Thomas Edison and Steve Jobs are regarded as geniuses who magically attracted success. In truth, they were decision-makers who relied on evidence-based methods – on structured data rather than pure gut feeling or adherence to proven best practices.
Because what worked yesterday may be outdated tomorrow.
And relying on individual managers is not successful in the long term. What happens if the successful manager leaves the company or fails to achieve success?
The solution is evidence-based decisions in a data-driven company. With concrete steps, a corporate culture can be focused on a successful decision-making basis.
Evidence-based: Deciding with data
When a decision has to be made in a company, the opinion of the most senior person often prevails. This is referred to as a “HiPPO decision”: the Highest Paid Person’s Opinion. This model is widespread – unfortunately. Because in the long term, the result is an increasing risk of incorrect decisions that cannot be corrected.
“Evidence-based decisions are based on data and facts, not on the position of the person making the decision in the company.”
Peter Spörri, Senior Consultant & Partner @ panter.consulting
The alternative is to make decisions on the basis of transparent data. If implemented correctly, all the people involved in the company have the same information at their disposal and can make value-oriented decisions based on this information that are comprehensible to everyone.
Data-driven company
The first step towards evidence-based decisions is therefore the creation of a common database. To do this, all data from different operational systems such as CRM, production planning, control, etc. must be aggregated and cleansed – turning pure data into usable information.
Based on this information, experiments can be carried out in order to achieve the strategic goals of the business unit or the entire company in small steps.
Schematic process of a data science project
Constant experiments
Constant evaluation is crucial to the success of these experiments. To this end, it is important to ensure the measurability of each experiment from the outset. In other words, a hypothesis is needed that can be confirmed or falsified with measurable data.
Experiments are not a one-off event, but go through constant iterations. Here are the four steps of an experiment cycle:
- Formulate a hypothesis: Based on empirical values, an idea is formulated as to how the goal can be achieved. Including ways to measure the success of the experiment.
- Carry out an experiment: A change within the scope of the hypothesis is implemented and the results are measured using the defined criteria.
- Check results: Did the change improve the results or not?
- Adapt objectives and approach: The procedure is adapted and further developed using the new information from the evaluations. The goal is also scrutinized in this step and – if necessary – sharpened or changed.
Experiment loop for data-driven companies
What does an evidence-based corporate culture look like?
The basis of a data-driven company is easy access to shared data for everyone. Only when information and knowledge are available across structures can they unfold their value – hoarding data in departments and silos is the direct antagonist of any evidence-based culture.
Clear goals
Once the data is available, clear objectives are needed. If everything has priority, nothing has priority. Company management must therefore set KPIs, decide which data is important and always be prepared to question the status quo.
Managers’ own intuition can certainly be a good starting point for this point – if they answer the following question: “What facts and figures can I use to test my gut feeling and confirm or refute it?”
Sensible data architecture
In addition to the objectives, it must be clear which data should be collected, how, where and with which tools this should be done. The access logic must also be defined. This ensures that everyone who is allowed to view the data can do so, while data security remains guaranteed.
Access and knowledge
Employees must be specifically informed and trained: What data is actually available, why is this data important, where and how does it benefit processes and decisions? When management and staff actively use the new data architecture, new energy is released that was previously tied up in old processes.
“The new data and information means that in many cases, for example, the time-consuming and error-prone maintenance of complex spreadsheets is no longer necessary. This is a welcome relief for many employees and ensures leaner processes.”
Peter Spörri, Senior Consultant & Partner @ panter.consulting
4 Advantages of an evidence-based corporate culture
- The quality of decisions increases when they are made on a rational, data-supported basis rather than on the basis of gut feeling or the personal experience of individual employees.
- Many processes, for example in reporting and controlling, are faster, simpler and cheaper, as the data architecture is a more efficient alternative to old and complicated spreadsheets.
- Any progress can always be measured and verified. Deviations can be quickly identified and corrected.
- The company and the individual decision-makers can continue to develop.
Evidence-based corporate culture in practice
When companies decide to make better decisions, the question of “How?” remains. The four cornerstones of a successful corporate culture are:
1. guidance from the front
As always, company management leads by example. Only if the top management in the company makes credible decisions based on data can the same be demanded of employees. The management level starts with small pilot projects that demonstrate the value of the company and serve as an internal case study.
2. central responsibility
If each department thinks and decides for itself, no company-wide systems can be created. One possibility is to create a position whose task is to create an overarching, automatable process for data in the company. This helps to avoid local individual solutions and data islands and to appoint a contact person for the transformation.
3. lay the foundations
Employees must not only be trained, but also recruited: Pilot projects limited to individual areas that work on real, local problems and demonstrate the benefits of evidence-based decisions can help here. This can then be prepared as a case study, which is used internally as a training basis and lighthouse project.
4. live a culture of error
A major challenge is the culture of mistakes and cooperation within the company. One reason why data is often hoarded locally is because the analysis gives others power over their own department – what if others discover errors or a lack of efficiency? Fear of transparency can cause the entire project to fail. It is therefore important to demonstrate a positive error culture within the company.
What does that mean? Mistakes are allowed, and it is never a question of putting someone down or looking for “culprits” afterwards. Instead, each error is followed by a precise analysis of how it could have happened and how the processes can be improved to prevent it in future.
Once the four cornerstones have been established, the new culture can be spread step by step to other areas and departments until the entire company is working together on the basis of evidence.
Example case: Building a “Science Team” at AIG
In 2012, Peter Hancock, CEO of the insurance giant AIG, saw a great opportunity. Evidence-based decisions had the potential to generate major competitive advantages in an industry still characterized by individual expertise and assessments. This is how the “Science Team” came into being. Within just under two years, it had grown to over 100 people – from mathematicians and data scientists to behavioral economists and psychologists.
The new team created the data basis for evidence-based decisions throughout the organization, but also focuses on identifying new business models, knowledge transfer and change management towards a data-driven company.
Successes came quickly, for example through the development of new and more powerful tools for the automatic detection of insurance fraud. At the same time, tests were underway from the outset that could change the business model in the long term, e.g. through the use of new technologies and sensors in damage calculation.
Key factors for the success of the “Science Team” were:
- Focus on issues that are important for the company.
- Supporting the change and learning process throughout the company with tools and workshops.
- Involvement of early adopters in the company and with direct contacts in order to make successes quickly visible.
- Building a portfolio of initiatives that is constantly being further developed instead of focusing on individual approaches.
- Rapid and iterative development steps for the initiatives, as implementation creates the greatest possible learning effect. Even if the first version is not yet perfect.
- Planning of initiatives for short, medium and long-term impact with varying degrees of potential impact on the core business.
Learning from evidence: advice for success
A transformation process towards becoming a data-driven company can be a major challenge. Management consultancies and coaches therefore offer ways to guide you through the process. There is a wide range on offer – but which techniques and services really help?
At Panter, we rely on a handful of techniques that we have successfully tested in practice: We start with a “walkthrough” in which we guide you through the entire process, followed by an individual “Paint Point & Solutions” workshop in which we analyze your challenges together.
We also offer technical reviews – such as an audit of your existing data architecture – and targeted workshops on overarching topics such as “Design with Data”.
Every consultation is individual. Try it out for yourself: Contact us free of charge and without obligation.
