'Success requires the ability to look at data in novel ways': 5 steps to creating a data-driven business

Success often requires the ability to analyze data in novel ways and then act on the resulting insights. For instance, a cereal brand that comes to realize that its customers are affluent sports enthusiasts can develop creative and personalized campaigns to target these customers.

While every organization will encounter challenges along the way, the good news is that there are basic steps that can be taken to build a truly data-driven business. In my experience, I have found that leading companies take five dimensions into account when developing data strategies and systems:

• People

The first step is assessing whether you have the right staff with the right skills and whether staff are organized around clear goals. Data-driven marketing programs shouldn’t be placed in a silo. Instead, they have to be part of a central organization that focuses on analytics or marketing.

The projects with the best results tend to have an executive sponsor and a program manager who together are accountable for success. Creating a steering committee with cross-functional representation is also crucial to ensuring that these programs aren’t just satisfying one team’s goals and benefit from group buy-in and diverse perspectives.

• Processes

Companies need to establish clear processes for using data insights, including testing and learning from marketing campaigns. These processes will allow an organization to continually refine how it targets consumers.

Companies should also develop processes for tracking how publishers, media buyers, and creative agencies are executing a specific data-focused marketing plan. It is far too common for a media planner to resort to the usual tactics and strategies. The right processes and oversight help ensure that everyone is on plan.

• Tools

The nuts and bolts also need to be in place. Data often exists in silos, but success or failure hinges on the ability to bring it all together. This means assessing what in-house or third party tools are used to manage data; whether there is connectivity between systems; whether there are APIs that can push the data in real time; and how easy is it to extract that data and draw insights from it.

• Data

This is the most obvious aspect of any data strategy. However, it requires going much further than understanding what data a company possesses. It requires careful and thoughtful management around governance. This includes setting rules that govern how data is managed and developing an understanding of the data’s integrity and any gaps that may be present.

Further, as this is people data, privacy considerations must be addressed, along with any sector-or region-specific regulatory considerations, to ensure the implicit and explicit contract that exists with the end consumer is honored every step of the way.

• Goals

This last dimension rounds out the other four. Goals keep a company on track by creating clear accountability and alignment on the purposes of the overall data strategy. What are the core objectives? Are they widely communicated? Is there any inconsistency? Are the goals reinforced and rewarded? These are the kinds of questions that data-driven marketers should be asking themselves almost every day.

In the end, technology is the means, not the end, and it is only as effective as the business context in which it is deployed. Developing the right capabilities to get the most out of data systems is often a multi-step process.

Successful enterprises take a holistic approach, across all dimensions, to ensure data is translated to insight, and insights into action, all with a direct and measurable impact on business results.

Paul Bates is vice-president, solutions at Krux