“We need to become more data-driven,” says the CEO. Six months later, decisions are still made on gut feeling and by whoever talks the loudest in the meeting. The problem is not the technology – it is the culture. Here is how Swedish companies actually succeed with the transition.
At Excel Department we have followed the journey of more than 40 organisations from intuition-based to data-driven decisions. Some succeeded within a year. Others struggled for three years without results. The difference was rarely about tools or data – it was about culture, leadership and processes.
This guide does not present theories from Harvard Business Review. It shows what actually works in Swedish mid-sized companies.
What data-driven decisions really mean
Data-driven decisions do not mean ignoring experience and intuition. They mean complementing them with verifiable facts.
It is the difference between: – “I think we should focus on segment X” → “Data shows that segment X has 40% higher profitability” – “It feels like sales are going well” → “Sales increased 12% compared with last month” – “We seem to have high costs” → “Administrative costs have risen from 18% to 24% in six months”
A real-world example: one of our clients in the service sector was about to expand to Gothenburg. The CEO was convinced based on “everyone says Gothenburg is growing”. We analysed their customer data. It showed that 78% of customers came from areas with a certain demographic profile. Gothenburg had a low concentration of that profile. Stockholm’s southern suburbs had a high concentration.
They expanded to the southern suburbs instead. The result: 40% faster growth than forecast. Data saved them from a costly misjudgement.
Three levels of data-drivenness
Most companies move through these steps:
Level 1: Reporting (where most start) – Data is used to explain what has happened – “Here are last month’s sales” – Historical focus – Reactive decision-making
Level 2: Analytical (where value emerges) – Data is used to understand why something happened – “Sales fell 15% in segment X because of Y” – A combination of history and understanding – More proactive decision-making
Level 3: Predictive (maturity stage) – Data is used to foresee what will happen – “Based on trends we expect 8% growth next quarter” – Future focus with forecasts – Proactive and optimising decisions
Most Swedish mid-sized companies are at level 1 or early level 2. That is perfectly fine – level 2 is where 80% of the value is created.
Read more about how to build the foundation: Business intelligence for mid-sized companies
Why most data culture initiatives fail
We have seen it dozens of times: the company invests in Power BI, hires a data analyst and sends people on training. A year later the tools are barely used and decisions are made as before.
Mistake 1: Believing tools create culture
“We’ll buy Power BI and become data-driven.”
No. Tools enable data-drivenness, but they do not create it. It is like buying a gym membership and expecting to get strong without training.
A manufacturing company we worked with bought expensive BI licences for everyone. Six months later they were used by three people. The problem: nobody had explained why data mattered or how it should be used in everyday work.
Mistake 2: Data is owned by IT or controlling
Data is treated as a technical or financial issue, not a business issue.
Consequence: Business people wait for “IT to fix the report” instead of exploring the data themselves. Decision-makers become passive consumers instead of active users.
Solution: Data is owned by the business. IT and finance provide support, but each department is responsible for its own data analysis.
Mistake 3: No connection to decisions
Beautiful dashboards are built but not used in actual decision situations.
One of our clients built 15 dashboards in the first year. Impressive. But when we asked “which decision was influenced by this dashboard?” they struggled to answer for 12 of them.
Principle: Only build dashboards that connect to concrete decisions.
Mistake 4: Data quality is ignored
“We have data, so we can make data-driven decisions.”
If data is inconsistent, incomplete or wrong, decisions become worse than gut feeling. Garbage in, garbage out.
A retail company had customer data in the CRM, the accounting system and Excel. The same customer could have three different customer numbers and four variations of the company name. Analyses based on this data were worthless.
Read more: Data quality in a BI strategy: How to build trust in your numbers
Mistake 5: Lacking data skills
Expecting everyone to be able to analyse data without training.
Reality: Interpreting data correctly requires training. Understanding when correlation is not causality. Seeing when a pattern is statistically significant versus random variation.
How to build a data-driven culture (that actually works)
Here is the process we have seen work in Swedish organisations.
Phase 1: Management must lead the way
No data culture is established bottom-up. It has to start at the top.
Concrete actions for management:
1. Base your own decisions on data – Start every board meeting by looking at the KPI dashboard – Require data support for major investment decisions – Ask “show me the data” when someone makes claims
2. Reward data-driven decisions – Give positive feedback when someone backs up a proposal with data – Question (kindly) decisions without data support – Share examples where data led to good decisions
3. Invest visibly in data infrastructure – Set aside budget for BI tools – Give people time for data training – Hire or develop data skills
A CEO we worked with started every monthly meeting by opening the Power BI dashboard and going through the key figures. In the first month people looked puzzled. After three months people came prepared with their own data insights. After six months it was standard.
Phase 2: Make data accessible
Data that is locked inside systems or held by individuals does not create data-drivenness.
Principles for data accessibility:
1. Self-service should be the standard Employees should be able to get answers to simple questions themselves without waiting for IT or controlling.
Example: Salespeople should be able to see their own figures, compare with colleagues and see trends – without requesting a report.
2. One source of truth When the same figure exists in three places with three different values, people stop trusting data.
Solution: Define master data. “Sales figures come from system X, full stop.”
3. Relevant to the role The CFO and the production manager need different data. Build views tailored to different roles.
A logistics company we worked with built role-based dashboards: – The CEO saw the whole business – Regional managers saw their region – Drivers saw their own deliveries and efficiency
Everyone had access to relevant data, and nobody drowned in irrelevant information.
Phase 3: Build data skills
Tools do not help if people cannot use them.
Training programmes that work:
For management (2-4 hours): – Why data-driven decision-making? – Which questions can data answer? – How do you interpret dashboards correctly? – Case: Decisions that improved with data
For managers (4-8 hours): – How do you ask the right questions of data? – Basic data analysis – How do you read Power BI dashboards? – How do you run data-driven team meetings?
For specialists/analysts (2-5 days): – A deep dive into Power BI or the tools you use – Data modelling and analysis – Visualisation principles – Hands-on with your own data
At Excel Department we tailor BI training for Swedish companies based on your maturity level and your tools.
Phase 4: Integrate data into processes
Data-drivenness becomes reality when data is built into how you work.
Concrete processes to change:
Monthly meetings: – Start with a KPI review from the dashboard – Everyone has seen the data before the meeting – The discussion focuses on “why” and “what do we do”
Investment decisions: – All proposals include a data section – A template for how data is presented – Decisions are documented with supporting data
Customer meetings: – Salespeople have access to customer data in real time – History, purchase patterns, profitability – Personal offers based on data
Product development: – Customer data drives which features are built – A/B testing of new functions – A continuous feedback loop
Phase 5: Measure and improve
How do you know whether you are becoming more data-driven? Measure it.
Indicators of data-drivenness: – % of decisions backed up with data (measure in meetings) – Number of active users of BI tools – Time from question to answer (is it getting faster?) – Decision quality (measure the outcome of decisions over time)
Quarterly review: – Which decisions did we make based on data? – Which decisions should have had more data support? – Which data tools are used? Which are not? – Which new data needs have arisen?
Practical examples from Swedish companies
Theory is good. Real examples are better.
Case 1: E-commerce company – From guesswork to precision
Before: The product manager chose which products to add based on “what feels trendy” and supplier recommendations.
Change: Implemented a dashboard with: – Sales per product category – Margin per product – Inventory turnover – Customer review data
After: Product decisions are now based on actual profitability and customer demand. Result: 22% higher margin on new products, 30% lower capital tied up in inventory.
Culture shift: Product meetings start with “what does the data say?” instead of “what do we think?”.
Case 2: Consulting firm – Data in the sales process
Before: Salespeople had no idea which customers were profitable. All customers were treated equally.
Change: Built a CRM integration with Power BI that showed: – Profitability per customer – Purchase history and patterns – Potential lifetime value
After: Salespeople prioritise high-value customers and can personalise offers. Result: 35% higher average deal value, 18% increased conversion.
Culture shift: From “chase every lead” to “focus where we create the most value”.
Case 3: Manufacturing company – Proactive problem handling
Before: Production problems were discovered when they had already become expensive. Reactive firefighting.
Change: Implemented a dashboard with real-time production data: – Machine downtime – Quality deviations – Throughput per line
After: Problems are identified before they escalate. Result: 40% less unplanned downtime, 15% fewer quality defects.
Culture shift: From reactive to proactive thinking. “The data warned us” became a standard phrase.
Next steps – Start your data culture journey
Becoming data-driven does not happen overnight. But you can start today.
Week 1: Map the current state
Questions to discuss in the management team: – Which important decisions have we made in the past month? – How many were backed up with data versus gut feeling? – Which data do we have access to but do not use? – What is stopping us from being more data-driven?
Weeks 2-3: Choose a pilot area
Do not try to change everything at once. Choose an area where: – Data is available – Decisions are made regularly – The impact is measurable – Champions exist who drive change
Examples: Monthly sales meetings, production meetings, board follow-up.
Month 2: Implement in the pilot area
- Build or improve a dashboard for the area
- Train those involved
- Change the meeting agenda to start with data
- Document decisions and outcomes
Months 3-6: Measure, learn and scale
- How did the decisions change?
- Did they get better (measure the outcome)?
- What worked? What did not?
- Which area should be tackled next?
Need help?
Excel Department helps Swedish companies build data-driven cultures. We do not start with tools – we start with processes and behaviours.
Our approach: 1. Workshop with the management team – identify where you are and where you want to go 2. Pilot project in one area – prove value quickly 3. Training adapted to roles – from CEO to analyst 4. Step-by-step scaling – one area at a time
Book a free workshop (2 hours) where we map your maturity and create a concrete action plan.
Summary
Becoming data-driven is more about culture than technology. Tools such as Power BI make it possible, but without organisational change they are just expensive letter openers.
Success factors: – Management that leads by example – Data available to everyone who needs it – Training adapted to different roles – Integration into existing processes – Continuous improvement and learning
Start small: – One area, one team, one decision – Show value quickly – Learn, adjust, scale
Swedish companies that succeed with data-drivenness do it step by step over 1-3 years. Those that fail try to change everything at once.
Which journey will you choose?
Related guides: – Business intelligence for mid-sized companies – The technical foundation for data-drivenness – KPI framework for Swedish companies – Define the right key figures – 5 Power BI dashboards every CFO needs – Concrete tools for decisions
Need help with this? Our Excel consultants can help you – the first consultation is free of charge.
Book a free call or phone 010-264 20 20.
