What if you were responsible for analyzing album, fan, social media, and merchandise data for more than a thousand musicians? That was the task that Naras Eechambadi faced in 2021, when he joined Universal Music Group (UMG), which represents current recording stars from Lady Gaga to Eminem, legends like the Beatles, and numerous up-and-coming artists. As the company’s first chief global data and analytics officer, he needed to find a way to make UMG’s information available to its many business units and partners. So he and his team compiled data from brick-and-mortar stores, e-commerce sites, social media, marketing campaigns, emails, and a CRM system and used it to build a reporting and analytics tool called Fan Analytics, Marketing, and E-commerce (FAME), to help UMG’s partners, including labels and artists, identify growth opportunities.

FAME provided granular data and insights on the behavior of every fan and automatically suggested follow-up actions for each person. Soon listener engagement and conversion rates for marketing campaigns rose significantly, driving revenue growth of more than 30% in e-commerce channels. FAME also gave UMG an edge over its competitors when it was signing new artists and labels. By packaging UMG’s far-flung and unorganized data into an integrated, easy-to-use tool, Eechambadi’s team had found a way to grow the business while remaining aligned with the company’s principal mission—to connect artists with fans.

Making money by commercializing customer data (and the insights gleaned from it) isn’t a new idea. Credit-reporting agencies, which tell lenders whether a would-be borrower is likely to repay loans, have been around for more than a century, and grocery stores have sold shopping data (collected via their loyalty programs) for decades. In the digital age companies are learning more about consumers by following their online behavior—the products they buy, the websites they visit, the reviews and comments they leave, and so on. Now AI is making it easier to analyze and gain insights from that information, increasing its value even more. And in an era of slower economic growth, companies are becoming more interested in finding ways to monetize this asset.

Some companies are already seeing success. Although Amazon’s retail business remains its largest source of revenue, the firm has used its deep knowledge about customers’ interests to grow its advertising business, which brought in $56 billion last year. More recently, Walmart used a similar model to launch its online ad business, which now generates $4 billion annually. Much of LinkedIn’s $16 billion in revenue is tied directly to the user data it sells to recruiters. Financial services firms like Mastercard and Visa have set up entire consulting divisions—Mastercard Advisors and Visa Advisory Services—to sell companies the insights they gather from analyzing millions of transactions. Neither company formally reports exactly how much it’s making from these ventures, but Mastercard has suggested that the annual revenue from its value-added services division is growing at a double-digit rate. Some companies are selling user data directly to gen AI companies for use in training their large language models. When Reddit licensed its user data to OpenAI in 2024, the financial terms weren’t disclosed—but Reddit’s stock jumped 12% on the news. Stories like these have made more companies think about the best way to extract profits from their data.

Yet our research shows that companies still struggle to choose the right way to do that and don’t know where to begin. It’s not necessarily their fault. Data monetization isn’t as simple as emailing a spreadsheet to a paying client. Companies must know how to collect, organize, and analyze their data. They also need to determine the best use cases and understand how they should price their offerings. And too many of them create data offerings that aren’t closely related to their core business—a move that often turns into a low-profit distraction.

To identify the reasons some companies struggle with data monetization—and why others succeed—we conducted in-depth case research with more than 30 organizations and interviewed 12 senior executives who are spearheading data monetization efforts across retail, media, tech, manufacturing, and marketing. Drawing on what we learned, we developed a framework for how to approach data monetization. In this article we’ll explain it and provide advice on how companies can get started on the journey. It requires asking three strategic questions.

[Here’s the first of three strategies.]

Many companies are sitting on what they think is valuable proprietary data. Potential buyers for it may include tech companies, data brokers, hedge funds, and companies in adjacent businesses. But selling it involves more than running a report.

During our research we observed many instances where business leaders began the monetization process by building out their technical infrastructure. Often they spent a couple of years on it, only to realize that they didn’t know what products to develop or who might buy them.

The most successful organizations begin the monetization process by focusing on use cases within their core business and with existing partners, specifically their suppliers and customers. Why? First, existing partners understand the potential value of your data better than others do, because it’s specific to their industry and key goals. Second, because they already have relationships with your company it’s easier to work with them to identify good use cases for the data. Your established operations and sales relationships with them also make it easier to generate more revenue from the data and to capture and distribute the data to them once a project starts. Your sales teams and relationship managers can sell the new offering to them as a line add-on and scale it up quickly. And the final reason it’s best to work with businesses in your existing ecosystem is privacy. An organization’s proprietary data is often subject to strict sharing and custodian agreements, such as those precluding the sale to data brokers or other nonaffiliated third parties.

Even companies that understand that logic may be tempted by offers from data brokers that package and sell data to hedge funds or other nonstrategic partners. On the surface these seem to be quick-and-easy deals that require little effort: You sell the raw data, you make money, and the value ends there. However, such deals may be tricky to navigate. Identifying potential customers and settling on prices across multiple parties with no partnership history is difficult. These opportunities also may present significant risks to your customers and suppliers, such as data leaks that could endanger their core business and strategic priorities. We aren’t saying that it never makes sense to do deals like this, but our research shows that they typically are riskier and create less value than deals with strategic partners.

To protect yourself and your customers, you need to manage the privacy, regulatory, reputational, data security, and other risks data monetization creates from day one. Leading transaction companies, consulting firms, and tech companies, for instance, take great care to aggregate and anonymize benchmark data when sharing it with customers. But even if you do this, you should check whether the use of data in a product or service could be misconstrued by your partners. For any data-based offering, you need to work closely with legal and risk managers right from the start to assess potential problems and create mitigation plans.

In parallel, and guided by the data needs of the prioritized use cases of potential buyers, you can progressively build a modern data platform and companywide data assets. Efforts to monetize data that’s poorly organized, of low quality, or incomplete will backfire. Yet many companies have a ways to go to build strong data and tech foundations: While they may already have the technology in place to collect data, they can’t transfer it into a central repository. Or they can’t organize the data or verify its quality. Or they have no easy way to run analytics or produce data visualizations.

Most businesses aren’t entirely unprepared. They’ve begun to feed data from internal and external sources into data lakes and data warehouses and model how all the data will fit into cohesive data assets such as a customer 360 or a supplier 360. By leveraging flexible tools such as Databricks, Domo, and Snowflake, they can rapidly build data products to test with customers.