Big companies are notoriously slow to adopt new technology, but that isn’t how things are shaping up in the area of artificial intelligence, according to a new survey of nearly 500 business executives.

The research, which was conducted by Harvard Business Review and sponsored by Apttus Inc., found that 60 percent of executives at business-to-business companies “believe their future success depends on AI implementation.” More than one-third of that group are already in pilot or production with machine learning technology.

Respondents were nearly unanimous in seeing a need for machine learning in their organizations and 52 percent said they believe they understand the benefits AI can deliver. By a wide margin, that’s in the area of predictive analytics, which forecasts future outcomes based upon historical data. More than four in five respondents pegged this as their greatest opportunity. Other areas of perceived value include text classification or mining, fraud detection, e-commerce and behavior or sentiment analysis, but those benefits were cited far less often.

The results dovetail with a recent Bain & Co. survey of 334 executives, in which more than two-thirds said their companies are investing heavily in big data, 40 percent expect to see a “significantly positive” impact and 8 percent predict “transformational” results.

Thirty-eight percent of executives in the HBR/Apttus research said their leaders can articulate an AI strategy. That’s good, because guidance on how to use the technology should come from the top of the organization and goals should be clearly defined, researchers said.

B2B is different

The report focuses on some of the unique aspects of B2B organizations that require different approaches to AI than what has been used in consumer markets. There’s no place for quirky bots in business scenarios, said John Frémont, an AI evangelist and former go-to-market lead for AI at Accenture LLP, who is quoted at length in the report.

“A B2B virtual assistant has to be action-driven,” he said. “The right approach is to say, ‘Let’s make our jobs more efficient,’ so that it’s in the background helping us without us even realizing it.”

Researchers stress that AI can spark fear of job loss, so it’s best positioned as an assistant to people, rather than a replacement for them. “Apprehension and demoralization in sales may arise due to alternate channels that combine AI with self-service models, posing a risk to income from incentives on recurring or repeat business with existing clients,” researchers write. “It’s not about replacing humans; it’s about enabling humans to focus on the things they are uniquely good at.”

For example, a machine learning system at Nasdaq Inc. enables advisers who disagree with an AI-based prediction to enter their own prediction alongside it. “Then we can track who’s actually right, and if the computer was wrong and the analyst was right, we can take that feedback and put it into the machine learning algorithm, which allows the system to get better and better,” said Michael O’Rourke, head of machine learning intelligence and data services at Nasdaq.

AI ultimately “holds the potential of providing a practical approach to fundamentally improving how different stakeholders in selling, commerce, and revenue management can improve practices through superior insight, intelligence, and guidance,” researchers conclude. “Companies that commit to and effectively adopt these capabilities are poised to achieve dramatic competitive advantages in revenue generation.

Apttus sells a cloud service that streamlines many of the logistical steps involved in closing large business deals, from the delivery of a price estimate to billing. The company last month announced a $55 million funding round.

Image: Flickr CC