This article originally appeared on HBR.org.

The financial services firm USAA prizes its relationships with customers and consistently earns the highest customer loyalty scores in the industry for both its U.S. banking and insurance businesses. The high level of trust affords USAA access to copious customer data that it uses to inform personalized cross-selling and upselling. The company does deep data mining through multiple sources to spot signature events in customers’ lives. Those events trigger USAA to contact the customer at just the right time, with just the right offer, such as auto insurance when a customer’s daughter is about to turn 16. As a result, USAA outperforms most competitors in the number of products held by its customers.

USAA is something of an exception, though. Many firms still underinvest or underachieve in growing share of wallet with existing customers, compared to their initiatives in acquiring new customers. In the past, marketers have struggled to deliver the higher response rates they need from existing customers—a smaller group than potential new customers.

Several trends have converged in recent years to break through the barriers to higher response rates. The proliferation of customer data and the greater computing power to organize and analyze that data make it feasible to create much more dynamic and insightful profiles of customers. Digital channels now allow companies to fine-tune marketing messages based on observed behavior. And many companies have made significant investments to improve their customers’ experience, earning them greater leeway to cross-sell.

These favorable changes have motivated marketers to reassess cross-selling opportunities, which can be substantial. For example, Bain & Company’s recent analysis of the U.S. telecommunications industry found that up to 60% of customers split their services across multiple providers for mobile phone, landline, TV, and internet services. For one telecom provider, convincing just 10% of those customers to switch one service from a competitor was worth up to $480 million in incremental annual revenue. A similar story can be told in retail banking, insurance, credit cards, retail, and other industries.

Managers wanting to grow share of wallet and raise the productivity of their cross-selling efforts may have to confront longstanding practices that stand in the way. To that end, here are five guidelines for expanding share with current customers.

Take a balance-sheet view. Many businesses organize around product lines and focus on achieving quarterly, monthly, or even weekly targets. Although this centers the organization on achieving in-year goals, it hinders efforts to maximize the long-term value of customer relationships. Effective cross-selling organizations, such as American Express, complement the P&L perspective with a longer-term, balance-sheet view of the business and a multiyear view of customer value.

Create dynamic, high-resolution customer profiles. With the exponential growth in data and the increased computing power available, companies can now combine internal and external data that spans several years to build more-useful customer profiles. Knowing how the customer’s product usage has changed over time, how he or she has migrated among products, and which triggers or leading indicators caused changes in behavior is essential for designing effective share-of-wallet strategies.

Focus on discrete customer growth missions. When companies set customer growth goals, they too often pursue overly broad or diffuse objectives, such as “Let’s sell home insurance to all high-value auto insurance customers,” rather than identifying and focusing on the pockets of greatest opportunity.

Defining a high-value customer growth mission, by contrast, narrows the aperture to focus on the organization. An example of such a discrete customer mission might be: “Let’s target our high-value customers who are in the market for home insurance and are customers of competitor X, which has low loyalty scores.”

Defining customer missions takes work. Companies must understand the most profitable customer segments, how their behaviors and preferences have changed over time, the products and channels they use, and how they stack up against the competition in each area. And they must calculate the real economic value of these segments, so that they can create targeted strategies that will be profitable.

Just get started. It’s a common misconception that a company must invest in new data warehouses or CRM systems before taking action. In fact, organizations can mobilize quickly around internal data that resides in existing databases, assemble external market data in a few weeks, and often use analytical tools that exist in-house.

MGM Resorts was looking to raise booking rates among existing customers at its casino and resort properties in Las Vegas. Using existing customer data, MGM launched a multivariate campaign, testing different offer packages (a room discount versus a coupon to play slot machines), travel windows, brand messages, email frequency, and more. For one set of target customers, the best combination of variables achieved a 180% lift in bookings over the control offer and was worth millions in incremental revenue when extrapolated to the full set of customers. Moreover, MGM learned how different groups of customers—say, gamblers and nongamblers—responded differently to each offer. To keep the insights coming, MGM has set up a new cross-functional test-and-learn team, which has continued to introduce new variations with these customer groups.

Build a repeatable model. The test-and-learn approach allows a company to log some early wins, build new capabilities, and iron out the kinks as the organization learns how to work to a different rhythm. Leaders can then decide if it makes sense to create a new team dedicated to share-of-wallet growth and possibly to invest in building new capabilities or installing new technologies.

Conditions are right for companies to reinvigorate their cross-selling strategies. Now it’s up to managers to harness the wealth of customer data, advanced analytic techniques, and the power of digital channels for customer growth missions. Companies that outperform in these missions will be able to systematically expand their share of spending among loyal customers.

John Senior is a partner with Bain & Company’s Customer Strategy & Marketing practice. Tom Springer is a partner with Bain & Company in Boston, and is co-leader of the firm’s Advanced Analytics practice. Lori Sherer is a partner at Bain & Company in San Francisco and is co-leader of the firm’s Advanced Analytics practice.