Building a Data-Driven Organization
As data science and big data came into our vocabulary, more and more business leaders want to leverage the power of data as well as equip managers with the tools offered by data to help grow their businesses. However, when it comes to transforming an organization that previously was not used to working with data, it can be quite a challenge. There can be conflicting interests, lack of knowledge transfer, switching costs, and many other reasons to resist a change in culture. As a data scientist who has helped introduce the data culture into startups, there are a few ways to go about building a data-driven organization.
What is a data-driven organization?
First, let’s be a bit precise with what we’re talking about. A data-driven organization is one that understands and (tries to) fully leverage the power of data across all parts of the organization. Even though some organizations have massive amounts of data, they rather rely on their intuition and past experiences to streamline business processes. Other organizations have successfully weaved a data-driven culture into the very fabric of their business processes, as it gives them a much clearer view of everything as well as a competitive advantage.
Why build a data-driven organization?
‘Just because others are doing it, why should we also be doing it?’ is a common sentiment among business leaders. Here are some of my reasons to build a data-driven organization.
Decreasing risk, increasing success
Operating a business inevitably comes with a lot of uncertainty. Uncertainty leads to a lot of risks that the business has to manage over long periods of time. Business leaders also have to be more careful with decision-making while taking into consideration the risks associated with each decision. Data is the antidote to uncertainty, as it gives more clarity on virtually everything, ranging from competitors, customers, and even internal operational efficiency. Once uncertainty is minimalized, business leaders are more likely to make the “right decision” and achieve the success they need.
A stronger understanding of customers
One of the key success factors of a successful business is a solid understanding of who the customer is, what the customer wants/needs, and why they want/need it. Data collected on customers and user behavior not only can help you better understand your customers better but ultimately iterate a product in a direction that is able to align even more with your customer’s needs, thus creating more value for them, maximizing customer retention and loyalty.
A backup for internal politics
One of the biggest challenges that organizations face is conflicting interests between departments and managers. Managers with a lot of experience will have a tendency to make decisions based on intuition, as they have succeeded many times in the past. Although there isn’t anything necessarily wrong with making decisions with intuition, it is often difficult to argue with someone who has more experience than you in their particular field. After all, the reason why they are in that position is due to their experience and expertise. Data is often a good way to help validate or invalidate intuition-based decisions, and it is often difficult to argue with hard evidence and facts, as opposed to raw intuition. This is also why big corporations often hire management consultants to help validate their assumptions; it’s often just to settle disputes on major decisions, from an objective, data-backed perspective.
For more concrete reasons to build a data-driven organization, check out this other post that discusses research done by PwC on data-driven organizations. Although there are good reasons to be data-driven, there are also reasons to not become over-reliant on data, some obvious ones include:
- Bureaucratic and inefficient
- Neglecting the emotional and subjective nature of consumer-oriented products
- Overlooking external factors that cannot be seen in data, such as socio-cultural trends, technological change, innovation, etc
Although being data-driven does not magically solve all problems, it does, however, have many of the advantages mentioned above. It’s also important to consider its shortcomings, as most things tend to possess.
How to build a data-driven organization
As much as business leaders would like to build a data-driven culture, they often don’t know where to start. Here are a few suggestions on how to get started.
Make data accessible and easy to understand
A common challenge when it comes to cultivating a data culture isn’t that people are unwilling to use data, it’s that data isn’t accessible to them. Especially at a big company where data is sparse yet minimal data engineering has been done, data is often inaccessible to non-technical personnel, which ironically, are often the people that need to access data the most. Even when transactional data is dumped into a data warehouse, non-technical personnel have trouble understanding ambiguous column names and data formats. The solution to this problem is to first build up a data architecture by understanding the business requirements of the end-user; in this case, would be managers or decision-makers. Then work with data architects to design data schemas corresponding to the requirements and finally set up data pipelines to funnel data into the data warehouse or data lake. Once an analytics data storage solution is in place, the next step is to make that data accessible through some data tool, such as Tableau, Power BI, or any off-the-shelf tool with a GUI that requires no coding or scripting, as the assumption is that most end-users don’t know how to write code or SQL. For an open-source tool, I would also recommend Metabase.
Assign data evangelists
Once the data is ready to go, business leaders need to begin spreading the word as well as provide guidance to people as they transition into using data as part of their day-to-day work. The most effective evangelists of a data culture usually come from the top of the organization, as they can move the needle and implement initiatives when it comes to making a shift in business processes. They can also work with data-savvy employees, such as data analysts, or data scientists to help educate and mentor others in understanding or leveraging data for their own jobs. Business leaders could even go as far as assigning a dedicated data scientist or data analyst in every team as a residential data mentor. Though it needs to be clear to each team that the assigned data personnel isn’t the only person who deals with data, but rather every person is held accountable for using the data tools provided to them.
Design a culture around data-backed decisions
Once all parts of an organization begin embracing the data culture, it is crucial to implement solutions for the culture to persist. The most obvious low-hanging fruit is partnering with data engineers to automate key metrics and routine analyses once each team has settled on a certain direction. This eases the amount of unnecessary work people have to do and helps them focus on managing and executing other crucial tasks. Another way to help employees transition is to gradually lower the capacity data personnel spend on data requests from non-data personnel, as this will force them to not rely on data personnel to get critical analyses done. The last solution is a bit of a tough-love solution, which involves a soft requirement of including data analysis or research for every proposal or presentation. This will remind employees of the importance of the data culture and the commitment the company has made to become more data-driven.
Conclusion
Building a data-driven culture is not easy, as most people are generally resistant to change. However, when people are able to see the fruits of the cultural change, as well as the change in efficient decision-making and getting things done, it becomes obvious that the transition was a good idea from the start. A persistent reminder of the overall purpose of the culture will ultimately sway people to your side, and you’ll be on track to becoming a data-driven organization.
