A new study from Gartner, a research company, shows a clear link between how good a company is at using AI (Artificial Intelligence) and how long their AI projects actually last. The study found that almost half (45 per cent) of leaders in companies that are very good at AI said their AI projects stay in use for three years or more, helping them get lasting benefits. This is a big difference compared to companies that are not very good at AI, where only 20 per cent of projects last that long.
The study was done in late 2024 and asked 432 people from companies in the US, UK, France, Germany, India, and Japan. Gartner measured how “mature” a company’s AI efforts were using a special model. Companies that were “highly mature” (very good at AI) scored high (around 4.2–4.5 out of 5), while “low-maturity” companies (not very good at AI) scored much lower (around 1.6–2.2).
Trust is key for AI success
Gartner found that strong AI companies make their projects last longer by choosing AI tasks that clearly help the business and are possible to build. They also set up good rules and engineering practices.
“Trust is one of the big differences between an AI project succeeding or failing,” said Birgi Tamersoy, a top analyst at Gartner. The study showed that in 57 per cent of highly mature companies, business teams trust and are ready to use new AI tools. This is much higher than in low-maturity companies, where only 14 per cent of teams trust new AI solutions.
Tamersoy explained that “building trust in AI makes people use it, and using it is the first step to getting value, so it really helps with success.”
Big hurdles: Data and security
No matter how good a company is at AI, getting good quality data and having enough of it are still top problems. 34 per cent of leaders from low-maturity companies and 29 per cent from high-maturity companies pointed this out.
However, other big problems were different. For companies good at AI, 48 per cent of leaders said security threats were a main barrier. But for companies not so good at AI, 37 per cent of leaders struggled most with simply finding the right way to use AI.
The study also found that measuring how well AI works helps get better results. Companies good at AI can get big benefits from their projects over time because they regularly check how AI helps them financially and measure its impact on customers. 63 per cent of leaders from highly mature organisations do financial checks on risks, calculate how much money AI brings in (ROI), and clearly measure how customers are affected. This helps them keep AI successful.
Also, 91 per cent of leaders from highly mature organisations said they already have special leaders just for AI. These AI leaders focus on encouraging new AI ideas (65 per cent), providing the tools and systems for AI (56 per cent), building AI teams (50 per cent), and designing how AI systems work (48 per cent).
Finally, almost 60 per cent of leaders in highly mature organisations have gathered their AI plans, rules, data, and tools into one central place. This helps them be more consistent and efficient with their AI efforts. As Tamersoy noted, “This shows a smart way of managing AI, which needs dedicated AI teams.”
