ABSTRACT: Meeting the business where it is creates a heavy load for data teams and slows down delivery. Time for the business to meet the data where it is.

"Meet the business where they are." If you're on the data team, that's what you're expected to do to empower stakeholders with data. If the organization is not able to leverage its data effectively, if the business doesn't use the lakehouse or dashboards that you created, then you've supposedly failed to meet them where they are. But how far should you go to meet the business? And shouldn’t the business be expected to try meeting the data where it is by learning about it?

In my last blog, I wrote about the duties of data citizenship in a data democracy. I asserted that for data democratization to work, data citizens must have a basic level of understanding and knowledge of data. In this article we’ll explore what happens when data citizens don’t have the requisite knowledge to participate in a data democracy. We’ll look at how that lack of knowledge creates a heavy load for data teams to carry, slows down solution delivery, and contributes to the persistent gap between data ambitions and results.

Data teams are expected to make data understandable for any group they’re working with

Meeting the business where they are puts the onus on the data team to figure out how to work with colleagues that vary in data proficiency. They must accommodate stakeholders who lack a basic understanding of data, while also working with advanced analysts who have sophisticated requirements. Data teams are expected to make data understandable to any group they’re working with based on their needs. To do this, the data team must learn a lot about a given business unit, such as:  

• Business unit goals, strategies, processes, pain points and opportunities 

• How data is created by and used in business processes

• How data issues cause operational pain and missed opportunities, and how it can alleviate pain and enable business units to recognize and seize opportunities 

• The source systems that capture and use data

• Gaps in data quality, such as incompleteness, inconsistency, inaccuracy, late arriving data, and duplicate data

Ideally, these facts are uncovered through a collaborative process with each domain-based business team. They should know a lot about these subjects and be able to explain to data teams how they use data to run their part of the business. Unfortunately, that’s often not the case. Without timely business input, the data team has to spend time learning the business domain. There are many techniques for revealing hard-to-surface requirements, such as prototyping and iterative development. But the less business teams know about their data and what they need from it, the more time it takes the data team to arrive at a solution.  

Stakeholders expect quick delivery from the data team no matter how much they contribute to the end solution. This dynamic plays out to a different extent with every business unit that needs help with data. As a result, the data team’s backlog grows faster than they can satisfy requests, resulting in dissatisfaction and distrust from business unit stakeholders. 

Data teams carry a heavy load on behalf of their business colleagues

With the business dissatisfied and desperately looking for solutions, data software and service vendors target them with promises that the business can solve their data problems on their own. All they need are the self-service features the vendor’s tool offers, such as low-code-no-code data pipelines and drag-and-drop dashboard building. They often make it sound so easy that the business believes that they can get value from data without any additional knowledge or effort.

Vendors can make data sound so easy that anyone can get value from it without knowledge or effort

These features do help those in the business who put in the effort to learn the tools and their data. But history shows us that those folks are in the minority. If they weren’t, we wouldn’t still be talking about how to achieve data democratization. Business users who put in the effort and become recognized experts in their domain become the bottleneck as colleagues inundate them with requests.

The business has to meet the challenge to evolve, or perish

There’s no avoiding the massive change that organizations and their employees face today due to the explosion of data. Technology can help, but it’s not a cure-all. The chief data officer (CDO) can’t drive the change by herself, nor can the data team. The business has to meet the challenge to evolve, or perish. 

Therefore, leaders need to make basic data literacy a requirement of employment. Here are some approaches to consider:

• Upskill current employees through education on basic data topics such as descriptive versus quantitative data, precision versus accuracy, correlation versus causation, and how operational processes affect data quality.

• Test job candidates on data literacy to determine where and to what extent they need data training.

• Manage data literacy performance, for example by providing incentives for those who author data as part of their jobs to beat minimum data quality standards.

(For a deeper understanding of data literacy programs see my colleague Dave Wells’ excellent work on this subject, Building a Data Literacy Program: What, Why, and How.)

As a member of data teams for many years, I understand how hard it is to keep up with the pace of technology-driven change because I’ve had to evolve constantly over the course of my career. Data teams really do want to help their business colleagues. They have empathy for the magnitude of the challenge their colleagues face.

Leveraging data is not just a service provided by the data team, but a fundamental aspect of everyone’s work

The path forward requires acknowledging mutual responsibilities to bridge the knowledge gap. Data teams must continue to educate and empower their business colleagues, tailoring solutions to meet their needs while pushing for a basic level of data literacy across the organization. On the other side, business units must commit to a journey of continual learning and collaboration. They must recognize that leveraging data is not just a service provided by the data team, but a fundamental aspect of their own work.

Data democratization is a gradual process, not an instant solution – evolution, not revolution. Leaders must provide the time and resources required for employees to learn and practice new skills. Doing so will impact project schedules and budgets. But as the saying goes, no pain, no gain.