How to Build a Data-driven Culture
VIDEO: How to Build a Data-driven Culture
The power of analytics now extends into every corner of a business. For product profitability, analytics is used to maintain healthy margins.
Over in marketing, there is a great deal of analytics to determine customer behaviour and preferences. How do we encourage them to buy and what features will convince them to buy?
For producers and sellers of physical products, we have the field of inventory and supply chain analytics. The more items you carry, the more complex inventory management becomes.
One of the biggest challenges as we progress into all of these analytical areas is the ability to instill a data-driven culture. With it being such a difficult endeavour, why would we wish to do this?
An MIT professor reported that companies that embrace a data-driven decision-making culture have output and productivity that is 5-6% higher than those that don’t.
An analyst at Forrester was quoted as saying that “For a typical Fortune 1000 company, just a 10% increase in data accessibility will result in more than $65M additional net income.”
We start to see that being a data-driven company can yield both financial and productivity benefits.
However, 72% of respondents to a survey by NewVantage stated their companies had not yet forged a data-driven culture. If you have been trying to establish a data-driven culture and haven’t quite gotten there yet, rest assured, you are not alone.
Building a data-driven culture takes time and I can say that it is a journey of several years to educate and change the mindset of many business professionals.
The Four Components of a DaTa-driven Culture
We need 4 things to become a data-driven culture.
A data-driven culture needs data. Makes sense, right?
We need the tools by which we can analyze our data.
We need the skill sets to be able to use the tools to analyze our data.
We need a mindset shift to get people thinking about and using data for practical decision-making.
Which of these do you think is the hardest to achieve?
Component 1: Data
Let’s talk about data. We need the right data and by that, I mean we need the data which will allow us to answer the business questions we have in mind.
We need a data quality level appropriate for the analysis we wish to run. If we only need a directional idea about some aspect of an analysis, then we can accept a lower level of data quality. If we need precision in our analysis, then we need a high level of data accuracy and data cleanliness.
We need what data people call, “a single source of truth.” In cases where the same type of information exists in more than one system, we need to agree with our fellow analysts in other functional areas which system should be used. It is not unusual for these systems to be accurate but slightly out of sync with each other depending on when one system replicates data to another.
All in all, this section of establishing a data-driven culture is reasonably easy.
Additional Data Considerations
There are a few additional considerations on the data section of this discussion. With data access comes great responsibility. There needs to be specific guidelines and policies regarding data governance and data privacy.
Who can access the data?
Who can extract data from the system?
How will data be shared while still preventing sensitive information from reaching people it shouldn’t?
Data Regulations
For data protection and privacy, there are existing regulations and there are new regulations being formed. These regulations aim to balance protecting data privacy while also providing transparency to customers and employees as to what you plan to do with their data.
In the EU, there is the GDPR. This provides employees the right to be removed from an HR system once they leave the company. It also gives you the right to be deleted from an email list. If you have ever unsubscribed from an email list, that’s not the same as being deleted.
The state of California put similar rights in place in 2020. I think we can expect to see that expand into other states. As of October 2021, Canada is drafting its own regulations which will be similar to the GDPR in that you will have the right to be removed from your employer’s HR system if you request it. Even if you don’t request it, your employee record will likely be scheduled for removal in about 7 years.
Component 2: Tools
Next, we need tools… and we need those tools to be appropriate for the type of analyses we wish to conduct and, for practical reasons, we need those tools to be available within our budget constraints.
As an example of choosing an appropriate tool, if I’m going to run an analysis once and it is statistical in nature, I may choose to use SPSS, Mintab or Excel. If it is an analysis that I know I will have to run repeatedly, I may choose to code an R script to make conducting the analysis much faster in the future. Of course, you do have to invest the time up front to code a script.
When we speak of budget constraints, you may have noticed that software tools have moved away from letting you buy the software once and toward charging you an annual fee to use it. For example, you may have purchased Minitab in the past and have been using the same version ever since. You will find that this and other statistical tools now charge an annual fee.
This trend has increased the popularity of tools like R and Rstudio which are free. In fact, I’ve seen entire analytics teams in global companies use nothing but R for this reason.
Component 3: Skill Sets
Once we have data and the appropriate tools to conduct an analysis, we need people with the skill sets to know how to analyse data and how to interpret an analysis within context. By context, I mean that if you conduct an analysis for a specific business area, then you need a certain amount of knowledge about that business area to be able to understand what your analysis is telling you. Often, analysts do not have this knowledge especially when they’re new to a company so teaming up with people in that business area to interpret the analysis together is highly recommended.
An additional skill set needed here to help encourage a data-driven culture is the ability to explain an analysis in multiple languages. Here, I’m not talking about English or French, German or any other language. I’m saying that you need to know how to speak to highly numerical people like Ops and Finance but still be able to explain your analysis in non-technical words for functions like HR who are not used to mathematical conversations.
Planning for Efficiencies
You will need one more skill in this section and that’s the ability to plan for efficiencies. As your company begins to embrace data, as an analyst, the number of requests you receive for data and analysis will ramp up quickly. If you don’t plan for efficiencies such as creating automation scripts or building self-service tools, you will quickly become overwhelmed. Plan for efficiencies today to free up resources for tomorrow.
Component 4: Mindset Shift
Finally, we have the most difficult quadrant and that’s changing the mindset of people in your company. This quadrant will take the longest and it will take several years… but it is possible.
If you read articles online about creating a data-driven culture, you’ll see that it always says to start with executive support. That will certainly speed the journey, but in reality, most people will not have that support. You can still succeed from the bottom up like I did.
To succeed, you need to realize that people can’t envision what data can do for them if they’ve never seen an example of it. You need to take people on an educational journey, even if that’s a few people at a time.
Once people see examples of analyses you’ve conducted that generated value (revenue generation, cost savings or increased productivity), then they come up with new ideas of what can be done to help them in their own business area.
It takes patience and the willingness to teach the basics, like data literacy.
What is a trend?
How are conclusions from a chart distorted by poorly designed graphs?
What’s the margin of error mean on a survey or poll result?
These are fundamental skills that need to be taught to all business professionals.
To start spreading a data-driven culture within your company, you need an intentional plan. Perhaps you form a community of analysts throughout your company. Maybe schedule 30-minute “lunch and learns” to share successful analyses you’ve conducted and the value it provided to your company.
Building a data-driven culture takes time, but it can be done.
Resources
To help you on your journey, I’ll share links to several resources.
● Downloadable templates and books.
● And of course, there’s a variety of educational videos on my YouTube channel.