Two-in-five businesses expect major AI benefits but lag on ROI metrics
Three years on from ChatGPT being made accessible to the public, talk around AI is finally moving beyond hype, toward measurable results. But even as the vast majority of organisations say the technology is now saving them time at work, more than two-in-five have no ROI mechanisms in place to actually measure the benefits.
Since the public launch of ChatGPT, the excitement surrounding the “game-changing tool for businesses” of AI has reached feverish new levels. The reality has often failed to live up to that grand billing, though. An estimated 85% of AI-related business projects fail outright, while only 10% ever see a return on the investment.
But despite the high rates of failure, and heightened concern among investors that the demonstrable results of AI do not justify the capital sunk into various unicorns and tools, many organisations are neglecting the need to measure its impact. Wavestone researchers spoke to 500 technology and business leaders based in the US, the UK, France, Germany, Singapore, and Hong Kong – and found that while many felt they were moving beyond teething problems to actually demonstrate value with AI, a sizeable minority did not have metrics in place to determine its return on investment (ROI).
When asked what they were hoping to gain from adopting AI in the next two years, the executives gave some familiar answers. For the longest time, efficiency savings, customer service and productivity have been the ‘potential’ at the heart of AI’s sale’s pitch.
It is not especially surprising to see that in two years, 39% of respondents expect AI will streamline their operations, then. Nor is it unusual to hear that 37% said they thought in 24 months it could boost customer satisfaction, or help create better products, faster.
Expectations meet reality
Overall, 80% said they expected at least one operational benefit. Meanwhile, 73% expected at least one business related benefit – including a competitive advantage or revenue growth. And amid a time of economic uncertainty, that is the kind of thing investors will want to hear – as is the assertion that some of those impacts are already manifesting.
A 46% portion of respondents said they were “behind most competitors” now, in terms of their AI adoption, down from 75% in 2024. Meanwhile 99% claimed their organisation had saved time in some way by deploying AI – with the largest 32% saying it had driven greater value through revenue growth and client satisfaction.
What will please investors less, though, is this deep into the endless amounts of time and capital sunk into the ‘AI revolution’, the actual extent to which any pay-off has been realised is still up for discussion. It is one thing to say “AI is making a difference” anecdotally – but as 46% of organisations also admitted they still do not have a structured ROI measurement framework in place, proving whether that potentially slim improvement was really worth it is another question all together.
Of those, 32% said they tracked AI’s returned value “informally” through expert assessments, while 10% evaluated the technology on a case-by-case basis, without a systematic view. The efficacy of either approach is hard to pin down – as it depends on the quality of the experts leveraged, or the specifics of each case as to whether those results can translate to a wider organisational goal. Meanwhile 4% said they had no process at all.
