Companies are gathering more and more data, but many are still struggling to turn that information into business outcomes. In the book Monetizing Your Data: A Guide to Turning Data into Profit-Driving Strategies and Solutions, published last month, analytics experts Andrew Roman Wells and Kathy Williams Chiang explain how enterprises can begin cashing in on their information by using analytics.

Wells, CEO of management consulting firm Aspirent, and Chiang, vice president of business insights at Wunderman Data Management, have each been practicing analytics for more than 20 years. While Chiang focused more on analytics, and Wells worked with business leaders on strategy, each observed that these two major processes were often disconnected within companies.

“Data scientists were churning through data trying to find insights, while the planning people had ideas and went to find the data to justify those ideas, instead of taking a more objective approach,” Chiang said. “We saw the opportunity to marry these two together so they become more synergistic and supportive–bringing objectivity to the planning side, and meaning and purpose to the data side. It shows how the quantitative and qualitative can be connected in the business environment.”

SEE: 10 skills you need to become a great project manager

When they teamed up together for a project at a large hotel company, they found that the combination of their methods made for a new approach that links decision theory, data science, and agile analytics. Their methodology evolved over the course of several subsequent projects, and resulted in the book, Wells said.

“The old era of how you gathered requirements and put together analytical solutions for driving business value no longer works in the new era,” Wells said. “You have to change the conversation so that instead of wondering about the question, you’re trying to solve for the decision.”

Here are five steps outlined in Monetizing Your Data to help businesses use their data to get the biggest business impact.

1. Discovery

The discovery phase involves focusing on the nature of the problems or opportunities the organization is facing, and trying to get a full sense of their business landscape to determine what actions and impact they are looking to achieve, Chiang said. This phase is interactive, and involves workshops and interviews to gather input from all users up front.

The discovery step also creates alignment for the project, Wells said. “It sets the foundation for understanding business objectives and drivers, and developing hypotheses we align [solutions] to so we know we’re solving a real-world problem,” Wells said. “It’s the launchpad for the rest of the project.”

2. Decision analysis

This phase begins to structure the analytical problem a company is addressing, allowing staff to be more systematic about the problems they are trying to get to the root cause of, so that they can test hypotheses and develop actions around them, Chiang said. This phase also involves looking for biases that the team may be bringing to the problem, to ensure that members have not already decided on a solution and are now seeking data to confirm it. “It’s being more intentional about what alternatives are worth considering, and being aware of predispositions the team may be bringing to the project,” Chiang said.

This works whether the decision involves anything from AI and machine learning to predictive modeling, Wells said. “It’s about trying to enable decisions and actions for business users, not just answer questions,” he added.

3. Monetization strategy

After you determine the problem, develop hypotheses, and discover ways to address it, the next step is to determine if its solution has value, and the impact on business performance, Wells said. This involves looking at the actions you plan to take and examining which part of the balance sheet those actions will affect.

“With big data, what’s often missing is value,” Chiang said. In this part of the process, you can test your analytics to see if they move a lever, she added.

4. Agile analytics

Agile analytics allow a company to enable new decisions and create a solution, using guided analytics and decision theory to create an end solution, Wells said.

This system means that every manager in the organization can access a rich set of data at its most granular form, as opposed to using reporting systems that generate reports based on averages and aggregations and lose insights. “Data modeling becomes critical–where before you had daily summaries that were simple and flattened, now, marketing, business, and IT analysts will be able to work with a rich spectrum of data,” Chiang said.

This stage helps end users see the value of the journey, Chiang added. She likens guided analytics to a walled garden–you want to give people freedom to play with the data, but also want to have guard rails up so they can work through it without getting stuck in analysis paralysis, she said.

5. Enablement

Enablement involves hardening the process, Wells said. “It’s making sure the data is valid, the calculations are correct, and the end users are engaged in the testing process,” he said.

Though agile analytics ensures that users are involved throughout the process, this last step is the most important for them to sign off on, Wells said. “If you put in an analytic solutions and users don’t trust the data, they won’t use it,” he said. “You have to have the buy in along the way, but especially at the end for validity.” Then, you can go through training and rollout processes to enable users to use it.

Chiang echoed his point. “A lot of great analytics projects end up sitting on the floor, because they are not tested with the monetization strategy, or they collapse at the finish line,” she said. “Users need to understand how it can change their daily work, and see the value.”

You can find the book, as well as templates, checklists, and examples to help you apply the methodology in your business environment, here.