you’re speaking in technobabble, not explaining things simply, not showing how it links back to strategy, you’ll lose your audience.
That communication piece is so important—but often counterintuitive, because many of us didn’t get into these roles by doing that.
Greg Freeman:
Absolutely agree. I have strong opinions on how data leaders should be using the time they get with the executive team—and I think it’s fair to say you’ve handled that exceptionally well over the course of this programme.
Let’s move on to some questions from the audience.
This one’s quite a literal one, but: what’s been the makeup of your rollout so far? What’s the plan moving forward? You mentioned horizontal groups—what else have you done?
And secondly, from the same person: what’s been the biggest enabler of your success so far?
Ylan Kazi:
I’d say in terms of rollout, we knew the initial stages might be a little bumpy. So we didn’t jump straight into a full-scale launch.
We started with a pilot group—a smaller, cross-sectional group of individuals from across the organisation. That was deliberate. As we ran the pilot, we captured all the hurdles that came up, identified what could have gone better, and used that learning to refine our approach.
When we moved to our second cohort, which was significantly larger, it ran much more smoothly. So my advice would be: definitely get started, but start small. Use that phase to iron out the wrinkles before scaling up.
Looking ahead, one of our big focuses is on how we enable this at scale—making it accessible for all employees across the organisation. You mentioned the on-demand platform earlier—that’s one of the things we’re exploring, along with other options.
Another important step for us is creating a community of practice. That’s key because we don’t want to lose the momentum we’ve built so far. Ultimately, we want data and AI literacy to become embedded into our culture—something everyone is part of.
That means it has to be internally driven. We need the right people and the right business leaders involved. It can’t just be the data team delivering everything—we need support at every level.
Greg Freeman:
Yeah, I think by far the biggest enabler of success for this programme has been, firstly, your ability to engage the executive team and get them on board. But most importantly—and I think this is what’s been particularly powerful with your programme, because I’ve been directly involved—is the way you’ve maintained ongoing engagement with them.
Unlike in some other organisations we work with, you’ve had recurring support from the exec team. You’ve consistently been able to go back to them in short, sharp bursts and say, “This is what’s working. Here’s the success we’re seeing. This is how the organisation is changing.”
I always say I’d love it if every organisation gave us that kind of partnership with the executive team. And I think that’s what’s kept things moving so well—that the execs are part of the feedback loop and are continuously kept up to speed. The way you’ve done that has been really impressive.
Right, two final questions. First: is there anything you wish you’d started sooner?
Ylan Kazi:
In every data leadership role I’ve had, I always feel like we, as a data team, overcommunicate. But the truth is—it’s never enough.
I don’t know what it is, but sometimes you feel like a broken record. However, I had a mentor once tell me, when you’re leading change and transformation, the moment you start to feel like a broken record is exactly the point when the organisation is just beginning to absorb the message.
That really stuck with me. So yes, I might be repeating myself day in and day out—with my team, with others in the organisation—but if you’re an employee in one of our departments, how often are you actually hearing that message? Maybe once or twice a month.
Even though I might have repeated it 100 or 200 times that month personally.
So I don’t think you can ever overcommunicate. I’ve tested this in several roles. I’ve never once had someone email me to say, “You know what, please stop sending these updates,” or, “I don’t need another session about this.” Never. That tells me there’s still more I can do to improve how I communicate and where I communicate it.
As data leaders, this is all basic stuff to us—Data 101. But for many across the organisation, depending on where they’re starting from, it can feel like a postgraduate-level course. Understanding where others are on the journey is just as important.
Greg Freeman:
Exactly. It’s classic marketing and communications, isn’t it? I always use Nike as an example—there aren’t many people in the Western middle class who haven’t already paid Nike for a pair of trainers. Yet Nike still bombard you with messaging every day to make sure you buy the next pair.
It’s the same principle here. Data leaders often struggle with the idea that it needs to be constant. It should become the white noise running through the organisation—repeating the message: “This is working. This is where we’ve delivered value.”
Saying it once is never going to be enough. That’s a brilliant point you made.
Now, you’ve already given us one tip, so let’s round it out—can you give us three tangible tips to help any organisation build their data literacy? You’ve already mentioned constant, recurring comms and being passionate about that. What else?
Ylan Kazi:
Yes, the second would be: the organisation has to be honest about where it currently stands and how it wants to improve.
If there’s no recognition that data literacy is a challenge—or that current outcomes aren’t what they should be—it’s difficult to present data literacy as part of the solution. In our case, that recognition was already largely in place. But in other organisations I’ve worked in, I’ve had to build the case and present evidence before there was any real understanding.
So that’s crucial.
The third would be: radical transparency. I don’t know how else to describe it.
Throughout our data journey, one of the key reasons for our success has been being open and honest—about what my team does well and what we don’t do well. Leading with that level of candour puts people at ease. You build much stronger relationships because you’re not pretending to be perfect.
You’re showing what’s real. You’re acknowledging that your team has vulnerabilities or areas for improvement—but you’re working on them. That’s made a big difference, especially during the tough conversations. If we miss a deadline, I’ll be the first to say, “We messed up. Here’s how we’ll fix it, and here’s how we’ll make sure it doesn’t happen again.”
Blaming other people or deflecting responsibility just doesn’t work. It’s not a good way to lead.
Greg Freeman:
And, unsurprisingly, the answer is not “more technology.” Throwing more tech at the problem isn’t going to cut it—and that’s always reassuring to hear.
Ylan, thank you so much. This hour has absolutely flown by.
It goes without saying—I’m genuinely proud to be partnering with you on what I think is one of the best data literacy programmes we’ve ever worked on. And as a result, probably one of the best out there globally.
That success has come down to your ability to win over the executive team, keep them engaged, and create an environment where your own team operates with humility and provides safe spaces for their colleagues.
When you’ve got those conditions in place, you’ve got the foundation for something exceptional.
We’ve heard a huge amount of value today—and I just want to say it’s a real privilege to work with someone who truly understands why this work matters.
Greg Freeman (closing remarks):
Thank you to everyone who joined us. If you’d like to continue the conversation about data and AI literacy, visit our website—there’s a “Contact Us” button where you can book a one-to-one chat with the team.
If this isn’t the right moment in your journey, that’s totally fine. Stay engaged with the webinars, the newsletters, and all the other resources we offer. I’m sure the time will come when this is the right step for you.
It’s been an absolute pleasure. Thank you again for joining us—and we look forward to seeing you next time.
.png)
.png)
.png)
.png)
.png)
