We don’t have a problem with data. We have a problem with decision-making
In a world where data seems to be the most valuable resource, many organisations fall into the trap of hoarding information without making any progress on decision-making. Why does this happen?
When I worked as a data scientist, I believed that if the analysis was good, the decision would almost follow naturally. If the model was well-built, the dashboard was clear and the metrics were well-defined, the next step seemed obvious: someone would look at it and make a better decision.
Spoiler: it doesn’t always happen. Over the years, as I got closer to operations, customer intelligence, the business and management teams, I began to notice a recurring pattern: we had pretty decent data, analysis and charts, but the decision was still on the table. Then came the questions: “Can we look at this by segment?”, “What if we cross-reference it with another variable?”, “Do we have historical data?”, “Can we wait another week?”
Sometimes those questions were necessary and provided context. But other times they masked something more human: the fear of getting it wrong. That’s when I realised that the challenge for data teams wasn’t just technical, but also about getting an organisation to trust the information, the judgement and the people enough to make decisions.
How often do we ask for data when, in reality, we’re struggling to make a decision?
In any organisation, it is common practice to present an analysis that makes sense and points in a clear direction. However, on many occasions, new requests or doubts arise following the presentation, which can delay the decision.
Sometimes it is simply curiosity, but often it is unease, a fear of making a mistake, or the need to justify a difficult decision. We ask for more data not to understand better, but to feel more confident, and that rarely solves the problem. In fact, it often only delays it.
For example, I have often encountered the following situation: in a company, the analytics team prepares all the reports and dashboards needed to launch a new product. However, after every presentation, new requests arise: “one more dimension”, “another analysis”, “what if…?”. Decision-making gets put off for weeks, until eventually a competitor beats them to market. Over-analysis and the quest for absolute certainty end up costing them the opportunity. Does this sound familiar?
This experience is not unique; it happens frequently in teams that seek the perfect decision rather than moving forward with sufficient information.
When everything is metrics, nothing is a priority
This is particularly noticeable in environments with many teams and initiatives. Each area optimises its own, each team has its metrics, and each dashboard tells a story. And in the midst of all that, someone has to decide.
The result is usually this: a lot of analysis, little direction. When there are too many open threads, decisions become scattered, and when everything is important, nothing is. I’ve seen this in projects where the analysis was flawless, but the fundamental question—what is the priority?—remained unanswered.
Making decisions based on data isn’t just about looking at a dashboard
Another idea we’ve oversimplified is what it means to ‘make decisions based on data’. It’s often reduced to looking at a graph, picking the figure that suits us best, and acting on it. But the reality is more complex.
Making decisions based on data involves understanding the context, connecting information from different teams, discussing, assessing risks and, above all, accepting that the data doesn’t decide for you. Data informs, but the decision remains a human one and is often uncomfortable.
Analysing data is a technical process, but making decisions involves taking risks and accepting consequences. Many organisations invest in better analytical tools, believing that this will solve their problems, when what is really lacking is leadership and alignment around decision-making. Without a culture that values learning and action, data remains untapped potential.
And this is where the usual cultural barriers come into play.
Common cultural barriers
Organisational culture is often the biggest obstacle to informed decision-making. Common barriers include resistance to change, fear of making mistakes, working in silos, and a lack of trust in teams’ judgement. Overcoming these challenges requires leadership, training, and opportunities for dialogue in which mistakes are seen as learning opportunities.
For example, many companies punish mistakes rather than analysing them and learning from them, which breeds risk aversion and stifles innovation. Fostering a growth mindset and celebrating learning, not just successes, is key to moving forward.
But even in organisations that are working on their culture, misconceptions that hinder action still persist. Here are some of the most common myths about data-driven decision-making:
Common myths about data-driven decision-making
- ‘We need all the data to make a decision.’ In reality, you will never have all the information. Waiting for perfection often leads to inaction.
- ‘Data eliminates risk.’ Data helps to reduce uncertainty, but it never eliminates it entirely. Risk is inherent in any decision.
- ‘The best technology will solve our decision-making problems.’ Without a culture that promotes action, even the best tool will prove insufficient.
And just as we begin to understand what data-driven decision-making entails, a new promise appears on the horizon: artificial intelligence.
AI-first: between excitement and vertigo
When I hear “we have to be AI-first”, it triggers two very clear reactions in me. On the one hand, excitement, because it implies that the organisation is starting to take data seriously. On the other hand, vertigo, because it is often interpreted as “let’s cram AI into everything”, “let’s automate and see what happens”. And that is dangerous.
AI is the visible part, the one that makes the headlines, but there is much more behind it: data management, data governance, data quality, processes, culture and people. That part isn’t particularly appealing, but it is what determines whether everything else works. AI is the result, not the starting point, and we often try to start at the end.
My fear is not that companies will use too much AI, but that they won’t realise what it takes to use it properly. Because it’s not just a technological issue, but also an organisational, cultural and leadership one. And that requires time, discussions and difficult decisions.
AI has brought about a fundamental change. Before, getting answers was expensive. Now it isn’t. You can generate analyses, hypotheses or proposals in seconds. That is a huge advantage, but it also changes the rules of the game: answers lose value and questions gain importance. Asking good questions is no trivial matter: it requires judgement, an understanding of the business, and knowing what matters and what doesn’t.
As Tom Davenport, an expert in business analytics, states: “Companies don’t fail because of a lack of data, but because of a lack of courage to make decisions based on it.” Or as Satya Nadella, CEO of Microsoft, puts it: “The true value of data is unlocked when combined with human intuition and experience.”
Before introducing AI into all processes, there are three key questions a company should ask itself:
- Are we clear about the problem? It is not enough to say “we want to use AI”; the important thing is to identify what we want to improve and which decisions we want to make better.
- Is our database ready? Without reliable data, AI solves nothing. It merely scales up errors.
- Do we have the culture to sustain it? Because this isn’t just about tools. It’s about how we work, make decisions and trust one another. And that can’t be implemented with software.
Another common mistake is wanting to carry out major transformations right from the start: huge platforms, ambitious projects, sky-high expectations. But in reality, it usually works better this way: test things out bit by bit, learn, adjust and consolidate. And consolidation isn’t just technical; it’s cultural too.
This is particularly relevant in the technical sphere. There’s something that’s common to many technical teams: we think that being technically very good is enough. It isn’t. We also need to communicate well, explain the value of what we do, propose initiatives and connect with the business. We often expect the organisation to give us space, but part of that space also needs to be created.
None of this is really about AI. It’s about decisions. And in an environment where answers are increasingly accessible, the difference will be made by those who know how to ask the right questions and who are capable of deciding when they have enough information, even if it isn’t perfect. In my experience, what truly transforms teams and companies isn’t the latest tool or the most sophisticated dashboard, but the collective ability to ask questions, prioritise and decide. The rest—the technology, the models, the AI—is merely the means. The aim is always to move forward, even if that sometimes means making decisions whilst still in doubt. Because waiting for the perfect answer is, at heart, just another way of never deciding.
Have you ever wondered how many times a decision in your organisation has been put on hold whilst waiting for ‘that missing piece of data’? How does this affect the pace of innovation and team motivation? What if, instead of seeking absolute certainty, you opted for informed action and continuous learning?
To move towards a culture of data-driven decision-making, organisations can start by promoting training in critical thinking, creating safe spaces to debate and make mistakes, and celebrating learning, not just successes. Furthermore, it is essential that leaders lead by example, trusting their own judgement and that of their teams, even when the information is not perfect.
The next time you face a difficult decision in your organisation, ask yourself: do I really need more data, or is it time to trust my judgement and move forward? Because deciding, even in the face of uncertainty, is what truly drives change.
If you are a leader, challenge your team to make decisions based on the information available. If you are an analyst, ask what the real purpose of your work is. If you are part of a technical team, seek to connect your knowledge with the impact on the business.
You’re not alone: even the most innovative companies face these barriers. What makes the difference is your ability to move forward, learn and make decisions.
What do I do when nobody makes a decision?
In my experience, the turning point isn’t about doing more analysis, but about changing the conversation.
I’ve learnt to put projects on hold when there’s no clear decision behind them. Before building anything, I ask:
“What decision are we going to make about this?”
If there’s no answer, there’s no project.
I also make sure to bring business and data teams together around the same table. Because the value isn’t in the most complex model, but in someone understanding what it means and acting on it. A good data team isn’t the one that produces the best analyses, but the one that gets decisions made.
Data doesn’t replace decisions. It makes them inevitable.
The problem isn’t a lack of data. It’s a lack of direction.
An organisation that doesn’t make decisions accumulates analyses and ends up holding the business back.
The value of data isn’t in understanding. It’s in acting.
The difference is made by those capable of deciding based on the available data. Waiting for the perfect answer is not analysis; it is avoiding a decision. Organisations that fail to make decisions don’t make more mistakes: they simply make less progress.
This article draws on my personal experience of grappling with the paradox between data and decision-making in technological and business environments. Within the Manfred community, there is the talent to transform organisations, provided we dare to make decisions and learn along the way. Have you come across these barriers in your work? Share your ideas or experiences in the comments or on LinkedIn/Twitter.
Want to know a little more about who wrote this article? 👇
Leader in data and analytics, specialising in transforming organisations through the real value of data and artificial intelligence. She has led teams and projects in Data Analytics, Data Engineering and Data Science across sectors such as digital, retail, consultancy and marketplaces, driving evidence-based decision-making.
Trained in mathematics, statistics and applied economics, with a specialisation in artificial intelligence and business analytics, she combines business acumen, technical expertise and leadership in digital transformation.