Bain research finds the best odds lie in selling new products to existing customers. Walmart and Nvidia are pursuing that strategy, but growth alone will not prove the bets are working.
Most attempts to grow beyond a core business fall short. A 2003 Harvard Business Review article reported that Bain & Company's five-year study found roughly three-quarters of adjacency efforts fail. A later Bain analysis of nearly 200 companies put overall odds of success at about 20%, Bain's Chris Zook wrote.
Those odds vary by strategy. Selling new products to existing customers offers a stronger path than building a new business around an existing capability. Established customer relationships, brand recognition and distribution can give a company an advantage when it introduces a related offering. A new business, by contrast, may require unfamiliar capabilities and expose the company to competitors it has not faced before.
Walmart and Nvidia are both expanding into adjacent markets by offering additional products and services to customers they already serve. Their businesses are different, but each illustrates the appeal of extending an established commercial relationship. The strategic question is whether these additions strengthen the core or create new dependencies, costs and competitive risks.
Where expansion goes wrong: the Haven venture
The risks of moving too far from the core were evident in Haven, the health care venture launched by Amazon, Berkshire Hathaway and JPMorgan Chase. The companies disbanded the initiative in early 2021, after roughly three years.
A STAT News examination pointed to an unclear mission and strategic missteps. The partners also continued pursuing their own health care initiatives, complicating the venture's purpose and coordination. Haven's closure does not establish that diversification generally fails, but it illustrates how difficult it can be to build a new business when partners lack a sufficiently clear and aligned operating model.
The higher-odds route: new products for existing customers
In data centers, Nvidia's CPUs and networking gear fit the higher-odds category. CEO Jensen Huang said the expansion follows customer demand, as customers now treat computing capacity as a direct source of revenue.
Nvidia told analysts that its revenue opportunity per gigawatt of data center capacity rises from $18 billion with Hopper to $25 billion with Grace Blackwell and $40 billion with Vera Rubin. These are company estimates of potential revenue opportunity, not realized revenue. The platforms combine increasingly broad sets of computing and networking components, with Vera Rubin also incorporating Groq technology. It expects data center CPU revenue to more than double in fiscal 2028.
The company's recent results show the scale of demand across its existing business. Second-quarter revenue reached $96.2 billion, up 106%, with data center revenue of $89.0 billion, up 117%, the company reported. Networking revenue rose 18% from the first quarter, and Spectrum-X Ethernet was up 2.6 times from a year earlier. Grace CPU revenue exceeded $5.0 billion over the past 12 months, Nvidia said.
These figures demonstrate the strength of Nvidia's data center business, although they do not establish how much future growth will come specifically from adjacent products.
Retail offers a second example. Walmart's global advertising business grew 46% to nearly $6.4 billion in the fiscal year ended Jan. 31, 2026, the company said. Walmart Connect, its U.S. advertising arm, grew 43% excluding Vizio in the second quarter of fiscal 2027, while global advertising rose 38%.
CEO John Furner told analysts that higher-margin lines, including advertising and membership fees, are improving the economics of e-commerce. He joined Walmart as an hourly associate in 1993 and succeeded Doug McMillon on Feb. 1.
Reorganizing and measuring around the new business
The retailer bought Vizio, a smart TV maker, to add screens where it can sell ads. Analysts say Vizio's value lies in closed-loop measurement. Combining TV viewing data with Walmart purchase data lets advertisers see whether an ad led to a sale, a capability that retail media networks without screens cannot easily match.
The structure around the bet changed as well. Walmart named Seth Dallaire chief growth officer, placing Walmart Connect, Walmart+, Walmart Data Ventures, Vizio and its marketplace under one executive. "Even the best teams need the right structure to win," Furner said when he announced the changes in January. Centralizing the platforms, he said, speeds shared capabilities and lets operating segments stay closer to customers.
The structure matters because an adjacent business needs more than investment to succeed. It needs access to the parent company's assets and customers without becoming disconnected from the needs of the core operation.
The test ahead: size, trust, concentration and staying power
Advertising remains small next to the whole: less than 1% of the $713.2 billion in revenue Walmart reported for the year. Its strategic importance is greater than that share suggests, however, because advertising and membership income can contribute disproportionately to profit. The business's weight lies in profit. Chief Financial Officer John David Rainey said in May 2025 that advertising and membership fees together made up about a quarter of company profit. Some advertising is also booked as a reduction to cost of sales, which complicates comparisons with rivals.
Results abroad trail the global figure: advertising at Walmart International grew 19% for fiscal 2026, against 46% globally.
Customer trust is a second concern. Walmart's data-driven model invites scrutiny. Furner wrote in a recent letter that the company will not use shoppers' income, purchase history or AI tools to set prices. "We price the product, not the person," he said.
That statement addresses pricing, not the separate questions surrounding the collection and use of customer data for advertising. The distinction matters as Walmart expands its ability to connect shopping behavior with media exposure.
Buyer concentration is a separate issue in Nvidia. Hyperscale customers contributed $49 billion of the $89.0 billion in data center revenue, tying a large share of revenue to the spending plans of a small group of buyers. Nvidia's third-quarter outlook of $108.0 billion, plus or minus 2%, assumes no data center compute revenue from China.
Staying power is the last question. Walmart said on its earnings call that Vizio's triple-digit advertising growth cannot last, though it expects overall advertising progress to continue.
For both companies, the long-term question is whether adjacent businesses can continue contributing to earnings and reinforce the core once their initial growth slows. Walmart must show that advertising can expand without undermining customer trust or retail relationships. Nvidia must demonstrate that a broader infrastructure portfolio can deepen customer relationships without leaving it overly dependent on a concentrated buyer base.
Adjacency may offer better odds than diversification into unfamiliar markets. It still requires discipline: the new business must create lasting value for existing customers, not simply add another source of revenue.
