“Data is the new black.”
“Data is the currency of the future.”
“Data is the most valuable asset your organization has.”
“Is your organization data-driven?”
If you have heard any of these statements or questions, then you have probably wrestled with the issue of creating a data-driven culture.
Having a data-driven culture means that data is the fundamental building block of your team. It means that every team member has a data-driven mindset. It means that every single decision maker uses data as their main evaluation asset. It means that every project uses, generates and pivots on data. It means that your team is constantly leveraging data as a strategic asset.
But how do you get there?
Creating a data-driven culture depends on cultivating a mindset of experimentation, having the right infrastructure in place and developing the skills to interpret the signal from the data, while ignoring the noise in it.
For each team member, there are four steps that they must take on the path to becoming data-driven.
Step 1: Do You Know Where You’re Going to?
The first step is to know the questions that you are trying to answer with data.
With unlimited resources, you and your team could monitor and store every single bit of data that you generate, that you use for your mission, or that you think may be relevant in executing your strategic objectives. Unfortunately, this is a very expensive proposition and you are not guaranteed that the data you have in your possession will be helpful.
Similar to other activities that must function under constraints and within resource budgets, defining the end state is extremely important.
The questions you want to answer with data provide the needed focus for data-driven success. Do you want to make a process more efficient? Do you want to decrease the time taken to successfully complete a transaction? Do you want to increase the number of customers that you can serve?
In your context, the questions that you want to ask of the data determine the data that needs to be collected. Yes, this is an obvious statement. However, it needs to be explicitly stated.
Know what you want to get answered from the data, then figure out the specific data items that needs to be collected and stored.
Be firm in clearly defining the data items, the units used and the meanings of each data point. Standardization and consistency will be essential when it comes to implementation and scaling your data infrastructure.
Step 2: Do You Know Whom You’re Going to?
Once you know how you will interrogate the data, which helps you define what you should collect, you now need to understand the audience that this data will be presented to. Will the decision maker be yourself, a data scientist, an executive or your grandmother? This knowledge will help you determine both the transformations that need to happen to your source data and the correct visualization to use for maximum impact.
Knowing your intended audience also forces you start thinking about the actions that you want them to take when you present the data to them. When they see the visualized data, should they take out steps from a process? Should they increase the number of staff members on a particular task? Should they start more closely monitoring a specific business area?
Step 3: Do You Know How You’re Getting There?
The hard part is mostly over. Now it is time to design and implement your ETL (Extract-Transform-Load) pipeline. Essentially, this is where you create the process and supporting mechanisms (technical or otherwise) that allow you to get data from the desired data source, cleanse and massage it into the right form with the right semantics, and then store it in the data management system of your choosing.
Start small. Pilot your ETL with your simplest use case. When you have it working satisfactorily, expand the scope of scenarios that your ETL pipeline can handle until it covers all of your needs.
Step 4: Do They Know What They’re Looking at?
In the end, the presentation and the interpretation of the data is what decision makers interact with and what facilitates the creation of a data-driven organization. A choropleth map with multiple data variables on it may mean nothing to your boss if they don’t intuitively understand both the visualization type and the message that is trying to be conveyed.
For this reason, it is critical to use information you gathered in Step 2 to create the right visualization for your audience, which should lead them to interpret the data in the right way and make the right decisions.
Conclusion
A word to the wise. Data is a reflection of the world around us. Unfortunately, the world around us is flawed and has deep systemic problems. So, be careful in your quest to be data-driven. Be careful in your exclusive use and trust of data.
It is better to be informed by data rather than only data-driven.
