Data Culture - The Invisible Force Behind Data Product Adoption
This is Part 3 in a four-part series exploring why data products fail to achieve adoption, and what to do about it. In this edition, we're focusing on data culture.
You can have the most sophisticated data product. A beautifully designed dashboard, a well-governed data model, and an intuitive self-service experience. And yet, if the culture isn’t right, people won’t use it.
I’ve watched this play out more times than I’d like to admit. A data team invests months into building something genuinely valuable; something that answers real questions, built on reliable data and then the adoption never takes off. The product exists. People just don’t use it.
When we investigate why, the conversation usually turns to training gaps, change management, or stakeholder engagement. These all matter. But underneath most adoption failures, there’s something more fundamental: the culture of the organisation doesn’t actually reward those making decisions with data. And no data product, however well-built, can overcome that on its own.
Culture is the invisible force that either enables adoption or quietly kills it.
The HiPPO in the Room
There’s a pattern that will be familiar to anyone who has worked in data for any length of time: the data product gets built, the insights are sound, the recommendations are clear, and then the most senior person in the room says something that starts with “my instinct is...” and the analysis gets set aside.
This is HiPPO culture: decisions driven by the Highest Paid Person’s Opinion, and it is one of the most reliable predictors of poor data adoption.
It rarely announces itself. But you’ll recognise it from the signals:
Analysts produce rigorous work that gets acknowledged in a meeting and then quietly ignored
Teams learn, over time, which leaders are actually open to using data and which have decided to “trust their gut”.
People stop opening dashboards before making a decision because they’ve learned that the decision doesn’t flow from there, making that action pointless.
The damaging part is how repeated HiPPO behaviour affects adoption over time. If people consistently experience that using a data product doesn’t influence outcomes, the rational response is to stop using it. Why build the habit around a tool that doesn’t change anything?
Sustainable adoption requires people to believe that data genuinely adds value and helps to drive decisions. Without that belief, even the most well-designed product becomes as useful as a chocolate teapot.
Culture is the Adoption Problem Nobody Wants to Talk About
When a data product isn’t being adopted, the first instinct is usually to look at the product and diagnose potential problems:
Is it too slow?
Too complex?
Does it answer the right questions?
These are worth examining. But I’ve seen products with genuine usability issues and poor designs get adopted enthusiastically, and well-designed products that are left untouched because the underlying culture either supported data-driven decision-making or didn’t.
The reason culture often gets overlooked as an adoption driver is that it’s harder to pinpoint and fix than a filter or a dashboard load time. You can’t raise a ticket for it. There’s no magic button for delivering a culture change.
If you’re serious about adoption and data products that actually influence decisions, not just usage metrics, then culture is where you have to look.
A culture that supports data adoption has a few specific characteristics that are worth naming, because “data-driven culture” has been repeated so often that it’s become almost meaningless:
Decisions are connected to questions before they’re connected to products. People ask, “What do we need to decide, and what data would help?” before they consider which dashboard to open. This orientation, question-first, tool-second, is a hallmark of cultures where data products actually get used.
Uncertainty is made visible, not hidden. Where data products are genuinely embedded in decision-making, people are comfortable saying “the data is incomplete here” or “this is our best read of performance, but we’re not certain.” Cultures that demand false confidence from their data discourage people from engaging with it honestly.
Senior leaders use data products publicly. When leadership references a specific metric, cites a trend from a report, or visibly changes a plan based on something they saw in a dashboard, people notice. It communicates that these tools are for decisions, not just for the analytics team. Leaders lead by example.
Empowering People to Act on What They Find
One dimension of data product adoption that gets underappreciated is agency. Not just: can people access the product? But: do they feel empowered and trusted to act on what it shows them?
There’s a meaningful difference between an organisation that provides data products and one that trusts people to use them.
I’ve worked with teams where access wasn’t the problem. The products existed, the self-service capability was in place, and the data was reasonably reliable. But decision-making authority remained tightly centralised. Every insight had to travel up the hierarchy before anything changed. The people closest to the problem, with the most context, were looking at accurate data and weren’t empowered to act on it.
When that’s the dynamic, adoption suffers, not because the product is poor, but because the feedback loop is broken. People stop engaging with tools that generate insight they can’t use.
Genuine empowerment requires a few things to be in place:
Explicit decision rights. When it’s unclear who has the authority to act on an insight, data prompts a conversation but rarely prompts a change. Clarity on who owns which decisions is a prerequisite for adoption.
Training that goes beyond tool proficiency. Knowing how to navigate a dashboard is different from knowing how to use it to make a business decision. Capability programmes that stop at “here’s how to use the product” leave out the most important part: how to use data to make decisions.
Psychological safety to act and be wrong. If using a data product to make a decision and getting it wrong results in blame, people will stop. They’ll wait for certainty that never arrives, or escalate rather than decide. Adoption thrives in environments where acting on evidence is respected, even when the outcome isn’t perfect.
Are People Excited About Working With Data?
Here’s a revealing question I recommend asking your teams: “Is anyone here genuinely excited about working with data?”
Not “do they see value in working with data?”, as most people can articulate that. But excited. The kind of engagement where someone opens a dashboard because they want to, not because a meeting starts in fifteen minutes.
Enthusiasm is easy to dismiss as a personality trait, but I think it’s a leading indicator. Organisations where data products achieve strong, sustained adoption almost always have visible pockets of genuine enthusiasm; people who talk about what they found, who share interesting data insights with colleagues, who start conversations with “I was looking at the data and...” That energy is contagious. And so is the absence of it.
The conditions that cultivate enthusiasm are worth being deliberate about:
Celebrate the insight, not just the outcome. When a data product contributes to a good decision, make that story visible. Not just “we hit our target”, but “we hit our target because we spotted this trend early and acted on it.” Connect the tool to the impact.
Give people data that speaks to their world. Company-level aggregates are abstract. Data that reflects someone’s own customers, their own team, and their own work is much more personal. Personal data is engaging in a way that enterprise reporting rarely is.
Create space for exploration. Not every interaction with a data product needs to answer a specific business question. Some of the most valuable adoption moments come from giving people permission to browse, experiment, and discover. Showcases, peer demos, and discovery sessions create the kind of curiosity that turns occasional users into habitual ones.
Build a community around the products. Where I’ve seen the strongest adoption, there’s usually a visible community around it; people sharing what they’ve found, discussing what the data means and helping others. Shared enthusiasm is what converts occasional users into consistent ones.
Practical Levers for Building a Data-Positive Culture
Diagnosing a culture that doesn’t support adoption is straightforward. The harder question is: what do you actually do about it?
The following suggestions aren’t linear. They’re levers you may consider pulling, depending on where your organisation is right now.
1. Start with a culture audit, not an assumption
Before designing any adoption initiative, understand what you’re actually working with. Survey your teams. Run focus groups. Ask directly:
Do you trust the data in our products?
Do you feel confident acting on data insights?
Have you ever seen a data product genuinely change a decision here?
The answers will tell you whether you have a trust problem, an access problem, a skills problem, or an authority problem. These each require very different responses. For example, I’ve seen teams try to solve a trust problem with better dashboards. It doesn’t work, and it’s an expensive way to find that out.
You may find that different business domains have different problems. You won’t necessarily be facing the same challenge across the entire business.
2. Find and amplify your data champions
Every organisation has them. These are people who are already enthusiastic about using data, doing great work quietly, and who their colleagues already turn to with questions. Find them. Give them visibility, resources, rewards and a platform.
Data champions are your most credible adoption advocates because they’re peers, not central teams or external voices. When someone sees a colleague like them using a data product to make a better call, it lands differently than any top-down communication. Formalise this through a champions program or community of practice and treat it as a core adoption mechanism, not a nice-to-have.
3. Make data literacy a genuine core competency
Data literacy programmes focused purely on tool proficiency miss the most critical skill: knowing what questions to ask before you open the product. In fact, I would argue that tool proficiency should be taught outside of data literacy programmes.
The most effective data literacy programmes I’ve seen combine technical capability with analytical thinking, including how to interpret the data presented to you, how to frame a business question, how to interrogate your own assumptions, and how to know when the data is telling you something surprising, or potentially misleading. Build this into onboarding, career development frameworks, and team expectations so that it’s a continuous standard and not a one-off session.
4. Reward the behaviour, not just the outcome
Culture follows incentives. If recognition goes to people who confidently hit targets but not to people who surface uncomfortable truths or change course based on evidence, you’ll get a culture of performance, not learning.
Actively celebrate moments where a data product changed a plan, where someone flagged something that prevented a bad decision, or where an unexpected finding led somewhere valuable. Make these stories as visible. Over time, this will reshape what people believe is valued.
6. Give leadership a specific role to play
Culture doesn’t change from the bottom up. It cascades from the top down. However, statements like “leadership needs to support this” are too vague to be actionable.
Be specific: ask your leaders to reference dashboards in meetings (preferably live and not via screenshots in slide decks), to publicly credit an insight when it shapes a decision, to ask “what does the data say?” before offering their own view, and to explain their reasoning when they override analysis. These are small, concrete behaviours. But they signal to the entire organisation what good looks like. People are watching more carefully than leaders tend to realise.
7. Measure the culture, not just usage
Adoption metrics such as views, active users, and query volumes help to identify whether people are accessing your products. They don’t tell you whether those products are influencing anything.
Build in regular qualitative checkpoints such as pulse surveys on data confidence and trust, retrospectives on decision quality, and post-project reviews that ask “was data used to drive outcomes here?” If you’re only measuring platform activity, you’re measuring habit formation at best. What you actually want to know is whether the culture is shifting.
A Note on Patience
Cultural change is a slow process. That’s not what anyone running an adoption programme wants to hear, but it’s true and pretending otherwise leads teams to declare success before the change has actually taken hold.
Culture shifts through consistent signals over time: what gets rewarded, what gets ignored, what leaders do in unscripted moments. A single training programme or webinar won’t shift it. A product redesign won’t either. What moves it is the accumulation of small moments, including decisions that visibly draw on data, conversations that model curiosity over certainty, leaders who ask “what does the data tell us?” and actually wait for the answer.
The organisations I’vheree seen achieve genuine adoption and thriving data cultures treat it as a long game. They’re not running a campaign. They’re building habits, one meeting, one decision and one data product at a time.
Up Next
We’ve covered technology, process, and culture. In the final part of this series, I’ll look at the fourth piece of the adoption puzzle: People. Who your data community actually is, how to grow it, and why the most important investment you can make in adoption isn’t in your products, it’s in the individuals who use them.
What data culture challenges are you navigating right now? Drop a comment or reply directly.
Great post. Culture is huge. Everything could be done right but if there is no data driven culture then the impact of any data product from models to dashboards is low. Anecdotally speaking though I've seen that the data culture problem exists more in older organizations as compared to start ups. Would love to hear your experience on that.