Linkedin
Tom Davenport
Visiting Professor at Saïd Business School, University of Oxford
March 17, 2022
Executive Summary
by Joaquim Cardoso MSc.
AI Powered Health Care Unit @ Digital Health Care Institute
The Health Foundation
March 18, 2022
What was the background 3 years ago (2019)?
- In 2019 only 70% reported minimal or no value from their AI investments (MIT Sloan Management Review and Boston Consulting Group AI survey).
- One of the reasons for poor returns was that relatively few projects were deployed into production; they were too often research exercises.
What are the results now?
- In the 2022, 92% of large companies reported that they are achieving returns on their data and AI investments (NewVantage Partners Survey)
- That’s up markedly from 48% in 2017.
What about the investment plans and making money?
- Companies aren’t just planning on spending serious money on AI in 2022 — they are already making good money from the technology.
- Companies are increasing investments in data and AI, equaling last year’s percentage.
- 26% of companies have AI systems in widespread production — more than double the 12% in last year’s survey.
- Other surveys suggest that companies around the globe are also registering more value with AI (Deloitte survey)
What are the results of value creation with AI?
AI adoption and value are increasing. The number of companies reporting AI adoption in at least one function had increased to 56%, up from 50% in 2020 (A 2021 McKinsey global survey on AI)
More importantly, the survey also indicates that AI’s economic return is growing.
- The share of respondents reporting at least 5% of earnings (EBIT) that are attributable to AI has increased to 27%, up from 22% in the previous survey.
- Respondents also reported significantly greater cost savings from AI than they did previously in every function, with the greatest improvements coming in product and service development, marketing and sales, and strategy and corporate finance.
- 78% of companies also reported improved cultural enhancements (nonmonetary benefits), such as collaboration within teams (MIT Sloan and BCG Survey)
What are some some of the best practices?
- Companies seeing the biggest earnings increases from AI were not only following practices that lead to success, including MLOps,
- but also spending more efficiently on AI and taking advantage of cloud technologies to a greater extent.
- 80% of companies are already using some form of automation technology or plan to do so over the next year (IBM Survey) — around a third driven by the pandemic.
What is the final message?
- There is still substantial room for improvement in the economic returns from AI, of course, and these surveys tap only subjective perceptions.
- The biggest remaining stumbling block, according to a recent small survey of data scientists, is that the majority of machine learning models are still not deployed in production environments within organizations. Companies and AI leaders still need to work on this issue.
- There are strong signs that AI is here to stay in the business landscape.
ORIGINAL PUBLICATION (full version)
Companies Are Making Serious Money With AI
Linkedin
Tom Davenport
Visiting Professor at Saïd Business School, University of Oxford
March 17, 2022
With the start of each year come predictions, plans, and surveys from consulting firms.
When it comes to artificial intelligence, multiple recent surveys indicate that companies aren’t just planning on spending serious money on AI in 2022 — they are already making good money from the technology.
… companies aren’t just planning on spending serious money on AI in 2022 — they are already making good money from the technology.
A bit of context might be helpful.
Despite some AI successes, one of the challenges in recent years has been that projects involving the technology have frequently lacked sufficient economic returns.
In a 2019 MIT Sloan Management Review and Boston Consulting Group AI survey, for example, 7 out of 10 companies reported minimal or no value from their AI investments.
One of the reasons for poor returns was that relatively few projects were deployed into production; they were too often research exercises.
Production deployments admittedly can be difficult, since they usually require integration with existing systems and processes, worker reskilling, and the ability to scale AI technology.
Despite some AI successes, one of the challenges in recent years has been that projects involving the technology have frequently lacked sufficient economic returns.
One of the reasons for poor returns was that relatively few projects were deployed into production; they were too often research exercises.
Just a few years later, things are beginning to change.
In the 2022 survey of senior data and technology executives by NewVantage Partners (where Randy Bean is CEO and cofounder, and Tom Davenport is a fellow), 92% of large companies reported that they are achieving returns on their data and AI investments.
That’s up markedly from 48% in 2017.
The same percentage (92%) said that they are increasing investments in data and AI, equaling last year’s percentage.
Twenty-six percent of companies have AI systems in widespread production — more than double the 12% in last year’s survey.
The survey also asked respondents whether their organizations were data driven, and only 26% said they are. However, that doesn’t seem to be preventing them from making progress on AI.
Just a few years later, things are beginning to change. …In 2022… 92% of large companies reported that they are achieving returns on their data and AI investments
That’s up markedly from 48% in 2017.
The same percentage (92%) said that they are increasing investments in data and AI, equaling last year’s percentage.
… only 26% said they are data driven .However, that doesn’t seem to be preventing them from making progress on AI.
Returns on AI Around the Globe
The NewVantage survey respondents largely represent North American companies.
But other surveys suggest that companies around the globe are also registering more value with AI.The State of AI in the Enterprise survey by Deloitte (where Tom is a senior adviser to the AI practice), fielded in mid-2021, found that two types of companies are getting value from their investments.
Twenty-eight percent of survey respondents were classified as transformers — companies reporting high business outcomes and a relatively high number of production AI deployments (six on average).
- This group has identified and largely adopted leading practices associated with the strongest AI outcomes, including having an AI strategy, building an ecosystem around AI,
- and putting organizational structures and processes in place (such as machine learning operations, or MLOps) to keep AI on track.
The other group getting value, accounting for 26% of respondents, was labeled pathseekers. They reported high outcomes but a lower number of deployments.
- They have also adopted capabilities and behaviors that have led to success with AI, but on fewer projects. They have not scaled to the same degree as transformers.
… two types of companies are getting value from their investments. (1) 28% of survey respondents were classified as transformers, (2) 26% of respondents, was labeled pathseekers.
Still, that’s more than half of the global respondents reporting positive business outcomes from AI. As we’ve noted, it’s difficult or impossible to benefit from AI without deploying it, but these results suggest that you don’t need a lot of deployments to get value.
Still, that’s more than half of the global respondents reporting positive business outcomes from AI.
A 2021 McKinsey global survey on AI also found that AI adoption and value are increasing. McKinsey found that the number of companies reporting AI adoption in at least one function had increased to 56%, up from 50% in 2020.
More importantly, the survey also indicates that AI’s economic return is growing. The share of respondents reporting at least 5% of earnings (EBIT) that are attributable to AI has increased to 27%, up from 22% in the previous survey.
We’re not sure how survey respondents would calculate the percentage of earnings attributable to AI, but their responses do suggest high value.
… The share of respondents reporting at least 5% of earnings (EBIT) that are attributable to AI has increased to 27%, up from 22% in the previous survey
Respondents to the McKinsey survey also reported significantly greater cost savings from AI than they did previously in every function, with the greatest improvements coming in product and service development, marketing and sales, and strategy and corporate finance.
Respondent … also reported significantly greater cost savings from AI than they did previously in every function
And echoing the Deloitte survey, McKinsey found that progressive AI practices are being rewarded.
Companies seeing the biggest earnings increases from AI were not only following practices that lead to success, including MLOps, but also spending more efficiently on AI and taking advantage of cloud technologies to a greater extent.
Companies seeing the biggest earnings increases from AI were not only following practices that lead to success, including MLOps, but also spending more efficiently on AI and taking advantage of cloud technologies to a greater extent.
A survey by IBM offers some insight into the impact of the COVID-19 pandemic on AI adoption, with a particular focus on automation-oriented technologies.
It found that 80% of companies are already using some form of automation technology or plan to do so over the next year.
Just over a third of the organizations surveyed said that the pandemic influenced their decision to adopt and use automation as a means of improving productivity.
The respondents to the IBM survey were IT professionals, which may have influenced the results; IT process automation (known as AI for IT operations, or AIOps) is a popular use case for the technology.
… 80% of companies are already using some form of automation technology or plan to do so over the next year.
Nonmonetary Benefits
We should also mention an interesting 2021 survey conducted by MIT Sloan Management Review and Boston Consulting Group that set out to assess not the monetary benefits of AI but its cultural enhancements. Because no one (to our knowledge) has asked these types of questions before, we can’t make comparisons to the past.
… 58% of all respondents who had participated in an AI implementation agreed that their AI solutions improved efficiency and decision-making among teams (MIT Sloan MR & BCG)
A majority of that group (78%) also reported improved collaboration within teams. Are improved decision-making and collaboration indicators of cultural benefit? We’re not sure, but they could certainly translate into economic value.
58% of all respondents who had participated in an AI implementation agreed that their AI solutions improved efficiency and decision-making among teams.
The survey also found that AI yields strategic benefits, but they mostly accrued to companies that use AI to explore new ways of creating value rather than cutting costs.
Those that used AI primarily to create new value were 2.5 times more likely to feel that AI is helping their company competitively compared with those that said they are using AI primarily to improve existing processes; they were also 2.7 times more likely to agree that AI helps capture opportunities in adjacent industries. It’s easy to see how these traits could turn into economic value.
The survey also found that AI yields strategic benefits, but they mostly accrued to companies that use AI to explore new ways of creating value rather than cutting costs.
For those who want the current “AI spring” to bloom forever, this is all great news.
There is still substantial room for improvement in the economic returns from AI, of course, and these surveys tap only subjective perceptions.
The biggest remaining stumbling block, according to a recent small survey of data scientists, is that the majority of machine learning models are still not deployed in production environments within organizations.
Companies and AI leaders still need to work on this issue.
There is still substantial room for improvement in the economic returns from AI, of course, and these surveys tap only subjective perceptions
The biggest remaining stumbling block, according to a recent small survey of data scientists, is that the majority of machine learning models are still not deployed in production environments within organizations.
Companies and AI leaders still need to work on this issue.
However, the fact that so many business leaders responding to so many surveys on the topic feel that their organizations are capturing substantial value from AI is a definite improvement over the recent past, and a strong sign that AI is here to stay in the business landscape.
About the authors
Thomas H. Davenport (@tdav) is the President’s Distinguished Professor of Information Technology and Management at Babson College, a visiting professor at Oxford’s Saïd Business School, and a fellow of the MIT Initiative on the Digital Economy.
Randy Bean (@randybeannvp) is an industry thought leader, author, and CEO of NewVantage Partners, a strategic advisory company that is now a division of Wavestone, a global consultancy based in Paris. He is the author of Fail Fast, Learn Faster: Lessons in Data-Driven Leadership in an Age of Disruption, Big Data, and AI (Wiley, 2021).
*This article was originally published by MIT Sloan Management Review on February 17, 2022.
Originally published at https://www.linkedin.com.
