IWD 2025: Setting up for AI success

The idea of using AI to automate, enhance, and accelerate business processes may feel daunting, particularly in organisations where there is little or no experience or expertise with this emerging technology. It's not an uncommon situation. One-quarter (24%) of leaders surveyed for our Trends Report 2024-25 tell us that insufficient technological knowledge and experience is one of the major barriers to implementing new technology.

The secret to successful AI implementation is all in the groundwork and there are important things to bear in mind from the outset to set up for success:

1. It is important to understand the business function or process where AI is to be applied. Successful projects don't use AI for AI's sake, but we have discovered many organisations may be rushing in without completing this vital step. Our report shows almost half (46%) of organisations who failed with an AI project admit they did not fully understand the business problem they were trying to address.
2. Organisations need to secure sufficient expertise. Almost one in five (19%) leaders say they have tried and failed to implement an AI project this year, with 48% of these naming lack of skills and experience to manage the project as the primary reason for failure. Organisations can either recruit new talent, upskill existing IT teams, or consider outsourcing some or all of their IT to a managed services provider (MSP) that employees highly skilled, experts to do the heavy lifting.
3. Understanding the importance of data is next on the checklist. When dealing with AI, the quality of the data going in has a direct effect on the quality of data coming out. Our survey shows  one-third (35%) of organisations whose AI project was described as unsuccessful, say they did not sufficiently review and clean up data before embarking on the project. 29% also said they underestimated the amount of data-expertise and resource required for the project.
4. Bring everyone on board before implementing an organisational AI strategy and policy. Our data shows that organisations are at differing stages of maturity in doing this. The biggest group (22%) tell us that certain teams or departments are using AI tools independently, without a cohesive, company-wide strategy. Implementing AI in silos means that results are likely to be very much down to the expertise, or lack of it, of particular individuals or teams. When considering the development of their AI strategy, the largest group (27%) of respondents say that different functions within their organisation procure their own software, make their own policy, and have implemented their own guardrails and best practice. This means that the entire organisation could unknowingly be at risk, due to the actions of one team or function leader.

The successful implementation of AI must be approached as an organisation-wide strategy and ensuring all employees are sufficiently prepared may require some specific training and ongoing support. It is crucial that all users understand why and how AI will be used, and have the critical thinking skills to be able to assess where it can have real value. They don't need to be able to understand the nuts and bolts of the technology, but should have insights into its application. This means setting up markers for success and deciding how this success will be measured.

This won't happen overnight, but an organisation that adopts the right culture will be able to ensure that teams can collaborate on their AI journey, sharing lessons learned along with new  knowledge and expertise. By embedding an AI-ready culture within an organisation, leaders can move forwards.