The dashboards, the tools, the shiny platforms that promise to transform your business, they’re not what’s holding you back. People are.
That was the core message from a recent LinkedIn Live hosted by Orbition Group, featuring Greg Freeman (CEO of Data Literacy Academy) and Pete Williams (Director of Data, Penguin Random House UK). And if you’re serious about data, AI, or digital transformation, it’s one worth sitting with.
Here’s the real talk they shared, stripped down, practical, and grounded in the messy reality of enterprise change.
Where we are now: Everyone’s talking AI. Few are ready for it.
According to Greg, data and AI literacy isn’t new, but people doing it is. For years, the industry’s been shouting about democratising data. But outside the data bubble? Most of the business is still stuck.
There’s a silent majority of employees, frontline managers, sales leads, ops teams, who don’t just lack confidence with data. They’re actively avoiding it. Not because they’re lazy or resistant. Because they’ve never been given a reason to care, or the tools to engage.
And then along comes AI.
Now the expectations are sky-high, the budgets are moving, and execs want results. But if the only people engaging with data and AI are the same 10–15% who always raise their hands, what happens to the rest?
Let’s stop assuming people “just get it”
Pete told a brilliant story from early in his career. He’d rolled out a new MI system to a big UK supermarket chain. After a year, he noticed usage was low. When he asked why, the answer was blunt:
“I look at this, but I don’t know what to do with it.”
It’s a perfect metaphor for how many data programmes still operate today. We build better charts. We train people how to use filters. But we don’t help them think with data. We don’t connect dashboards to decisions.
So what happens? People revert to gut feel. Or worse, they disengage entirely.
Misconception #1: “This isn’t my job”
When non-technical staff say “I don’t do data,” it’s easy to roll your eyes. But here’s the thing: they’re not wrong. Most haven’t been trained, supported, or incentivised to think differently.
That doesn’t mean they’re off the hook. As Pete put it, “There are no decisions in business that are truly free of data.” Even the gut feel merchants are using inputs, they just happen to live in their heads.
The opportunity? Help them connect the dots. What are the five questions they should ask on a Monday morning? What signals should they look for before taking action?
It’s not about turning everyone into analysts. The goal is always to make decisions more informed, more consistent, and more aligned to what the business actually values.
Misconception #2: “Tech will fix it”
One of Greg’s sharper points: “Silicon Valley has spent 30 years making you believe software will solve everything.”
The reality? You can pour millions into data platforms, AI tools, and cloud migrations, but if your people don’t know how to interpret the outputs, challenge assumptions, or spot the opportunities… it’s money down the drain.
Buying the Ferrari is easy. Teaching people to drive it confidently, in your business context, that’s the work that needs to be done to get those pie in the sky outcomes tech companies promise.
So what does good look like?
A few things stood out from both Greg and Pete’s stories:
- Confidence > capability. Skills matter, but confidence is the unlock. If someone’s scared to ask a “stupid” question, they’ll never use the tools they’ve been given.
- Context is king. Teach people to connect data to their roles, not some abstract use case. What does ‘good’ look like in this team, on this system, in this market?
- ROI isn’t just commercial. It’s human. Pete shared that 36% of people who went through their data literacy programme were promoted. 44% took on more responsibility. That’s not just training, that’s real business transformation.
- Executive engagement matters. Not just in sign-off, but in behaviour. If your C-suite says data matters but still makes decisions on gut feel and vanity metrics, your culture won’t shift. And people will notice. That's why role-modelling starts at the top.
Measuring success? Start where the business already cares
Greg laid out a simple way to show impact:
- Find the value hotspot. Where is the business already spending money? Where does data literacy support a bigger transformation?
- Quantify it (as best you can). Even if it’s directional, attach a number to the upside of changing behaviour.
- Tell the story. Don’t just collect feedback. Share it. Make the invisible visible. Success stories are contagious.
Final thought: This is a climb, not a leap
If you’re wondering why your data strategy isn’t landing, or why AI adoption feels like pushing water uphill, it’s not because the tools are wrong.
It’s because the people haven’t come with you.
Changing that takes time. It takes trust. It takes a proper, human-centred approach to learning, one that starts with curiosity and ends with impact.
But if you get it right, you won’t just build skills. You’ll build belief. And that’s when transformation starts to stick.
Catherine: Hello and welcome everyone to this LinkedIn Live. We're gonna be talking all about climbing up that AI and data literacy ladder. My name is Catherine King and I'm the Head of Marketing here at Orbition Group, and I am excited for this discussion today, and we've got two fantastic guests who are gonna be joining me very soon.
But just a reminder, if you are watching us live on LinkedIn or whether you are on YouTube tuning in as well, please drop in any questions or comments you have in the comments section. I'll be keeping a close eye on them over on this screen here. So please do let me know your thoughts. But without further ado, I am going to bring on our wonderful guests here today.
Hello to you both, Greg. Pete, how you doing today?
Greg: Hey Cat, are you all right? Yeah. Really good, thank you.
Catherine: Good? Yeah,
Greg: I'm good as well. Thank you. Nice to see you both.
Catherine: Amazing. Now, for those of for those in the audience who perhaps haven't met you both before, I think a quick introduction is [00:01:00] probably worth.
Wow. Greg, I'll come to you first.
Greg: Thank you. I'm Greg Freeman, CEO, and founder of Data Literacy Academy. We help enterprise organisations to roll out Data and AI Literacy programmes with a specialist focus on non-data professionals. So we always say we teach business people data and AI. You just got a little bit of a, of an insight into today being quite a big day.
Actually. The OnDemand launch video only went live anywhere today, let alone on this, this event. Yeah, really excited to also launch the OnDemand platform to market. Gives us a whole new lens on the way that we deliver programmes, but really happy to be here as well, to have this conversation.
Catherine: Amazing. Yes, it does feel like we've got a bit of an event exclusive, which is always fun. Thanks for being with us, Greg. And Pete, if you could give us a quick intro.
Pete: Yeah, sure. Pete Williams, Director of Data at Penguin Random House UK. But I think I've got a, like a 30 year history.
Terrifying, it's say, out loud in a, in the data industry and probably about 10, 15 years worth of attempting to be a data leader, as we now call it within our echo chamber. So I've made lots of mistakes along the way as I've tried to roll things out, hopefully reflected and learned a lot on those.
And this conversation, one that's really important and I think gets to the heart of the problem. 'cause the tech is easy nowadays, but the people are a whole different ball game.
Catherine: Yeah, a hundred percent. A hundred percent. And I know we're gonna get into very much the people side of things when it comes to that change management, that growth but also, the technical skill behind it.
First question that I wanted to put with you both today is laying out the landscape of the current challenges, the current. State of data analytics when it comes to this literacy piece, and you know why it's more than just a technical skill. So really, Greg, talk to me about what you are seeing in the market at the moment of just, what's going on.
Greg: Yeah, absolutely. Cat, just have a check of the of the quick chat there and see what we're up to. I think there's some questions and comments coming in. Yeah, I think that the interesting thing goes of where we've been over the last kind of two to three years and I think the thing we've all gotta accept is that data and AI literacy as a concept isn't new, but people actually doing it and believing in it is new.
I think is probably the way that I'd put it. And it was a gap I saw at the time, two, two or three years ago where I couldn't really find a way to make myself or my peers more data literate. I. I was in a very data literate environment. Everyone was data first, and that in itself was quite challenging.
The dawn of AI has meant the the dawn of AI. That's a really crappy way of saying that. But the fact that everyone now talks about it all the time and really seems to care about it and thinks it's gonna save the world and solve all our problems means that actually everybody is more aware that.
The business side does need to be on the journey. And there's lots of different ways of solving the problem. We're delighted to be part of that solution, but actually I know Pete has done a lot of work in the apprenticeship space as well across Penguin Random House. There's plenty of different ways to get on this journey.
I think the only people I question are those who aren't on it at all. And if there's not really any investment going into this space yet, they've still got. Tens of millions of dollars going into your big data warehouse platforms and your big hyperscale tech providers.
Are you really using your budget in the best way for your long term success? In fact, one of the things that I see, and this is quite a, I guess a direct comment data leaders who seem to care about the people space and the data and AI literacy space and the data culture space, from what I observe.
Are you being more successful at keeping their position within the organisation, being successful and making themselves into peers with the business side versus those who are still on the technology hype and the technology solution hype, probably less so I think that journey from it not really being a thing that's recognised three years ago to where we are today is really exciting.
But it is really where we take it next that I think matters.
Catherine: Hundred percent. And that journey just feels like you say, it's gone from naught to a hundred. When I was busy crafting conferences, say five years ago you were having a far less mature conversation to, the content.
The transformation has been so fast, and you can forgive that some businesses are still, and I know we're gonna literally talk about that, but it is climbing that ladder of success and some are still, grappling with which ladder is it we should be climbing and why, and what business value that's gonna give us.
Pete, what are you seeing in terms of that overall state of where organisations are and what they're doing and why? It's more than just a technical skill and more cultural.
Pete: Yeah, I think this goes back a little away for me. Obviously I can only really talk about organisations I've been within, otherwise I'm just talking to my peers at environments like the great ones that you pulled together.
But I think the place where this started to really resonate for me, and this will be a short story, I promise, early two thousands, I was in charge of a branch management information system for a large UK supermarket and. I was on the tech dream of delivering six monthly upgrades to this thing and talking to the stores and asking what information they wanted to see.
And it was only a year later that people started to say I look at this, but I don't really know what to do with it. So it's like, why are the usage figures going up? And this is the trouble, you can put all the information that you think is important in front of people, but there's a decreasing number of people, I think who for want of a better phrase, about just being experienced and old hand operators.
There's an assumption from our side of the fence, from the tech and the data side of the fence, that just putting data in front of people means they know natively what to do with it. But quite often people have not associated a picture of performance in a chart with a series of commercial actions inside their organization that say, when this sticks up, I do X.
Or when it goes down, I do y. Or if I just come in on a Monday morning, and these are the first five questions I answer, then actually I've done 80% of my job. And so it got beaten into me and I was overruled at the time by the training function that says we have to build this. And then it became much more successful and I think this is what I carried forward.
That's 25 years later that. Quite often when you roll something out, you concentrate on the training of the system. How do I get in? Where do I find it on my desktop? How do I use the slices and filters and whatever it is, how do I sort columns what nobody ever says is? And so once I've done that, how do I take an action from this system?
And I still see that to this day, and not to denigrate my current employer, but the, there's a kind of a journey for people to go on that says data literacy. And in this case, commercial literacy as well. 'cause the two have to be combined. It is not just learning how to use the tool, it's learning how to use the tool in the context of the environment you're working within to drive commercial success.
And if you're not doing that, then your whole data strategy in any conversation you have is just, it's just meaningless. I continue to preach this message. It's based as I so many of my things on a mistake I've made in the past where I didn't recognize this. 'cause I thought that just building better charts and graphs was the way forward.
But I could definitely see. The aligning better presentation of data to better correlation of people with the outputs they want to get from that for the organisation is the way forward and will lead to success for individuals and companies.
Catherine: Yeah, absolutely. And I often talk about the battle scars of data leadership and I think there'll be many in the audience and beyond who, who share those similar experiences and have gone through that growth For sure.
It's certainly not a unique experience to, to go through that for sure. And it, for me from hearing both of you there, it's that evolution be. Of data literacy, AI literacy, not being a box checking task of, oh yeah we've done that training, or, I've sat through and clicked that PowerPoint as quickly as I humanly possibly could so I could get on with my job and what's important.
But actually that evolution is I. I'm doing this training, I'm learning this, and I've become more literate and it's almost like a Lego block of stacking on top of into my job as opposed to a separate thing. So I guess leading on from those points, what are some of the misconceptions that you hear from, say, non-technical users when it comes to, your, Pete your bringing data literacy into the organisation, Greg, you are doing that, but from a kind of vendor standpoint, what are some of the pushbacks and misconceptions where people say.
Not my job. Go speak to it. And they don't necessarily wanna get involved. Pete I'll throw that one to you first.
Pete: I think you just said exactly what I was gonna say. The so often when you go to a, particularly when you join an organisation, you go around the organization learning what his function does, you know how they contribute to the overall success of the business.
And quite often people say, oh, I don't do data that's not my thing. And so quite often every time, to be honest, whenever I start to write a data strategy or try and make data relevant for people, it's helping them understand that in the modern business world, I. There are no decisions that are taken that are free of data.
Everybody does data. Even people who think they're operating on a gut feel are using data they've manually collected in their head and they're rolling out over time. It's [00:10:00] just that nowadays we've got better methods to capture store process and present it. So usually it is exactly that is the data is not relevant to me.
I don't do that. I'm not a numbers person. You know the paralysis that comes from statistics, which I'm terrible at by the way. This doesn't mean that you can't have a successful career without being able to do statistics, but data literacy itself is really important and I think it's really good that you have to help people understand that quite often the questions they're asking are not necessarily deeply technical, deeply mathematical questions.
It's just, I didn't know I had that data. Where do I ask permission to see that data? Am I allowed to con to see that data? Is it confidential? How do I get more, can you get something to do this? And. Trying to pitch these questions in a way that isn't data related, but is information related to what they need to be able to do to gather knowledge, to be able to make better decisions in their jobs, is the key.
So I think the pushback is trying to get past the fact that, they don't do data because I can guarantee that almost every single one of them does in some part of their job, and definitely in their daily life when they leave the office and hit that smartphone for the journey home, they're doing data constantly on the way of doing that.
And then I think the other problem that I have is. And I've seen this so often in the last 10, 12 years, is the correlation of tech and data. Data is the output of a whole bunch of business processes, which you as a end consumer, are either gonna use to make a decision or in other parts of your role.
You'll be entering into systems and hopefully take accountability for its accuracy. Through the data supply chain that you're creating within that, within the organization. And quite often people correlate the fact that you're doing it inside a computer system with tech and systems as opposed to data and decisions.
And I think those clear divides have to be made so that people realize that being good with data and being how to, make a sensible decision and a well-informed decision from data. It's not the same as being an expert user on one of the systems inside your organization. So I think those are the two things that really start to hammer home to me.
I'm sure Greg has a whole bunch of examples given his exposure to.
Catherine: No, I love that and I love the inclusion of the information and data. And one of the many hats that I wear is lecture for the University of British Columbia. And the way that I explain the differences in data is the information is the flour, the eggs the water, the intelligence is the cake.
And I think as data leaders, we get really caught up in the ingredients. 'cause we know that's where the quality is. We know that if the flour is not great, then it's not gonna help the cake rise and all the rest of it. But actually, when it comes to bringing that to the business, we need to be selling the bakery.
We need to be selling the cake because that's what the end user cares about. And that. Makes a difference to them. So I love what you're saying that, Pete, just to extend that analogy further, but yes, Greg, I'm sure you and I'm gonna give you a time limit on this 'cause I'm sure you hear so many misconceptions when it comes to this work, especially, added into the layer of the AI is just everywhere you look now, you open a news app and it's gonna be, there's gonna be a story in the first 2020 articles.
So what are hearing and, perhaps developing onwards, how are you replying to those misconceptions.
Greg: Yeah I think the first misconception is one that's coming increasingly from. Both data people, data professionals, and senior business professionals actually, which is super interesting to me 'cause it's definitely the dawn of gen AI has probably brought this to the fore if I'm honest, but probably two and a half years ago when we said to people, the likelihood is that 80% of your organisation is were like at best, uncomfortable with data.
Somewhere between disinterested and actually afraid and uncomfortable of having those types of conversations doesn't understand why it's valuable. I think because we all, like I say, we all, but because everybody in the data industry, generative AI has become the absolute norm. And why would you not use it to help you write better code or whatever it might be, depending on which part of the industry you sit in.
It blows people's minds [00:14:00] that most of the organisation is still there, even though certain elements of data and AI have become so every day now. But I can tell you when you actually start speaking to the people on the front line, who, to Pete's point, definitely don't consider themselves data.
People outside of that, maybe top 15% who you do hear from on the business side because they do have appetite, and you're absolutely right. They've got a level of confidence. Absolutely those people exist, but the bigger challenge is the silent majority, and you just don't hear from the silent majority.
That's why they're the silent majority. But if you actually go and have some brave conversations with them and try and talk to 'em about this stuff in the right way, they will openly tell you that it's not their bag and that they don't get it. And it. Puts a knot in their stomach every time they see a graph or chart that they don't understand.
So that's the first thing. And interestingly, senior leaders on the business side I think are also becoming less aware of that somehow. And I think it's because the proliferation of data and AI has happened so quickly. The second point that, again, I probably position as a senior exec C-suite type thing on, in a lot of cases Silicon Valley is built.
On technology as a silver bullet, right? There is more money than cents going into and has been for the last 30 years into making the world believe that technology, software as a service heavily invested products. Can solve all the world's problems. So why would that not be the silver bullet if you've been hit by that messaging as a senior exec for the last five years, 10 years, 15 years?
Even though it's never really worked. And I think if we all look at it in the eye and say, has that new expensive X, y, or Z platform I bought really solved that problem? Probably not in most cases. Shout out if you have, 'cause you've done a really good job. But I think for me it's that actually. The messagings worked 'cause billions of dollars have gone into advertising those types of products as the golden solution.
So it's just getting through to people that the solution [00:16:00] here isn't just the technology and that is something that data professionals are now probably starting to get a bit better at communicating to execs, but they're still getting that pushback. So the cool thing about literacy and culture is it is quite people-centric.
So one thing that execs on the business side do understand is people typically, 'cause they wouldn't be senior execs if they weren't. So if you can bring it back down to what Sarah or John and their team might feel about this and how does that actually make managing them more challenging or how does that help make supporting them more challenging, bringing it very back, very much back to that human centered approach can be a really effective way of communicating the message that, yeah.
Thank you. You've signed off a big budget for the technology. But if you don't ably help me invest in the people, the culture, the literacy, you're gonna miss out on all that thing that you were promised, all the ROI, all those outcomes you were promised. So I think it's about it, it's an evolving challenge.
And the dawn of more common AI, I think is only enhancing that challenge, not necessarily solving it.
Catherine: I love that. I love that. And just to continue my random creative analogies, it is like buying a supercar and giving it to people without a driving license and going there go really fast and do these cool things.
And it's I don't even know how to unlock the door because anything beyond that.
Greg: And ca I used the I used the, like I remember when I first learned to drive and I felt like first time you ever put your foot on the accelerator, you felt like you were going a million miles an hour and it was the scariest thing you'd ever done.
Regardless of whether that car is a Ferrari or a Oxil course, you push that button for the first time and it feels fast and scary. That is what most people go through every time they pick up a data tool for the first time, or even look at data for the first time when it's presented in a meeting and they have to question it.
That's why we give people driving lessons. That's why we get them qualified because it allows them to do it in a way that they're confident with. And I love the car analogy as well. I also really liked your cake analogy that.
Catherine: Thank you. Thank you very much. Now we've got a lovely planned discussion here.
I've got wonderful questions in front of me, but I'm gonna go completely off script 'cause we've had a load of questions come in and that is much better use of all of our times is making sure we are answering questions from the audience. So guys, please keep sending them in. We will do our best to address 'em in the time we have, but first we have Gabriela, writing it, asking how do you suggest to measure the SU success of data and AI literacy initiatives or a program? And I think this leans nicely into what you were saying there, Greg, around, you spend a lot of money, but what are the outcomes? So I wonder if you could speak first to that point around measuring that success.
Greg: Yeah. Such a good question. By the way. It's one of the hardest things to get you, to get everyone's head around and actually execute on. But I think it's the same problem we face with actually measuring the effectiveness and the success of any kind of data initiative. I don't think it's that different, it's just the education's a little bit more ethereal, so it's harder to do.
I think Gabriela, the first thing to think about is what is the understood. Value and success that you're already attaching this to. [00:19:00] So typically when we work, we want to work within programmes or within kind of groups of users who've already got some sort of transformation or initiative going on because there's already a promised ROI or a promise success from it.
Because otherwise that big old programme wouldn't have been signed off for 10 million quid or 20 million quid, or a hundred million quid or whatever. It's, so the first thing to do is make sure that the groups you're working with, the initiatives you're suggesting are strategically aligned. The next thing to do is discuss with the business partners.
Where they see the value coming from those programmes. So we talk about this idea and I did a webinar on it an in, not an internal webinar, but like a Data Literacy Academy webinar, which is available on our website about measuring the success in the ROI of data literacy initiatives. And we believe it's the same way you should be handling that is snappy cat.
That's very good. So we we talk about this exactly the same way that we'd measure typical data initiative, which is that you need to firstly fork like hunt for the value, go around the business, hunt for those opportunities where yes, of course there's appetite. But is there appetite and potential value?
Because appetite isn't the only thing that matters here. There has to be potential technology adoption, value transformation, adoption value, ways of working, adoption value, whatever it might be. Then t-shirt size, that value, I. Sit down with your business people work out whether they agree that if these behaviours do change, what is their perception of a more data and AI literate workforce?
What should their teams be doing that they're not doing right now, and what would that mean to them from an ROI perspective or a intangible benefit perspective? Then the most important bit, get some sort of figure attached to it, get it justified by you as a data office, by the business partner. And even if we can get something signed off by the finance team to say, yeah, we agree that loosely these figures are about right and therefore this investment is gonna be credible.
The trick then is they are actually doing the measurement downstream. So candidly, we've just hired a number of people into a specific team for this now because we believe that the measurement bit is very difficult, and it's a really important part of us ultimately being more successful inside the business.
But you do need to dedicate the time downstream. To speak to the learners, speak to the successful learners, talk about the changes they've made, be able to aggregate that data up so you've got a success metric and a success figure, because actually it is about spending time after the learning, after the community work, et cetera, with the learners so that you can present that measurement back, and then you actually have to present the measurement back.
Because it's not worth presenting them. It's not worth measuring it all. If you don't then present it back and say, look, this is valuable. We now wanna scale it. So yeah, all those things. But I would recommend Gabriela go to our website. It's under the resources section somewhere. There's a whole webinar that me and my colleague and who's our VP of strategy did on measuring the value of data initiatives and data literacy initiatives more generally.
Catherine: Hundred percent. Thank you. Thank you, Greg. Yes. There's a QR code up on the screen there to take you to it. And as with most of these questions, they could be entire reports and white papers just based off of those. But thank you for that, Gabriela. Pete, how about you? When it comes to measuring success of these sorts of initiatives what can you add to what Greg said there on a kind of more broad perspective?
Pete: Yeah, I think Greg's covered most of the, the major program, why you would do this stuff. I think I might take a slightly parallel stance for that is trying to work out who your measure is success for. Because there's an individual success for the people getting involved in this, and quite often if the learning is not aligned to somebody's objectives or how they'd be better at their job, it's just gonna land flat anyway.
So I think, you do have your, large how do you prove this as an organisation, but part of that is by proving as people. And if you can get them bought into the journey that they're on, I think you can be more successful. And just some stats that I've gathered along the way from here in, in the data literacy training that we've been doing.
I think my people are 20% more productive, those who have been through the training as opposed to before. Obviously a lot of these people are already doing like Excel based manual tasks, so if you can get them more competent and confident in doing those tasks and seeing some better ways of doing it.
Gathering a day a week is a lot of time to gather back into organisation. If you do that across everybody, I don't think every role could achieve exactly those figures, but it's a great starting pace. The other plate stat that I've got from the data literacy training we were doing is that 30, 36% of the people who have been through my training have been promoted.
Now, if you want to increase your chances of being promoted, being able to be part of the. The bigger decisions, the more the more structured information that's gathered and the, and what that means for the part of the business that you're working within and how you could take advantage of it.[00:24:00]
If you got the skills to be able to contribute to that as opposed to just do the day job, then there's a good chance you're gonna get notice and you're gonna get somewhere from that. And I think even, I think the other stat that I've got from this is that I think 44% of my people have been involved in more significant conversations and taken on more responsibilities, even outside of the promotion piece.
I. Simply because they can now, they've got skills to bring to the table, which they wouldn't have before. So I think there's a definite personal advantage in the modern world to building these if you want to drive your career forward, if you wanna be more impactful in the organisation.
The other thing I'd say is that there are hard measurements which, that like Greg's outlined in your business case, you said you're bringing these people. Are those people contributing? Have you achieved your outcomes? But the other. The other sorts of stats you can use for this are around the individual confidence as well.
Starting out with some sort of survey based assessment of where the organisation stands. Is it different across different organizational units or different types of roles? And then regularly checking back in with that and seeing what people can now do, what [00:25:00] they feel confident about. Do they feel attached to the organization?
Do they know where to go for data? Do they understand the sort of principles by which you manage data? Do they understand their role? In making sure that data's good. And if you can see that build along the way, I'd be very surprised if that isn't growing alongside, the achievement of the objectives that Greg's very well laid out as well.
So I think there's a, there's more than one lens on this.
Catherine: Yeah, for sure. I love that. And I think, naturally we talk about business value, but I think that personal value in treating people holistically, and as you say that there's such a fantastic statistic on, personal growth, but also, promotions come with it.
Additional salary, bonuses, perks, et cetera, that is going to be making people's lives very different. So it's a case of, marrying up that, hey, it's not just gonna make your life easier, but your, your job prospects and your future. I think it's these moments where we really humanize. That actually we're gonna see, we're gonna see great success.
Amazing. Let's keep going. Let's get through some more of these questions. I've got Angeliki here. Lately everyone wants to implement AI. Yep. Even without being data literate or actually have any insights at all how do you deal with the new AI trend in the lifecycle of data? What are the necessary steps to get your company to need to follow in order to, oh, let's cut off the question.
In order to get to the level of then utilising AI. We've run with this climbing the ladder here, Greg. And as Angelique has noted there, you may not even be up enough rungs to even consider ai. So what's the steps from you, you want to get there, perhaps the business keeps harping on about AI and that's something you've been mandated with as a data leader.
What's the sort of steps you're gonna need to take?
Greg: Yeah, for me, firstly, I think because of the environment we're in this room we have to differentiate between what is actually AI and what most people now mean by AI. That's the first thing we have to do. Most people would redefine, I.
AI generally is generative AI now is my fairly sweeping [00:27:00] statement. But every room I've been in recently that wasn't with data professionals where AI was mentioned, generative AI was basically what they were talking about. It wasn't the complexity of other models and stuff. So I think firstly that stuff has to continue like I advanced data science, et cetera, has to continue to be done by data professionals in the background. And really then it's just about building a level of trust with your business partners to get 'em to trust the results because they are not going to understand why it's all happened, where it's happened and how it's happened.
That isn't what you need them to understand. However, and I'm gonna make a, another fairly bold, sweeping statement if I was the chief data officer right now. I would be doing absolutely everything in my power to roll out generative AI at scale. I believe it's the biggest win for business professionals.
I. In God knows how many years, but definitely around ocean AI, it's easily [00:28:00] understood. It helps them solve their day-to-day headaches and problems. They can feel it, use it, benefit from it multiple times, every single working day. And it can be the Trojan horse for the rest of the industry, right? So I think your problem is though, whereas this perception exists that everybody will do that.
Everybody, why wouldn't everyone pick it up? There is still this foundational work around. Confidence, trust, belief, like, how do I even open it? Yeah, all that stuff is still there. With generative AI, we do feel that we have a concept called the data and AI literacy curve, and we do believe that generative AI adoption is lower down the curve than typical traditional data adoption.
So in my opinion, you're more likely to find somebody or be able to develop somebody to open copilot or chat GPT than you are. To be able to power BI or Tableau or ThoughtSpot, right? So I think it's a massive win. Don't get me wrong. I'm not writing off all the key [00:29:00] problems we've got around governance and all that kind of stuff.
But if we really wanna win as an industry, we have to start to delivering value. And the speed to value from a decent generative AI engagement is just night and day compared to anything else. But to answer the original question, you are obviously then gonna have to do the more boring stuff, like your data quality stuff, like your governance stuff.
But actually you just need more and more people to have access in a trusted governed. So that they feel comfortable to use it, train them on use cases. We take a very clear use case lens in our education around AI and generative AI. 'cause like I said, they don't need to know how the underlying model works.
It's probably a bit boring actually. They just need to understand that they can trust it or they can't trust it. If they can't trust it. What might flag it to them that they can't trust? They can't trust it in that moment. That's the type of stuff that they need to be thinking about. I think. The human in the loop piece will definitely not disappear yet.
And it shouldn't disappear for a long time, but we need to be teaching them just like what it is and how they should use it and how they should challenge it. Not necessarily all the underlying stuff, which doesn't really matter. Yeah, focusing on use cases and outcomes is gonna be the best way to start deploying it.
Catherine: Yeah. And Pete, I'm gonna come straight to you. Do what have you seen work when it comes to going along this journey and, battling with the idea that the bright, whatever the bright shiny thing is of the moment? At the moment, it is AI. In the past it's been big data, et cetera.
We, we know there's usually something that's sparkly and exciting. So how are you dealing with that, that progress up the ladder?
Pete: I think the one thing I'm really grateful of is that Jenny AI has gone through the hype cycle faster, almost anything in history from the original, just apply magic and everything works to realizing that actually asking questions of a tool that's on your phone or on your desktop can be very different to actually achieving a distinct business outcome from that.
So I think there's a lot of work to be done. [00:31:00] I think Greg makes a lot of really good points around the ability to interact. Almost anybody who's used a chat bot function, you're using AI and gen AI has just made it more reliable and more able to understand what you're trying to ask for as opposed to the kind of formulaic match.
This question with this answer from my standard model, so as a data leader, gen AI has been a world of pain. Because people just think everything is so easy. And it's not, that's the real prob problem with this stuff. I think the gen AI part is brilliant for word-based stuff, for interacting around writing better emails, managing your calendar, asking about tasks, all that sort of stuff's great.
And, you can pick up that training from anywhere. Greg obviously does a brilliant way of doing that. I think the, what, the stuff that I worry about is. A partly the human in the loop stuff. We hold AI to this a hundred percent barrier, and yet all around us colleagues are failing every single day.
And nobody says, oh, there's gotta be a machine in the loop. But know people make mistakes all the time. So why can't a [00:32:00] machine make a mistake? It's, it is how you deal with the mistake as it is with supporting any colleague who's made a bad decision or backed the wrong horse at some point. So I think there is something around understanding.
What's a feasible sense of failure rate as opposed to if it's not perfect, it's rubbish. So something to get over there. But I do think that it is the way to get people in because it's really hard to launch the his training on the New BI tool. Oh, I don't wanna go to that.
I've got better things to do. Do you wanna play with ai? Oh yeah. When's the session? I never can't get in that one. Where's the next one? I think part of the. Part of the character set or the opportunity set of any leader of any sort, including data leaders, is understanding the zeitgeist and being able to use the energy that's currently out there to be able to achieve your outcomes.
At the moment, that's Gen AI. It could be almost anything else that people are excited about or that's is a current legal problem, et cetera. That's in the media to be able to use that to, to your objectives. So I think there's all sorts of different ways of going around this. I think the final point I'd make is.
There's no point just giving people at all that makes current bad practice easier to achieve. The thing you really have to understand is, am I rolling this tool out to improve productivity by looking at the processes people are doing and making that process better? Or am I just making this terrible existing business process faster and easy to achieve?
So I think you need to be careful as to what you're trying to automate and how you're trying to do. So before we just throw an AI toilet and walk away, declare it success.
Catherine: Absolutely. And I love what you're saying there about, riding the wave of whatever it is that's the thing in the moment.
But I think what's really. Different potentially now than we've ever seen before, is lay people are using these LLMs, but just think in my personal life, my, my mum has been using chat, GPT and Sawa with interior design. 'cause they're redoing parts of their house at the moment. The idea, five years ago, my mum, accessing a large language model would just make me giggle.
But now she's using it for something that is valuable to her.
Pete: In their personalized people are boundary free on these AI tools. And then you have to be careful how you use 'em inside the organisation. And I think that creates a great sense of friction and frustration as well. You can't just do anything you might do outside world with business data, which isn't your own to share, so they're understanding the.
Practices within which you're, you can do that inside your professional environment is a key part of the training. I think. Sorry to jump on your mom's story though.
Catherine: No. I, no, I think that's that, that is fantastic because like you say you're treating someone as holistic and saying, I know you know how to do these things, but you're, you are almost running with them.
So rather than taking them back to scratch and saying, here's this technology, and they, falling asleep. You can say, we know you've been using it. This is why you perhaps wouldn't use it for this based off of sensitive data. And it all clicks in. It's almost, not patronizing people with data initiatives as well.
And understanding that people aren't necessarily coming at this from scratch, which is, [00:35:00] should be a good thing in celebrated. And definitely the whole, security piece is a whole other piece of content, but unfortunately we've only got about. Five minutes left. And Greg you mentioned something earlier that I wanted to draw back on and probably where we're gonna finish off today.
You, you mentioned some keywords that I've noted down in my notepad here around trust, belief, feedback, and obviously we've spoken quite a bit about culture, how have you seen organisations really create that safe environment? And I know we, we use safe spaces and it's become a bit of a woowoo word, but it is a safe environment for feeling maybe a little bit vulnerable and you're sat around thinking, oh, this isn't really my bag usually.
So how have you seen people create successful upskilling environments where people actually can ask the quote unquote stupid questions and see success from that?
Greg: Yeah, this is really where it all starts. So there's the theory of education is called pedagogy and there is a pedagogical concept called taught fear.
Taught fear, which is something we talk about all the time and provokes [00:36:00] a lot of reaction from people typically actually, where. They'll see it for the first time and they'll be like, whoa, our people aren't scared of data. In fact, all they've got is appetite for it. They love data and AI and that comes back to my point earlier about the 80% silent majority versus 20%, but actually the concept of taught fear is something that we will have all experienced during our lives.
So it's the idea that as human beings, there are topics or subjects that we've taught ourselves, we can't learn. Mine is languages. I've had some really bad experiences in school with languages, and it just, I just will never learn a second language. It like makes my stomach churn just thinking about the idea of speaking Spanish in front of someone.
And unfortunately for a lot of people, maths stats, new technologies, new ways of working, these are taught fears. To talk Cat about Y Mum. My mum works for us. She works for Date Literacy Academy. She was a very early hire where she started by helping out, then she went full time, then we scaled exponentially and now she still works for us.
But she is. She will kill me for announcing this. But she's off sick today anyway with COVID. She's 62, I think, or 63, something like that. Quite frankly, like she's not grown up in a generation where all the digital technologies, all the data technologies, the idea of picking up an LLM on the first day it's released is her world and.
Only since she's been through the data literacy and AI education has she actually started to feel that level of comfort. And I think as an industry we jump to what I describe as usage. This concept that everyone will want to use this thing, whether it's data, whether it's AI, whether it's a technology, whatever it is, actually.
People don't use things they don't trust and they don't trust things they don't understand. So if we don't work on that foundational understanding, which in my opinion, albeit a bit biased, is the literacy piece, you're not gonna get them to trust. What they're working with, why it would be better than the way they've always done it, and therefore they won't use it.
So I think for me it will always come back to, this is far bigger than just an education piece. Our business is 50% change management, winning hearts and minds, internal communications programmes, and 50%. Education. So even we as the company whose name suggests that data literacy as an education point is important, do not believe that educational alone is the solution here.
This has to be far bigger and far more of a change process. And that comes to the point you made there. I think without necessarily even. Specifically asking about it. We have to build reinforcing behaviors within the organisation's culture beyond data and AI literacy and data and AI culture that allow things to be done within a safe space.
If the business, if the corporation has a terrible culture around failure, where it's not okay to fail, it's not okay to be the person who puts a hand up and says, we tried this thing we spent a bit of money on, it [00:39:00] didn't work. Then people won't put their hand up and say, actually, I don't know about data, or I don't know about AI.
Equally, they won't run a test because most data tests do fail and they don't get to production. So all those things that are much bigger than data and AI literacy, much bigger than training. I hate the word training actually in our context. I much prefer education 'cause I think it's more of a holistic word for a journey that someone goes on.
I think all those things play in to the cultural piece, and that's why we spend as much time in our business and as much money in our business on change management and winning hearts and minds as we do just giving people, whether it's live education or on OnDemand education Now.
Catherine: Yeah, no, absolutely.
I think that's, adding that kind of layer of understanding. And I love the piece around languages, and I'm glad you mentioned it 'cause I was allowed a sneaky peek of the product and it made me think of Duolingo actually, and actually how, languages as a whole can feel really overwhelming, but suddenly you've got [00:40:00] this, bite size makes it more manageable. So I think that, that analogy spreads into that piece as well. Pete, no pressure, but if you don't mention your mom in the next 30 seconds, you're gonna look like the worst child ever compared to me and Greg. But what are your kind of rounding out thoughts for this?
Pete: Oh, strangely, it's my mom's birthday today. Yeah I sent her some flowers this morning in Italy. I'm sure she'd be delighted to receive those, or I'll get a, an aggressive message later to, to the fact that she didn't receive them at all. My I'm not gonna give a story about my mom. I was about to launch into one, but I'm not gonna do that. I think what's interesting to what Greg's saying is I think Gen AI gives people that safe space to be able to shout and ask stupid questions without looking stupid.
The fact that you can interact in your own language and you do a pretty good job of understanding what you're asking and allow you to go back again and again to refine your question. And it would always come back with mostly positive, supportive answers that allow you to further your journey without having to put your hand up.
I think might in many ways help people forward to, voice questions that maybe they felt very conscious about voicing in the past. And I'm sure that Greg's, tool, which I haven't yet seen is gonna be great for doing that. So I think it's a massively positive development and I think the reason why people find it easy to adopt is 'cause they don't have to learn anything.
All you've gotta do is ask a question. All the learning's been done behind the scenes. It's a great message from my good friend Emma there as well. Data is indeed our language. Let's talk binary. So yeah, it is, it's all great stuff. And I think the more that we can deploy these tools to get over the bird, the barrier of this feels technical, this feels challenging, this doesn't fill up my world.
And I'm a massive believer that the way we're gonna do data literacy in future is not to teach people how to t trawl through pages and pages of filters and slices in their BI tool, but it will be conversational analytics. I'm absolutely convinced that's gonna be the way forward. As long as you can get your data into a state where.
The tool understands what it needs to bring back from your database and presents it back in a way that you can then make decisions from. And that's where you can start to build in your business logic on top of your ontologies and your knowledge graphs, and really be able to get to something that's genuinely the Star Trek moment.
Or the ORAC moment if you're old enough for Blake seven, like me, to be able to have a conversation, with your machine and actually have some benefit to the organization. And I'm desperate to see that.
Greg: The other thing that you asked about in your question, Cat, which I think is a really often overlooked thing when it comes to the way the data office engages with the rest of the business, is this idea of you, you mentioned feedback loops and like.
How do we know what's going on? And I guess having now run hundreds of data literacy programmes of different sizes, scales, varieties I think one of the biggest wins you can achieve through a data literacy program is the feedback loop. The what's going on in the front line that causes people problems, that ROI can be developed and achieved by solving that then gives you something you can hang your hat on as a data office and say, look.
We found this out. We found this out. Here's the solution we [00:43:00] built. Everybody's now happy. We've made more money. We've saved more money. And in just to, to plug on OnDemand 'cause it's gone live today as a product to market, we have actually been metering on OnDemand with four clients. One in airlines, one in charity, one in government, and one somewhere else.
Just from those four clients alone, we have already collected 19,000 inputs to from the front line of those four businesses. To suggest these are business problems that need to be solved with data. These are data quality problems that exist. These are success stories that we've already solved at a local level, but you reverse engineer those.
You get more data problems that the people are probably experiencing. So I think that feedback loop, the more you can collect from the front line of the business, that's something we are just massive on. For an outcomes perspective from a literacy program, regardless of whether we're doing it or someone else is doing it.
So yeah, to have nearly 20,000 inputs on that coming from the OnDemand platform already. Super exciting for us and obviously it'll become a key part of our arsenals a as a business, I think
Catherine: I'm so excited to see it, Greg, to be honest, I think there's gonna be so much appetite for this in it and yeah, super exciting.
Now I know we didn't get chance to go through all of the questions in the chat I'm sure you both will be open to some messages over LinkedIn as you usually are yet. Then pop those through to both Greg and Pete. But thank you both for your time today. It's been we've overrun on time.
We haven't answered all our questions, but we've had fun doing it. But thank you both for.
We've got some wonderful content coming up soon. And make sure you head over to our website to see all the latest and greatest that we are up to as well. But for now that is everything. Thank you very much, and I can see the flurries of applause. Thank you and stay safe and we'll see you soon.
Unlock the power of your data
Speak with us to learn how you can embed org-wide data literacy today.