You hear it more and more: the future belongs to organisations that use their own data to sell more. That means analysing the information you can get at, processing it and finding new insights into how to grow. Once large volumes of data arrive in real time from several sources it gets harder, but computing power has come within reach of companies without big budgets.
Plenty of organisations discover only after paying for sophisticated solutions that technology is just one part of the recipe. Below are the steps to take before you buy data technology: the changes without which no company becomes truly data-driven.
We wrote this article in 2022 and updated it in October 2026. Since then, generative AI has entered almost every conversation about data, and the steps below matter even more: an AI model is only as good as the data it is given. Without clean, accessible data and a clear owner, AI just automates mistakes faster.
The figures come from Eurostat and a Gartner survey of 1,203 data management leaders.
01Why customer data matters
The pandemic moved a lot of shopping online and hugely favoured brands with a strong digital presence and a well-run direct-to-consumer offer. For those that weren't ready, 2020 was a missed opportunity.
Demand hasn't dropped since. According to Eurostat, 78% of EU internet users bought online in 2025, and Romania saw the biggest rise of the past ten years. Tomorrow's winners will be the companies that get really good at using their own data on customers and how they shop.
But being data-driven takes more than hiring someone to look after it and waiting for miracles. It takes a change of attitude across the company and a shared understanding of what success through data looks like. Here are the main steps.
02A data strategy before technology
The first step is a clear strategy for collecting and managing information. Surprisingly often, companies buy the latest technical solutions and only then realise they've badly overestimated their ability to manage data. And managing isn't only collecting and storing: it takes a team that properly understands technology and data science. The plan should cover at least:
- An owner with a mandate. One person who is the undisputed owner of the data, with the authority and a clear mandate from leadership to make changes across the organisation.
- A mixed team. A member of management who sets up and coordinates a team with people from marketing, sales, data, legal and IT.
- Clear limits on investment. How much you are willing to spend now to earn more later, with the long-term benefits in mind.
The winning attitude treats data as a company asset, with investment in line with the profit it could bring. And the data owner's role has to become a process the whole organisation adopts, so data stays a profitable asset whoever holds the role and whatever programme uses it.
03A complete view of the customer
One of the best-known sayings in computing is GIGO, "Garbage In, Garbage Out": feed the system bad information and the result is just as useless. The biggest barriers are poor data quality and how hard it is to get at. Quick access matters, because success often depends on how fast decisions are made, and good decisions depend on the quality of the information behind them.
Almost every organisation runs into two obstacles at first:
- Silos. Departments that keep their information to themselves.
- Gatekeepers. Colleagues who block access to essential data, because for years information has been treated as a department's property or a perk of the job.
The hardest part is explaining why access is needed. The best argument is the customer: every colleague has to see why a single view of the customer matters, one that pulls together everything we know about them and every interaction we've had with them.
That view has to be kept up to date. The customer isn't carved in stone, and what we know about them can change from one day to the next. So projects built on customer data have to move faster than the customer's world does.
Subscribe to the newsletter.
Like how we think? Get an article like this one twice a month.
04Data is worth nothing without action
Customer data is like gold in the ground: you need a mine and miners. New insights come from analysing the context, and to solve a business problem with data you need a multidisciplinary team. The most useful insights come from putting two different data sets side by side, as in this simple example:
- The situation
- You click a Google result and land on a product that's no longer made.
- The cause
- Marketing didn't coordinate with sales, so SEO kept promoting a product that had been dropped.
- The minimum fix
- A process or system that stops promoting products that no longer exist.
- The data-driven fix
- Marketing and sales sit at the same table and rethink the whole offer together.
It's a simple example, but it shows how far a data-driven approach can reach: access to information for everyone, and colleagues encouraged to fix problems they didn't know existed.
05In short
These are only the basic steps. As long as the company hasn't put the fundamentals in place, the technology will only make money for the people selling it. Getting there takes experiments and tests, and every informed decision creates new data, new insights and new ways to grow.
Letting data drive your decisions means, above all, a culture willing to experiment until it finds the seam of gold in the mountain of information every company sits on, sometimes without knowing it. It isn't an easy road, but the payoff can be worth it.
