7 Steps to Building a Data-Driven Culture

The surge and complexity of data being generated every minute has changed how we do business in profound and unalterable ways. It has fueled a new era of fact-based strategy creation and innovation.

A research study by MIT professor Erik Brynjolfsson, cited in Forbes, found that companies that embrace decision-making based on data have output and productivity that is 5-6% higher. In the same Forbes article, Richard Joyce, senior analyst at Forrester, stated, “For a typical Fortune 1000 company, just a 10% increase in data accessibility will result in more than $65 million additional net income.” Forrester also estimated that data-driven businesses are growing at an average of more than 30% annually.

In light of this, it’s no wonder that companies are investing heavily in data collection and processing. They’re installing high-powered infrastructure and analytic tools, as well as employing the best teams of data scientists—all for the prospect of reducing uncertainty, optimizing supply chains, reducing costs, and clarifying strategy.

Yet many companies are struggling to grasp the full value that data and analytics can unlock: the value that comes beyond the technology, when data is truly driving decisions within the organization.

So what is creating this gap?

Surprisingly one of the biggest challenges to creating a data-based business isn’t a lack of technology but a lack of cultural implementation.

The culture of an organization is based on a set of shared values, objectives, attitudes, and practices. Culture directly shapes how the business assesses performance, allocates resources, and encourages people to act.

Data has the power to transform all of these areas. Hence, a data-driven company has a culture that influences every person within the company to make decisions based on data.

Sound easy? Not so fast: if you use incorrect or low-value data, you will corrode cultural trust in data-driven decisions and force employees to revert to risky gut-feel and opinion-driven decision-making.

Unlike technology, a data-driven culture can’t be bought from a vendor. Like any cultural shift, it requires more EQ than IQ. Though it is easy to inject data into a decision-making process, it is far more challenging to embed the practice of data-driven decision-making into people’s mindset and behavior, where it is seamless and instinctual within the organization.

At Accelerated Growth, we have worked with hundreds of entrepreneurial organizations as they scale. During those periods of growth, instilling and maintaining the right culture is critical to the sustainability of the organization. As such, we have found the following seven steps to be a proven process for creating a data-driven culture:

1. Start at the top

Heavy investment in data analytics is pointless without a commitment from the company’s leaders. That commitment must be demonstrated by more than sporadic proclamations—there must be an ongoing, informed conversation with key decision-makers who then set expectations for anchoring decisions in data and continually show how data impacted the decision-making process.

Such an approach has the power to influence behavior at all levels of the organization as team members want to communicate effectively with leaders by using the leader’s terms and language.

Leaders can also influence the team’s attitudes toward data by being among the first to use data and analytics software to exhibit data-driven decision-making in daily work. This leading-by-example showcase encourages a shift from risky, “gut-feel” management styles to more scientific data-driven and analytics-enabled models. 

2. Position data as a strategic asset

Data-driven decisions will create meaningful results only when they are closely aligned with the core business objectives. Organizations that struggle to create value through analytics tend to develop analytics either in isolation or far removed from the day-to-day operations, in scattered silos across the organization.

Once you have achieved the commitment from leadership described in step 1, the next step is to identify business objectives that you will dedicate data-management efforts towards achieving.

Don’t ask, “What kind of insights can I extract from data?”

Instead, ask, “What kind of analytics and insights can help to achieve [a particular business outcome]?”

Unless data is embedded as a critical constituent of the strategic initiative, your organization will struggle to generate and use insights, ideas, and innovation for accelerating business results.

3. Clearly define metrics

Having a clear, unambiguous, and accepted list of metrics—as well as their definitions—is critical to the success of data-driven decision-making. Otherwise, companies can waste countless hours trying to justify the subtler aspects of a metric that should be universal, and team dissonance will ensue as they derive their own interpretations from the data.

Begin developing the list of metrics by first reviewing the processes currently being used to drive decisions, then identifying either: a) the gaps in the available data, or b) the inefficiency in procuring the data necessary to make that decision.

Once you have drafted the list of metrics, visualize how they will be used. It is often not the fundamental definition of the metric where misunderstandings occur but how exceptions to the definition are handled. Hence all possible scenarios need to be articulated, debated, resolved, and recorded for communication and reference.

Last, edit the list down. Collapse multiple similar metrics into a single universal metric or define scenarios where it becomes imperative to split one metric into two or more.  

4. Democratize data access

You cannot foster a data-driven culture if access to data is limited to a few individuals, teams, or functions. To create a competitive advantage and trigger demand for data across the organization, you need to create a common platform through which people can easily access data.

This is not just a software solution. While employing the latest analytics tools and technology can free your data from its organizational silos, senior leaders need to simultaneously model transparency on every level.

Sharing insights across the company can give rise to diverse perspectives and enable prolific ideas and solutions. Also, expanding data access at all levels can empower your data scientists to focus on more strategic projects and encourage your employees to use data on a more regular basis to create value.

5. Assess uncertainty

While you need to ensure that your data is precise, accessible, and trustworthy, achieving absolute certainty is impossible. An organization that encourages experimentation, innovation, and exploration also must allow room for failure with data. 

That said, you can require teams to be explicit and quantitative about their levels of uncertainty, which will force a deeper inspection and reflection on their implied assumptions and beliefs.

How did the team approach the problem?

What alternatives did they consider?

What were the assumed trade-offs?

Why were their approach chosen over the other approaches?

These questions can force teams to assess their thinking and prompt them to expand their range of alternatives or to rethink fundamental assumptions. Conversely, by not asking difficult questions, you limit innovation and creativity and even hide potentially disruptive problems under a façade of continuity.

Also, embracing uncertainty and exploration compels teams to run experiments, eventually encouraging more and better ideas to be tested—sometimes uncovering insights trapped under systems and processes. This is also a good opportunity to embed processes with proof of concept.

6. Enhance data literacy

Team members company-wide must be able to easily extract meaning from raw data. The sheer level of sophistication and detail of data can often seem overwhelming, forcing data literacy to remain trapped within a single department or subject-matter experts, while most associates confine themselves to familiar but low-value, static reports.

Most bad data practices are more due to a lack of experience than deliberate design. Though it is critical to hire qualified data analysts and scientists, everyone who is using data should be provided with basic training on how to read and understand the data they’re expected to consume and use on a regular basis.

Beyond data literacy, team members should feel encouraged to articulate and share key insights with others to deliver value.  

7. Maintain a single source of truth (SSOT)

Regardless of how many data systems your organization has, there needs to be a single source of truth (SSOT).

Called the master data, a SSOT is a centralized and unified source of data from which the whole company can draw insights and eliminates data silos. In absence of the master data, different teams may revert to their own preferred sources of information and combinations of metrics to determine different results. This can include the use of low-quality or incorrect data and ultimately result in bad decisions.

A SSOT with defined outcomes and measures can reduce time spent hunting data across the organization, using that time instead to deliver actionable insights and increase the speed of progress.

 

Using data to drive decision-making is unarguably an optimum way to accelerate organizational growth. However, there are potential pitfalls to be understood and avoided before setting out on this journey.

Join the trend, but tread with purpose.

Need help in your journey? Contact us here.