Key takeaways
- Data analysis is the beating heart of business intelligence: without it BI is an empty shell without content.
- Investing in a data warehouse makes no sense if no concrete, continuous improvement actions result from it.
- When a figure deviates, the only right action is to look for the root cause instead of resorting to panic measures.
- Affinity with data grows by having every employee practise with data sets in pairs for about four hours a week.
- Preconception is the great pitfall: always combine several data sources and actively look for counter-evidence.
Use data analysis as a competitive weapon
Why do organizations find it so hard to analyze their data consistently? Thomas Abramse: “You can declare BI dead and replace it with AI, but then you quickly get bogged down in a discussion about definitions that does not help the customer.” Investing in BI only makes sense if you are able to transform data into usable information and use new insights for actions that strengthen your business. Data analysis plays the leading role in this. “Investing two hundred thousand euros in a data warehouse makes no sense if no concrete continuous improvement actions ultimately result from it.”
Data analysis as the beating heart
Data analysis is the beating heart of business intelligence. It has always been that way. Take it out, and all life force and energy drain away at once.
Daan van Beek sees people staring at their dashboard without understanding or being able to explain the figures behind the charts. “Suppose you are a mortgage lender and see that the inflow of new applications has been 10% lower for two months than in the same period last year. Which actions do you take? Under pressure organizations quickly opt for panic football. Flexible staff are dismissed and ill-considered marketing campaigns rushed out. That can be counterproductive and set a self-fulfilling prophecy in motion.”
Look for the root cause
“The only right action is to look for the root cause of the shrinking applications. Something may just as well have gone wrong in the recent upgrade of the website, so that conversion falters. You need data analysis to find out exactly where the problem lies and which knobs you can turn.” A second example: when infections rise on a pandemic dashboard, the tendency is to reach for crude measures. “With a sound root cause analysis you can work in a much more fine-grained way. Without it everyone is tarred with the same brush: vulnerable groups are insufficiently protected while the majority suffers unnecessary damage.”
Why data analysis is skipped
Abramse: “Professionals prefer to stay in their comfort zone. Data can be quite confronting. An analysis may show that the old familiar way of working is now outdated. In addition, some of today’s managers are simply not data literate enough.” Van Beek sees, besides practical causes such as lack of time, emotional blockages: lack of interest, fear of the truth and resistance to change. Too many managers underestimate the importance of data analysis, lack the analytical view and the skills, and so do not give professionals the time and space to analyze data thoroughly.
Organizations also expect miracles from a hired data analyst, while an analyst by definition stands on the sidelines and often has insufficient domain and process knowledge. “You only cultivate affinity with data by having every employee practise with data sets in pairs for about four hours a week. Teams keep each other sharp, and you steadily build knowledge. It also prevents superfluous work and ensures that the right priorities are set.”
Get started yourself
Should organizations call in specialists first or experiment themselves? Abramse: “As with so many changes, getting started with data yourself and clocking up flying hours is necessary to learn to deal with data. That applies to every employee.” Van Beek: “But advisers can make the difference in a number of areas: training and coaching employees, independent of suppliers, and giving objective advice on a solid data infrastructure. In addition, mapping the strategy, KPIs, processes and decisions in a practical way lets you carry out data analyses in a much more targeted manner. That is perhaps where the greatest gain lies.”
What makes a good data analyst?
Someone who has occasionally put together a dashboard is not one, says Abramse. “A good data analyst regularly provides relevant new insights. Those insights are based on reliable data, can be applied quickly and are aimed at improving or accelerating processes or increasing profit or quality. A good data scientist exposes what is invisible to everyone but very important to know: patterns, causal relationships and correlations.” According to Van Beek, weak performers, whom you encounter in every profession, are partly responsible for the idea that BI yields too little. “That is why we must train BI consultants, BI managers and business managers much better and feed them with inspiring ideas.”
Avoid the pitfall of preconception
How do you carry out a sound, accountable data analysis? Abramse: “Put simply, you link different data sources to the different types of business questions and look for the analysis technique that works best. In reality it is a continuous process of converging and diverging along different axes. Compare it with the shape of a cut diamond. It is important not to work towards the solution too quickly. People often already have a solution in their head and then look for the data to match. They only seek confirmation. Preconception is a major pitfall.”
Van Beek: “It starts with selecting good-quality data. Always combine several data sources. Be aware of possible biases and be prepared to set aside your own convictions. I have benefited a great deal from Karl Popper’s falsification theory: always look for counter-evidence.” There are two kinds of analysis: targeted analysis, in which you look for something concrete, and free data exploration. Abramse: “Data analysis can also be very misleading, so always ask how it was carried out. It is a creative process that you do not do on a Tuesday afternoon between your regular work. Always look for the question behind the business question.”
The techniques every professional should master
Van Beek: “Properly applied, data analysis is a very powerful weapon in the competitive struggle. Intelligent, data-driven organizations analyze their data daily, so that they can continuously improve their processes. The rest just muddles along.” Abramse: “Because there is still a lot of ignorance, you quickly gain a lead if you invest in data analysis and in the data literacy of your people.” Tools no longer make the real difference, he feels: time and energy are better invested in data literacy.
Five analysis techniques from the interview
There are dozens of techniques; these are the favourites of both interviewees.
Answers the question
Typical use
Ad hoc analysis
What is going on right now?
Fast gains, done thoroughly
Root cause analysis
Why does this figure deviate?
Finding the knob to turn
Correlation analysis
What is related to what?
Standard in every toolbox
Churn analysis
Why do customers leave?
Subscription and SaaS companies
Zero data analysis
Who did not buy from us, and why?
Action plan for account managers
Abramse puts ad hoc analysis at number one: “It is a must. You can make the most gains relatively quickly, without concessions to soundness.” Root cause analysis he calls the godfather of all techniques. Van Beek is charmed by zero data analysis, based on data that is not in your database: “Find out why customers did not buy a well-selling product from you while they did buy it elsewhere. Compare those segments and you have an almost ready-made action plan for your account managers.” Within artificial intelligence he encounters very powerful techniques too, such as ensemble learning, random forests and neural networks. “With AI you can automate away a large part of the analysis function. You still need data-aware employees, make no mistake. But the heavy task of making every employee data literate then feels a lot lighter.”
How Diamond Decisions puts analysis at the centre
The cut diamond from the interview is the Diamond Decisions method: diverge and converge until the question behind the question is clear. In business modeling we map strategy, KPIs, processes and decisions, in data analytics we choose the technique that fits the question, and we train decision makers to look for counter-evidence.
Conclusion
Data warehouses, data lakes and dashboards are means. The value arises in the analysis: understanding why a figure deviates, finding the root cause and acting on it. If you remove or neglect data analysis, BI amounts to nothing. If you invest in it and in the data literacy of your people, you quickly gain a lead.
Talk to the author
Daan van BeekDiamond Decisions
Discuss your question with Daan
Tell us where decisions in your teams get stuck today. In a no-obligation conversation we explore how everyone can make better decisions with data every day.
Schedule an introduction →Step 4 of our approach
Decision Makers
From data to better decisions every day, in every team.
Explore Decision Makers →Frequently asked questions
01
Why is BI pointless without data analysis?
Because data warehouses and dashboards only show figures. Only analysis explains why a figure deviates and which action helps. Without analysis no improvement actions follow and the investment yields nothing.
02
What is the difference between targeted data analysis and data exploration?
In targeted analysis you look for the answer to a concrete question, such as why applications are falling. In free data exploration you search the data without a fixed question for patterns and relationships that may be relevant.
03
How do you avoid bias in data analysis?
Select good-quality data, always combine several sources, do not work towards a solution you already have in mind and actively look for counter-evidence, in the spirit of Karl Popper’s falsification.
04
Which analysis technique should an SME start with?
With ad hoc analysis and root cause analysis: when a figure deviates, find out why with the data you have. Add correlation analysis and, if you have subscriptions or repeat customers, churn analysis.
Further reading
About the author
About Diamond Decisions
Diamond Decisions helps organizations become better by enabling better decisions. We don’t start with data, dashboards or AI, but with the decision that needs to improve, and work back from there to KPIs, data and technology. We do that in four steps: inspiration and alignment, business modeling, analytics and data platform and data-driven decision makers. That way BI and AI don’t remain an IT project, but deliver value every day.


