Today, most management teams can reference better data than they were able to five years ago. But the decisions were different five years ago in most of the organizations in the world. The instincts, politics, and habits – not the data – too often drive those decisions. The data looks perfect if they were the only change, but a tab on no one’s browser is where they remain.
From gut-feeling to data-interpreter
There is a type of data literacy training that shows managers how to create pivot tables, but we are not referring to that kind of training here.
What we are really talking about is transforming leaders so that they can use data to challenge their own perceptions, rather than blindly trust them. For instance, a manager with fifteen years in a given sector will likely notice patterns that would be invisible in a dataset. But they tend to overlook whether the data supports the pattern they expect to see.
We know that confirmation bias is most prevalent at the management level: more so even than the analyst level, because senior people get less pushback for ignoring the data that doesn’t fit. Upskilling must target this flaw. This entails giving managers real data and testing it against their immediate assumptions. It means giving them data that doesn’t support their hypotheses and looking for ways forward there – not just scenarios where the numbers logically follow their expectations.
Part of this is also KPI alignment. Many management teams are tracking metrics that made sense 5 – 10 years ago, but weren’t updated to match the current reality. This is also an upskilling challenge: not only do managers need to be guided on whether the data is telling them what they’re reading, but they also need to be pushed to ask if they are even reading the right thing in the first place.
Data storytelling as a leadership skill
Understanding technical details is essential, but not enough. A manager who can read a predictive model but can’t communicate what it means to their team hasn’t bridged the gap that matters.
Data storytelling – taking complex outputs and translating them into decisions a team can act on – is where most upskilling programs fall short. They teach the software and stop there. But the ability to frame a trend, explain why a forecast changed, or connect a drop in one metric to an upstream cause is what separates a manager who uses data from one who leads with it.
This is where soft skills and data skills intersect. Emotional intelligence isn’t separate from analytics fluency. It’s what makes analytics fluency useful to anyone other than the person doing the analysis.
The risks of skipping the strategic layer
When organisations invest in AI and analytics infrastructure without developing the leadership layer above it, the results tend to disappoint. An ai leadership failure often doesn’t look like a dramatic collapse – it looks like an expensive tool that nobody trusts, or a recommendation engine that managers override every time because they don’t understand how it works.
Algorithmic transparency matters here. Leaders don’t need to understand the mathematics behind a model, but they do need to understand what inputs drove a recommendation and where that model’s assumptions break down. Blind compliance with AI outputs isn’t data-driven leadership. It’s just a different kind of gut feeling – one that’s harder to challenge because it has a confidence interval attached.
Only 20% of analytic insights actually deliver business outcomes, largely because of gaps in data literacy and strategic alignment at the decision-maker level. That figure hasn’t shifted dramatically because companies keep investing in tools and skipping the curriculum that would make those tools useful.
The curriculum itself needs to move beyond tool training. Which questions should the data be answering? What does a good data governance process actually protect against? What’s the difference between a vanity metric and a leading indicator? These are strategic questions, and they belong in a management upskilling programme.
Upskilling without undermining autonomy
The challenge isn’t just teaching managers to trust data – it’s helping them trust it without becoming dependent on it. Over-reliance on analytics can be just as damaging as ignoring them entirely. When leaders defer every decision to what the model says, they abdicate the judgment their role requires.
Effective upskilling acknowledges this tension. It teaches managers when to lean on data and when to step back from it. Sometimes the numbers reveal a blind spot. Sometimes they reflect a measurement problem, not a business problem. The skill is knowing which situation you’re in.
This means building comfort with ambiguity. Data rarely gives a single clear answer, and the best training programmes don’t pretend otherwise. They teach leaders to triangulate: data plus context plus experience.
Building a culture where data isn’t surveillance
A barrier to data-driven management is how teams perceive the data’s purpose: supporting them or monitoring them?
Psychological safety and data culture go hand in hand. If managers primarily use metrics to look for someone to blame when numbers are off, their teams will have no trust in the data process. This is where the application of the “fail-fast” principle of agile, using data to quickly identify if a project is struggling and reallocate resources, becomes fraught with the danger of being seen as a performance management exercise.
Teaching managers to be a diagnostic user of data rather than a surveillance user is, in part, a change management discussion. How do they frame the purpose of looking at the data? How do they react when the numbers indicate a problem? Is the team of the opinion that the data works for them or against them?
Likewise, the area contains discussions around ethics. Algorithmic bias, data privacy, and questions around how customer data is applied are not merely compliance topics. These are discussions where if you don’t have a decision-making framework, defaulting on the wrong side will eventually be the outcome of good people. Nobody ever made a wrong decision because nobody showed them how to make the right one.
The management teams that succeed in data-driven environments won’t be the ones with the best tools. They’ll be the ones where leadership can hold the data, question it, explain it, and act on it without losing the human judgment that no dashboard can replace.
