Where does best innovation happen? Not in stand-alone labs, some companies say

The re-evaluation comes as big businesses feel the pressure to harness cutting-edge generative artificial intelligence for new discoveries. The best ideas come from within the business, these companies say, rather than from colleagues in detached groups who then have to get buy-in from the company.

Walmart in January announced plans to close Store No. 8, an idea-incubation arm that had some associates based in California’s Silicon Valley. The stand-alone lab was no longer needed, the Arkansas-based retailer said, because it had developed a new method that embedded innovation deeper across the company, giving more employees the opportunity to contribute.

“We’re taking a different approach," said Anshu Bhardwaj, chief operating officer of Walmart’s technology unit. “So associates are innovating while they’re driving the business."

Some tech leaders say it is a mistake to position innovation as something separate from the business itself, and that doing so can often lead to simply pursuing technology for technology’s sake. Companies risk falling into that pattern with AI, by focusing on where they can apply the technology rather than the business problems it could potentially solve, they said.

How it started

A wave of innovation labs started when companies, caught flat-footed by the sudden success of digital-first companies like Amazon and Google, set up outposts in tech hubs like Silicon Valley.

Some of those efforts paid off with increased e-commerce sales, entries into new markets and faster product development. Walmart said its Store No. 8 delivered several innovations, including a service that can deliver groceries directly to a customer’s refrigerator, if the customer arranges to be home at the time or provides access through a smart lock or keypad.

But there have been challenges as well. Efforts developed in stand-alone labs don’t always align with problems the business is trying to solve. And even when they do, business counterparts with existing ways of working aren’t always excited about using them, executives said.

Changing gears

Now some companies are rethinking their labs as they prepare for the new wave of AI development.

Walmart closed its Store No. 8 and shifted to a “Four in a Box" method, which brings together employees with business, product, technology and user experience roles to identify and solve challenges through tech.

The approach has allowed Walmart to more quickly deploy new solutions across the whole business, such as adding early-morning and late-night options to its in-home delivery service.

Assigning innovation efforts across business units makes sense for Walmart, said Bhardwaj. “It’s very difficult for any one single team to be the subject matter expert and the tech expert at the same time," she said.

Ford Motor’s innovation lab in Silicon Valley is now more about tapping into the region’s talent pool than its launch mission of working with startups on new technologies, the automaker said.

Responsibilities for innovation have evolved to be more integrated across many functions and departments within Ford, a spokesman said.

Former CVS Chief Information Officer Stephen J. Gold oversaw the creation of the drugstore chain’s innovation lab in Boston, at a remove from corporate headquarters in Woonsocket, R.I., to bring new digital products and services to market. “The goal is to basically operate in a mode that mimics a startup," he said in 2017.

Today, CVS is developing a new strategy and focus for its innovation labs, the company said, although it declined to get more specific on how the new focus will differ from the old one.

Payment-technology company Mastercard operates a unit called Mastercard Foundry where dedicated teams do more direct applied research on new technologies, such as quantum computing, but that is in addition to encouraging all employees to innovate with customers on real business problems.

“There’s a mistake sometimes where organizations say innovation is something off the side rather than something that you do every day and needs to be part of every team," said Mastercard Chief Technology Officer Ed McLaughlin. “Innovation is like ethics. And it has to pervade the whole organization."

Mastercard recently worked with Mercedes to enable digital payments via fingerprint biometric authentication directly from the vehicle’s dashboard at more than 3,600 service stations in Germany.

Sticks, not carrots

Some companies remain committed to their labs, despite the challenges. Jeff Wong, Ernst & Young’s global chief innovation officer, oversees a number of labs at the professional-services firm that develop technology both for its own use and that of its clients. The company gains a fivefold return on technologies developed in the labs—enough to keep them open, he said.

Wong said he strives to keep the stand-alone model in tune with real business problems. “It’s a very hard thing to do," he said.

At the financial giant BNP Paribas, Matthieu Soulé oversees a Silicon Valley-based lab that aims to bring cutting-edge startup technology into the bank and its customers.

Colleagues in other parts of the business aren’t always excited about the solutions he brings in, Soulé said. He said he has been working to build better bridges with the rest of the company and to position new technology as something they need to survive in a changing world, rather than just a cool addition.

“People are sometimes more reactive to a stick than a carrot," Soulé said.

There is definite value in companies setting aside resources to consider moonshot technologies 10 years into the future, albeit typically only in industries with long product development cycles such as aerospace, said Greg Larkin, who founded Punks & Pinstripes, a global community of entrepreneurs and executives, and has advised on innovation at Google, Uber and PricewaterhouseCoopers.

But labs tend to gravitate toward technologies that don’t have a clear path to driving real business value, now or in the future, he said.

“When you have an innovation lab, you’re only answering the FOMO [fear of missing out] question, meaning: ‘Is this a sexy thing that we might be caught flat-footed if it moves forward without us?’" Larkin said.

Write to Isabelle Bousquette at isabelle.bousquette@wsj.com