How Businesses Can Get The Most Value From Artificial Intelligence
Many businesses are experimenting with artificial intelligence but lack an understanding of the use cases that deliver real business value.
That’s according to a new study from digital transformation firm Mindtree, which surveyed 650 global IT leaders on their use of AI in the workplace.
It found that 85% of organizations have implemented a data strategy and 77% have invested in some AI-related technologies, with only 31% achieving a return on investment.
The research shows that businesses can do more to gain increased value from their investments, although they need to focus on use cases that drive ROI.
Suman Nambiar, head of strategy, partners and offering for digital at Mindtree, said: “Artificial Intelligence (AI) is transforming the way businesses across the world are operating, for instance, helping them deliver much more personalized experiences and services to their customers.
“AI is delivering measurable business benefits to enterprises which experiment using agile, fast, value-driven innovation models, which can then be rolled out for wider use. But there is room for improvement.”
Defined use cases are key
When developing an AI strategy, just 16% of global enterprises are focusing on a pain point and defining a use case. And smaller organizations (13%) are less likely to consider business impact compared to more established firms (18%).
Mindtree warned that although organizations are experimenting with this technology, they have not “found the formula to deploy at scale and add significant value”.
But which areas should organizations prioritize? Functions such as sales (35%) and marketing (32%) appear to be providing the biggest gains, offering improved customer experiences.
While AI can deliver business benefits, Mindtree said the “majority of enterprises have yet to find a formula for repeatable success”.
It recommends that firms “experiment with different use cases and technologies with agile and rapid innovation methodologies”. That said, just 29% of enterprises are agile enough to experiment with AI.
Progressive enterprises, according to Mindtree, are focusing on use case definition, experimentation and operationalization for scale. These firms are spending 25% of their budgets on emerging technologies. The most popular technologies are machine learning (34%), chatbots (34%) and robotics (28%).
As businesses implement digitial innovations, they must upskill their teams to boost value. Out of the respondents, 44% said they’re hiring the best talent externally, 30% have partnered with academia and 22% run hackathons on new challenges.
Outdated IT infrastructure can also affect the performance of AI technology, particularly when it comes to data, architectures and systems – with over half of large enterprises (51%) saying they don’t understand the data technologies required.
Nambiar said that for businesses considering implementing AI, they need to strike the right balance between rapid experimentation and business value.
He added: “It is important to get started, iterate and show early successes rather than wait to get a grand ‘AI Strategy’ exactly right, but even for these early innovation efforts, the focus should be on use cases that will deliver some measurable business value.”
“The test of AI’s ability to deliver business impact will be when the early innovation efforts scale to enterprise-level, and for this to succeed, the data infrastructure, architectures and systems are critical, else there is the risk of early successes being mired forever in The Land of Perpetual Proof of Concepts.”
Getting AI right
Observing these findings, Andrew Rogoyski – innovation director at Roke Manor Research – said he agrees that experimentation is key when investing in emerging technologies such as AI.
“You have to know you’re solving the right problems, for the right reasons in the right way. You also need to understand the ethics of what you’re doing, especially privacy and data protection,” he commented.
He said the biggest barrier to success is finding the skilled experts who can make sure an AI or machine learning solution is being properly designed and implemented.
Rogoyski added: ”The demand for such skills is insatiable at the moment, so you have to be able to present recruits with a compelling story of why they should come and solve your particular problems. It’s an industry-wide problem.”
Discover Past Posts