Why tools alone do not make a company data-driven, and the habits, roles and recurring meetings that do, with seven steps to start.
In short
A data-driven organisation is one where decisions are routinely based on trusted data rather than opinion or habit. It comes from behaviour, not software: leaders who ask for evidence, a few clear roles that own data and metrics, and recurring meetings where the same numbers are reviewed and acted on. Start in one business area, then spread what works.
Most companies that call themselves data-driven (datadriven) have invested in a data platform and dashboards. Far fewer have changed how decisions are made in the meetings that matter. This guide from Devies, a Gothenburg-based product and engineering consultancy, is for CEOs, CFOs and data leads who want a data-driven organisation in practice, and who suspect the next tool will not get them there.
What is a data-driven organisation?
A data-driven organisation is a company where people at every level routinely use trusted data to make, explain and review decisions. It does not mean that data decides everything; it means that opinions are tested against evidence, and that the evidence is easy to find and believe.
Microsoft's Fabric adoption roadmap gives a practical description of data culture: a set of behaviours and norms that encourage informed decision-making by more people, based on analytics and trusted data rather than opinion, hunches or undocumented tribal knowledge. It adds a useful warning: data culture is what you do, not what you say, and it is separate from governance, which is the set of rules.
That distinction explains why many digital transformation programmes stall. They deliver the platform and the rules but leave the behaviour unchanged, so the new dashboards sit beside the old spreadsheets and the meetings run as before.
Which habits make an organisation data-driven?
The habits that matter are small, visible and repeated by leaders. They signal what is valued far more than any strategy document.
- Leaders ask "what does the data say?" before discussing a proposal, and accept "we do not know yet" as a valid answer that leads to a measurement.
- Every important number has one definition and one source. Meetings do not open with a debate about whose revenue figure is right.
- Decisions are written down with the expected effect, so they can be checked later. A decision without a follow-up date is an opinion.
- People can find data themselves and know who to ask when they cannot.
- Bad news in the data is welcomed. If a dashboard that shows a problem gets its owner criticised, people learn to hide problems.
- Effort to improve data is recognised. The Microsoft roadmap notes that leaders show what matters through how they praise and reward people who take initiative.
Which roles does a data-driven organisation need?
A data-driven organisation needs a few clearly named roles, most of them part-time responsibilities for people who already work in the business. A mid-sized company rarely needs a large central data team to get started.
Role
Responsibility
Typically held by
Executive sponsor
Responsibility
Sets the vision, removes obstacles, asks for data in leadership meetings
Typically held by
CEO, CFO or COO
Metric owner
Responsibility
Owns the definition and target of a key metric and acts on it
Typically held by
Head of the relevant function
Data owner
Responsibility
Decides who may use a data domain and what quality is required
Typically held by
Business manager for customers, products, finance
Data steward
Responsibility
Maintains definitions, quality rules and documentation
Typically held by
Experienced specialist in the business
Data platform team
Responsibility
Builds and runs pipelines, models and the platform
Typically held by
Internal team, partner or both
Analysts and champions
Responsibility
Help colleagues use data and spread good practice
Typically held by
Power users in each department
Microsoft's roadmap also describes a Center of Excellence: a small group of technical and business experts who support others, set standards and mentor users. In a mid-sized company this can be three or four people who meet regularly, rather than a department. Our guide to data governance without the bureaucracy covers owners and stewards in more detail.
Which rituals keep data in daily decisions?
Rituals are the recurring meetings and routines where data is reviewed and acted on. They are what turn a dashboard into a habit.
- Weekly operational review: each team looks at a small set of leading indicators, notes deviations and agrees actions with owners and dates.
- Monthly business review: leadership reviews the main KPIs against targets, using the same dashboard every month. Our guide to KPI dashboards leaders actually use shows how to build it.
- Decision log review: once a quarter, look back at major decisions and their expected effects. Did the data support them, and was the outcome as predicted?
- Data quality review: data owners review quality scores and open issues for their domain.
- Community session: analysts and champions share reports, techniques and problems once a month.
Keep the number of rituals small and protect them. One well-run monthly review does more for a data culture than a dozen dashboards nobody discusses.
Seven steps to start building a data-driven organisation
The most reliable path is to start where a real business problem gives the work urgency, prove it there and then spread it. Microsoft's roadmap describes the same pattern: data culture often takes root in a smaller area before being replicated across the organisation.
- Pick one business area with a pressing question, such as margin erosion in one product line or delivery delays in one region.
- Name the sponsor and metric owners for that area, and agree the three to five metrics that answer the question.
- Define each metric once and build it on trusted data, even if the first version is simple.
- Start one ritual: a weekly or monthly review where those metrics are discussed and actions recorded.
- Train for interpretation, not only tools. Data literacy means reading charts critically, separating correlation from causation and spotting outliers.
- Measure the effect after three months: decisions taken, actions completed, time saved on reconciliation.
- Spread deliberately to the next area, reusing definitions, roles and the meeting format.
Tie this to your overall direction; a one-page data strategy keeps the areas pulling the same way.
The engineering habits of the data team matter too. At Hedin Mobility Group, Devies coached the BI team in code review, logging and reusable code while working alongside them on real changes, and the team enforces a common reporting approach across acquired companies so the group works from a single source of truth.
Keep reading
From our work
Client Case: Stena LineStena Line modernised its Azure data platform - improving Power BI reliability, reducing pipeline runtimes and scaling with medallion architecture.Read the caseDevies
Want a second opinion?
Devies, a Gothenburg-based product and engineering consultancy, works on questions like this with Nordic product and IT teams. Tell us where you are and we will suggest a sensible first step.
Frequently asked questions
What does it mean to be data-driven?
Being data-driven means that decisions are routinely made, explained and reviewed using trusted data rather than opinion or habit alone. It does not mean data replaces judgement. Leaders still decide, but they ask for evidence, use agreed definitions and check afterwards whether decisions had the expected effect. The visible sign is meetings that start from the same numbers and end with actions.
How long does it take to become a data-driven organisation?
A single business area can show a clear change in three to six months if it has a sponsor, a few agreed metrics and a recurring review. Spreading that across a mid-sized company usually takes one to three years, because it depends on changing habits, roles and trust in data rather than only installing technology. There is no finish line; data culture needs continuous attention.
Is a data-driven organisation the same as digital transformation?
No, but they are related. Digital transformation is the broader change of processes, products and business models through digital technology. Becoming data-driven is one part of it: using the data those digital processes create to make better decisions. Many transformation programmes deliver new systems without changing how decisions are made, which is why data culture deserves its own attention.
Do we need a chief data officer to become data-driven?
Not necessarily. A mid-sized company can make strong progress with an executive sponsor from the existing leadership team, named owners for key metrics and data domains, and a small group that sets standards and supports users. A dedicated chief data officer becomes relevant when data is central to the business model, regulation is heavy or the data estate is large and complex.
