Why do AI projects fail?
AI represents the most powerful capability organisations have ever had access to. So, business leaders naturally want to use it to tackle their biggest, most complex challenges. But few organisations have the operational maturity, data integrity, investment strategies and/or strategic partnerships in place to fully productionise, operationalise and subsequently support these behemoth projects.
Also, the technology is rarely a ‘silver bullet’. True progress can’t be made without understanding the multiple people and intricate processes impacted, and the success criteria everyone is striving to achieve.
CEO to CEO – where would you focus your AI efforts?
Recognise where you are on the maturity curve, and start small, but fast. Focus on some – or even one – of the many micro-processes that make the wheels turn in your business, affect lots of people, and don’t only shackle the workforce, but your productivity potential too. Map these workflows end-to-end, consult with impacted colleagues to drive early engagement, understand inefficiencies if there are any, and utilise AI’s capabilities to drive intelligent automation –and ROI – quickly.
The organisations that win, are those that nail the ‘POC to productionised’ flow, the fastest.
Can you give an example of a micro-process?
The creation of a board report might involve administrators through to C-suite execs, extracting data from various systems, exporting it into Excel, hand-cranking it, then dropping it into PowerPoint. Think of the time lost for something that isn’t actually complicated. Or expense claims. If 20% of a 1000-strong organisation claims expenses per month, it could take every employee two hours to file the claim and the expense team two to validate it. That’s 9,600 productivity hours that could be unlocked every year.
Who should drive AI adoption?
The most successful projects need a senior stakeholder to direct departments, drive collaboration and unity, and confidently build, articulate, socialise, and enforce the AI strategy, among the people who’ll live and breathe the change. Even the most technologically gifted individual cannot do this alone.
That senior stakeholder could wear one of many hats, but whether they’re the CEO, CIO, CTO, COO, director of business transformation, or another change leader, they must have an innovation appetite, and the ability to articulate – perhaps with the help of their chosen partner(s) – what budget is required and why. It also can’t be about ego.
What would you say to someone who believes their organisation is too complex to unpick?
It’s a common viewpoint, because businesses are distinct, with their own value chains, revenue generators, systems, and datasets. But behind every organisation, large or small, is a team of people, interwoven processes and interdependent technologies. So, if a more experienced AI specialist peels the layers back, they’ll likely have been exposed to the same infrastructure, the same problem statements and the same cultural challenges, many times. I’ve known the NHS face the same underlying difficulties as a Formula One team, for instance!
So, every business should be treated individually, but the same best-practice discovery methodology will help to unpick the complexities, without fuss.
In a noisy market, how do you identify an AI consultant you can trust?
AI is a big space. Many providers can help with things like Microsoft CoPilot and chatbots at the lower end of the maturity curve. But are they going to really move the needle for you in five years’ time?
Things start to get exciting when you combine chatbots with intelligent automation, agentic AI or predictive analysis. The right partner will have deep development experience across the AI landscape, an appetite to understand business-wide motivators, and the ability to design a strategic roadmap – rich with AI use cases that deliver almost immediate ROI while furthering the journey to your ‘north star’.
JOIN US for our next after work social – AI Chat Club – on 23 April, 5-8pm, Department @ Leeds Dock.
.svg)
