Why Most AI Projects Fail (And What Actually Works)

ยท
Listen to this article~4 min

Most digital transformation projects fail in execution, not strategy. Nick Allaert of SANDT explains why AI only works when it's truly integrated into daily operations.

### The Gap Between Strategy and Execution Digital transformation sits at the top of nearly every company's agenda. Yet most projects run late, blow budgets, and deliver a fraction of what was promised. According to McKinsey, roughly 70% of digital transformations fall short of their goals. So why do so many smart strategies die in execution? That's the question Virginie Claes digs into with Nick Allaert, co-founder of SANDT, on a recent podcast episode that's packed with hard-won lessons for entrepreneurs. The short answer: technology isn't the hard part. People are. ### The AI Hype Trap AI is everywhere right now. Every vendor promises it'll revolutionize your business overnight. But Nick makes a point that's easy to forget in the noise: > "Choosing the right technology matters far more than choosing the newest technology." That's a counterintuitive idea in a world obsessed with being first. But it's also why so many AI pilots stall. Companies bolt AI onto broken processes and expect magic. What they get instead is expensive confusion. AI only creates real value when it's woven into the daily rhythm of how a business actually operates. Not as a side project. Not as a shiny demo. As part of the workflow. ### Why Employees Resist (And How to Fix It) Change management gets treated like a checkbox. Send an email, hold a training session, done. That approach fails almost every time. - **People fear what they don't understand.** If your team doesn't see how AI helps them personally, they'll quietly sabotage it. - **The business-IT gap is real.** When IT builds without talking to operations, you get tools nobody uses. - **Trust takes time.** Nick's team at SANDT guides companies from strategy all the way through full implementation, precisely because handoffs are where projects die. Here's the thing: your employees aren't the obstacle. They're the key. Involve them early, ask what slows them down, and let them shape the solution. ### From Advice to Actual Implementation SANDT's model is a useful blueprint. They don't just hand over a strategy deck and walk away. They stay through deployment, training, and the messy middle where real change happens. That matters because the gap between a plan and a working system is where most budgets disappear. A 2023 Gartner study found that organizations often underestimate integration costs by 40% or more. Nick's advice for business leaders is refreshingly simple: start with the problem, not the tool. Ask what's actually costing you time or money. Then find the simplest technology that solves it. ### A Practical Takeaway for Founders If you're investing in digital innovation this year, here's a quick checklist: - Define the business outcome before you evaluate any tool - Get your operations team in the room from day one - Pick technology your people can actually adopt - Budget for training, not just licenses - Measure results in months, not days None of this is glamorous. But it's what separates companies that get a return on AI from those that get a cautionary tale. The full conversation with Nick Allaert covers far more ground, including change management tactics and the business-IT divide. If you're a founder or business leader planning a tech investment, it's worth your time. This podcast is an initiative from Creatic, a team that helps small and mid-sized businesses build their brand online through strong branding, thoughtful marketing, and high-performing websites. From brand strategy to a site that actually converts, they help companies grow with visibility and recognition.