Machine learning (ML) is driving AI adoption in business, with 80% of data and analytics leaders reporting that their organizations have already implemented ML initiatives, according to new Corinium research.
The global survey of 640 data and analytics leaders proves that there is enthusiasm for AI and ML in the business community. Virtually everyone we polled said their organization is open to using the technologies, and 99% expect to see ROI from their AI investments within two years.
Given the amount of media coverage AI and ML receive, it should be no surprise that so many businesses are experimenting with these technologies. However, only 5% said their businesses are ‘fully open’ to embracing AI and ML. This suggests that many are yet to be convinced about the benefits of AI and ML.
“There certainly needs to be education and level-setting around the actual use cases for AI and ML in all industries,” says Christopher Boone, VP, Global Medical Epidemiology and Big Data Analysis Lead at pharmaceutical giant Pfizer. “Unfortunately, many boards do not have an adequate understanding of the technologies to develop enterprise strategies.”
But while many organizations are still in the early stages of AI adoption, others are ahead of the curve.
Machine Leaning Driving AI Adoption in Business
ML algorithms are the most widely-used type of AI. But deep learning algorithms and chatbots are also popular, with 50% and 35% of respondents saying they use these technologies, respectively.
What’s more, a fifth of the executives we surveyed are assessing the business impact of AI technologies right now. Impressively, 25% thoroughly assessed whether they needed AI two years ago and a further 5% did so five years ago.
However, 49% of respondents said they won’t thoroughly assess whether AI and its subsets are benefiting their businesses for at least two years. This will give technology vendors time to develop their offerings and produce case studies as their technologies mature.
The contrast between early AI adopters and data leaders adopting a ‘wait and see’ approach shows how differently businesses are reacting to AI and ML technologies. This may be as a reflection of the different industries involved, or indeed the type of board and CEO in charge of each business.
In fact, securing board-level approval for AI and ML projects remains a stumbling block for data leaders in certain sectors. While 41% of respondents cited this as a key challenge, that figure jumps to 53% in the insurance industry. Either way, there’s clearly still work to be done when it comes to educating the broader organization about the benefits of AI.
Our survey results support this conclusion, with 64% of data leaders believing that it will get easier to secure board approval for these projects as understanding of the technologies involved improves.
“Many boards do not have an adequate understanding of the technologies to develop enterprise strategies, determine appropriate investment levels and effectively measure the impact of [AI] investments” – Christopher Boone, VP, Global Medical Epidemiology and Big Data Analysis Lead, Pfizer
That said, securing board approval is not the only hurdle data leaders must overcome to achieve enterprise-wide AI adoption. Almost two thirds of respondents are also wrangling with legacy technology, while 57% are finding it hard to secure the right staff for AI projects.
To help data leaders overcome these challenges, many technology vendors are developing AI products that aim to facilitate easy integration with older technologies. Others are working democratize data and create tools employees can use without help from specialist data scientists.
More than a third of respondents cited ‘picking the right problems to solve’ as a key challenge, suggesting that some organizations are still getting to grips with the technology themselves.
Data Leaders Expect AI Returns Within Two Years
Increasing efficiency is the most frequently cited benefit data and analytics leaders expect to see from AI, with 63% of respondents citing it as a key reason for adopting the technology. Meanwhile, 48% believe it will help them uncover useful insights, 45% say it will help them cut costs and 43% believe it will help their business make better decisions.
"We are using machine learning to identify buying patterns in our customers," explains Ricardo Rodrigues, Global Pricing and Operations Manager at vehicle manufacturer Opel Vauxhall. "Trying to say that, 'In this country, in this region we have a bigger appetite for Red Astras with special rims', for example."
He concludes: "That's what we are trying to find out – buying patterns that can help to know exactly how to locate those types of vehicles."
Perhaps the biggest surprise of this survey is that just 39% of data leaders cite ‘improving the customer experience’ as a key driver for AI adoption. Recommendation algorithms may have transformed the retail industry in recent years, but it looks like other client interactions and customer journeys may not be so easy to automate.
Almost all our respondents believe they will see returns from their AI investments within two years, with 59% predicting an ROI of at least 50%. Many data leaders will need to secure management buy-in, educate the wider organization and upgrade existing IT systems to realize those returns – but most clearly believe these investments will be worth it.
This article is an extract from our Four Data & Analytics Trends to Watch in 2020 report, published in association with Lynchpin. Click the image below now to access the full research and discover the other three trends.