In a previous post, we introduced a process and methodology for making data-informed decisions. The process defines the steps that should be systematically followed. The methodology outlines what needs to be done in each step of the process, including how and why to do it. In this post, we will add in the final piece of the puzzle: the skills that are needed to allow organizations to follow the process and apply the methodology accurately and successfully.
Data-informed decision making is a team sport. It is unrealistic to expect any decision maker to have all of the skills below. Some of the skills are important for everyone to have, but other skills are specialized for specific parts of the process. Below is a matrix that shows the top ten skills required within an organization to follow the Data-informed decision making process.
When an organization has an analytical question they need to answer, the key is understanding what data has information that will aid in answering that question. Once the data and information are identified, it needs to be extracted. With some analytical questions, that may just involve taking data as-is from an excel file. In more complex questions and situations, it involves extracting information from big data systems and technologies.
Once the data is extracted, it needs to be transformed and standardized to be made analytics ready. Research show that up to 80 percent of the time spent making data-informed decisions is on tasks related to cleaning, standardizing, and organizing data. This shows the absolute importance for organizations to have the right skills to be able to clean, transform, profile, tag, catalog, and standardize data.
Not everyone in an organization requires data science skills to make data-informed decisions, but basic math skills including a fundamental understanding of data are essential for everyone involved in the process. This includes a fundamental understanding of data, including types of data (categorical vs continuous), attributes of data, and various data aggregations and distributions. With this skill, everyone can understand and leverage descriptive analytics, which is a key step in the data-informed decision making process. This can range from someone responsible for building and maintaining a measurement framework which includes critical key performance indicators, to decision makers who need to apply meaning to the information they are seeing.
Foundational statistics skills are vital for an organization that wants to make data-informed decisions. The individuals who are ultimately making the decision do not necessarily need these skills, but they need someone working with them on the decision who can provide foundational statistics. This requires an understanding of probability and correlation, simple regression, as well as inferential statistics to ensure things like sample sizes are created properly.
While data science is not a single skill by itself, they are grouped together here to encompass everything that an organization needs to do with machine learning and artificial intelligence, including predictive and prescriptive analytics. With the vast amount of data available to organizations today, machine learning skills are essential to turn the data into insights to make data-informed decisions.
According to Edward Deming, 94% of problems in business are systems driven and only 6% are people driven. Systems thinking helps decision makers understand why people behave like they do. It is a way of looking at an organization (and the world) as a set of systems that all connect in some way. When viewing the organization this way, identifying causes vs symptoms is easier as decision makers can consider how each part interacts with each other.
Decision makers need to understand not just the data available to them, but they also need to know the meaning of that data and how it should be applied to the business. Insights are not found in data but in how humans assign meaning to data. Part of the data-informed decision making process is the ability to think critically about the data and recognize both the complexity of the decisions and the possibility of multiple valid positions. Decision makers need to understand possible limitations of the data presented and they need to be aware of, as well as mitigate against, any cognitive bias they may have. Very rarely will decision makers have a complete set of data at their disposal. Decision makers need to accept this, avoid analysis paralysis, deal with the uncertainty, and make the best decision they can with the data that is available to them.
During many phases of the decision making process, people are exposed to information. Whether it be requirements, insights from the analysis, or feedback on the decision during the assessment phase. It is human nature to apply meaning to that information based off our own cultural and personal perspectives. From that, people may draw conclusions. In reality, those conclusions may be based on what people think someone else said, as opposed to what they really said. This is where active listening, combined with critical thinking, is vital.
When going through the data-informed decision making process, one of the key influences on the process is the organizational culture. The ability of that culture to support the process depends on the quality of relationships within the process, which also depends on the quality of conversations. This means that relationship building is a critical skill required, from gathering requirements from the business, to communicating out to all the stakeholders, to gathering feedback on the decision after it is made.
Communicating decisions out to stakeholders is one of the most important skills an organization can have. Resources need to be able to translate data into a story that emotionally moves and motivates their stakeholders and helps them understand the decisions.