What’s in this podcast?
Creating a data-driven culture has become one of the most common ambitions in business. It’s written into strategies, discussed in boardrooms and increasingly sits alongside AI transformation plans. But what if organisations have been approaching it the wrong way?
In this episode, Jason Foster, CEO & Founder of Cynozure, challenges the idea that culture is something you can create through awareness campaigns, training or communications alone. Instead, he argues that culture is the outcome of how an organisation is designed – from its incentives and decision-making processes to the systems, behaviours and leadership it rewards every day.
Drawing on leadership theory and practical experience, Jason explores why organisations should stop trying to change the “shadow” and instead focus on changing what’s casting it. He explains why making data easier to access, embedding evidence into decisions, rewarding the right behaviours and solving real business problems will naturally create a stronger data culture.
As organisations race to become AI-first, the episode also offers a timely warning: don’t repeat the same mistakes. Design the environment first, and the culture will follow.
The following interview transcription has been edited for length and clarity.
Introduction
Jason Foster
Hi everyone, welcome back to the show. Today I’m going to be talking about data-driven culture and trying to create a data-driven culture.
Over the last decade, maybe even 20 years, creating a data-driven culture has become one of those phrases that seems to appear everywhere. You’ll find it in annual reports, transformation strategies and, more recently, AI strategies. Of course, it’s in data strategies, but also board papers, earnings calls and reports that public companies put out there.
It’s almost become the default ambition for organisations to want to become data-driven. That has evolved a little to become AI-driven in the current thinking and mindset around how organisations run.
CEOs have also seen this work in one organisation and, when they move to another, they expect it there too. They expect their leadership team to be data-driven. They expect the organisation to use data to drive the way it runs the business and operates.
Even in our own research this year at Cynozure, when we asked data leaders about their top priorities, creating a stronger data culture came out as the number one priority. That’s the first time we’ve seen it take the top spot, but it has always been prevalent in what data leaders are trying to create and what organisations outwardly say they want to do.
Why data-driven culture became the default ambition
I completely understand why. Every organisation wants to help its people make better decisions. They want evidence-based decision making instead of gut instinct and gut feel. They want to unlock the value from the platforms, products and capabilities they may have spent years building. And they want a greater return from the investments they’ve made in data and AI.
These are all great things: really good objectives and brilliant aspirations. We know the quantified benefits organisations can achieve when they apply data and AI properly to what they’re trying to do. They can get really strong results.
So this episode isn’t arguing against a data-driven culture. Quite the opposite. It’s asking whether many organisations have got the approach right or wrong. I’ve often seen organisations getting it almost backwards in the way they approach it.
I’ve been asking myself recently: have organisations been trying to create a data-driven culture in the wrong way? To answer that, I think we need to take a step back and think about culture more generally.
Culture is the shadow cast by an organisation
Quite recently, I came across a leadership metaphor that really resonated with me in the context of working through this question.
Its origins go back to the philosopher Ralph Waldo Emerson. He wrote in his famous 1841 essay that an institution is the lengthened shadow of one man.
Over time, that evolved into a leadership idea many people will have heard before: culture is the shadow of the leader. It’s the same idea, but an updated version. In other words, culture reflects what leaders model, reward and tolerate: the habits, the ethics, the toxicity or otherwise within an organisation.
More recently, organisational design thinkers, particularly in modern agile organisational design and change management circles, have taken that idea a step further.
Rather than saying culture is the shadow of the leader, they describe culture as the shadow cast by the way an organisation is designed. The object represents the operating system: the structure, reporting lines, software, KPIs and even the office layout. The shadow is the culture all of those things create.
I think it’s a really powerful way of thinking about it.
Imagine you’re standing outside on a bright, sunny day and you don’t like the shape of your shadow. You wouldn’t start trying to move or reshape the shadow itself. You’d change your body shape, the object casting the shadow.
The shadow changes automatically when you lift your arm, turn your body or put something else next to you. I think culture works in exactly the same way.
Culture isn’t something we build directly. It emerges from the things created around it. It’s the visible outcome of hundreds of decisions about how an organisation is designed.
In 2014, a corporate diagnostic tool was developed by Chief Executive Women and Goldman Sachs. It focused on the structural levers leaders use to shape the culture shadow, either knowingly or unknowingly.
It identified four things. The first was what we say: the strategic messaging. The second was how we act: the behaviours we model. The third was what we prioritise: where the budget gets allocated. The fourth was what we measure: the rewards and incentive structures.
Those are the things that cast the shadow. They are how culture is created in an organisation.
Given all of that, I started to wonder whether, in trying to create a data-driven culture, we’ve confused cause and effect. We’ve been trying to change the shadow instead of changing what’s casting it.
What actually shapes culture
If culture isn’t something we can create directly, what does create it? I think it comes down to the environment people experience every single day.
The diagnostic tool I mentioned starts to touch on this. It’s the structures of the organisation, the incentive systems, the systems themselves, the behaviours people see around them and the people who sit above, beside and below them.
Those things shape behaviour. When behaviours are repeated, they become habits. Those habits ultimately create the culture you see around you: the thing you can’t necessarily touch or feel, but which is present throughout the organisation.
Think about what gets measured. If leaders say collaboration is important, but only measure individual performance, what do people optimise for? They optimise for their own performance rather than collaboration.
You might rely on people being good citizens and collaborating for the greater good, and that often happens. But, ultimately, things get optimised around what you’re measuring.
If leaders say quality matters, but reward speed above everything else, what ends up winning? Speed.
Then think about what gets rewarded. Who gets promoted? Who gets recognised? Who gets invited into the important conversations where decisions are made? People pay far more attention to those signals than they do to posters on the wall or statements about the culture the organisation says it wants.
It’s what actually happens that matters.
Think about priorities. Where does the leadership team spend its time? Where do teams spend their time? What gets funded? Which initiatives survive budget reviews? When there are budget cuts, what continues and what disappears? All of these things create the culture around you and communicate what really matters.
The tools people use every day matter too, along with the insights and behaviours they interact with. If systems are difficult, people create workarounds. If accessing data takes 20 minutes, people stop using it. If the easiest way to get something done doesn’t involve data, they won’t use data.
The same is true of decision making. If leaders regularly ask, ‘What does the evidence tell us?’, that’s great. You start having a data-driven conversation.
But if the highest-paid person’s opinion wins every time, people quickly learn how decisions are really made. That drives culture.
Recruitment also plays a huge role in creating the culture you want. Every person you hire either reinforces, changes or breaks the culture you have.
If you recruit curious people, you’ll have more curious people and that curiosity will grow. If you want an experimentation mindset and recruit people who are comfortable experimenting, experimentation becomes normal. If you recruit people who challenge respectfully because you want more constructive challenge and feedback, that becomes part of everyday life.
The things we do, the behaviours we expect and the experiences people have day to day are what create culture. This happens either proactively or reactively.
None of the things I’ve just talked about involves rolling out a culture programme, communications campaign or awareness initiative. But collectively, they create the culture that exists, knowingly or unknowingly.
More accurately, all of these things create the conditions from which culture naturally emerges.
Testing the theory with innovation
Let’s test this theory and see if it really works.
Imagine an organisation that wants to build an innovation culture. What have I seen happen? People talk about innovation. They run innovation workshops and away days where people get a chance to innovate. They nominate innovation champions, hold innovation weeks and put an ideas portal online or in the reception area.
Maybe a few posters appear, reminding people to think differently and try new things.
But none of those things actually makes people more innovative. What creates innovation is the environment they’re in.
Do people have time to experiment so they can innovate? Can they access budget to test new ideas and do new things? Do they have the psychological safety to try something and fail without damaging their career prospects or opportunities? Are they rewarded for learning, even when something doesn’t work out?
SpaceX has become a recognised example of this. When things go wrong, such as a failed rocket test, that failure is almost celebrated inside the four walls of the business.
The point isn’t to create failure, and failure isn’t necessarily something to celebrate. The point is experimentation, and experimentation inevitably includes some level of failure.
If everything fails, that’s incompetence. But if every failure is treated as incompetence, people simply stop experimenting. Innovation disappears, and the culture becomes one of not innovating, even though the organisation says it wants more innovation.
The culture simply reflects the environment people are in.
Creating the conditions for data-driven working
So, what does this mean for data and a data-driven culture?
When organisations tell me they want to create a data-driven culture and I start exploring what they’re thinking about, the response is fairly predictable. Similar things come up each time.
People talk about improving data literacy: getting people more comfortable with data, helping them understand insights, graphs and charts, and showing them how to make data-driven decisions.
They run awareness campaigns, put posters on walls, send communications and hold away days or team meetings to explain what being more data-driven means. They might train everyone on a new platform or appoint data champions to act as cheerleaders inside the organisation.
Don’t get me wrong. All of those things are great. They’re valuable, and they’re things that need doing.
But I don’t think they’re what creates a data-driven culture. They’re really just the tip of the iceberg.
Instead, I think we need to ask a different question. Based on everything I’ve talked about in terms of how culture is really created, the question is: how do we create the conditions in which using data becomes the natural way of working?
There are many things we can do to achieve that, none of which involves running a data culture programme.
It’s about starting with the outcomes you’re trying to achieve, identifying the decisions that matter most and focusing on how to improve those decisions in the most effective way, day in and day out.
How do we build products, solutions and technology that solve real problems people have and improve the way they work? How do we make data more accessible? How do we measure the impact of the things we’ve done? How do we celebrate teams that improve by using evidence to drive their decision making?
We then need to make that success visible.
We need to show people that using data and insights to improve strategic or operational performance has created success, or that it didn’t and we learned something along the way.
When people experience data helping them do their jobs better, make a better decision or genuinely improve the outcomes they’re trying to achieve, you don’t need to convince them that data matters.
They naturally become curious about it. They naturally want to do more of it. They naturally become data-driven.
That’s a very different way of thinking about culture from a top-down, blanket initiative designed to get the whole organisation behind something.
Culture changes through small, iterative, day-to-day and ongoing improvements in the way an organisation operates. We create the conditions, and a data-driven culture emerges from the changes we make.
Avoiding the same mistake with AI
I’m concerned that we might go the same way with AI and make the same mistake again.
I’m already hearing organisations talk about building an AI-first culture. And I know what’s going to happen. We’ll follow the same playbook: AI awareness campaigns, AI champions, AI training and AI communications.
All of those things are useful, but none of them answers the much more important question: how do we redesign work so AI genuinely helps people succeed?
How do we embed AI into decisions? How do we embed it into workflows? How do we embed it into day-to-day activities in ways that create value?
If we get that right, we’ll create an AI-driven culture, and that will largely take care of itself.
Stop trying to create culture directly
I think it’s time we stopped trying to create and force a data-driven culture.
Not because culture doesn’t matter. It absolutely does. It’s the right outcome and the thing we want to create.
But we’ve been treating it as a destination in itself rather than the consequence of the work we do.
From my own experience of building and running a business, I don’t think culture is something you create directly. It’s something that emerges through what you do.
It’s the shadow cast by the way our organisations are designed.
If we don’t like the shadow, we shouldn’t try to reshape the shadow itself. We have to change what’s casting it.
We have to create the right environment. We have to create the right incentives. We have to create the right tools. We have to create the right behaviours. We have to create the right conditions.
Perhaps the data-driven culture we’re looking for will naturally emerge from doing all of those things.
I’d love to hear your thinking on this. A data-driven culture is a great ambition, but I think people have often approached it in the reverse way to what we need.
We do a lot of work in this space, and I’ve seen time and again that approaching it this way works much better than the alternative I’ve described.
I’d love to hear about your experiences, so please do get in touch. I hope you enjoyed the episode, and I’ll catch you again next time.
