Why Most AI Projects Fail Before They Even Start

Dr. Niklas Richter · 2026-09-08

Digital transformation is on every company's agenda. Yet most projects stall, run late, or quietly deliver nothing close to what was promised. So where does it actually go wrong? That's the question at the heart of a recent conversation between Virginie Claes and Nick Allaert, co-founder of SANDT. Nick has spent years helping companies move from strategy slides to real, working technology. And he's seen the same patterns trip up businesses again and again. ### The Gap Between Strategy and Execution Here's the thing. Most companies don't fail because they picked the wrong goal. They fail somewhere between the plan and the people who have to carry it out. A strategy looks clean on paper. Then it hits real teams, real workflows, and real resistance. Suddenly, the roadmap that made perfect sense in the boardroom starts falling apart on the floor. Nick points out that this isn't a technology problem. It's a people problem wearing a technology costume. - Leaders assume the tool will fix the process - Teams get handed new systems without real training - Nobody owns the outcome once the consultants leave Sound familiar? ### Why AI Is Overhyped and Underused Let's talk about AI for a second. It's everywhere right now. Every vendor promises it'll change everything. But Nick's take is refreshingly blunt: AI only creates value when it's woven into how a business actually runs day to day. Not bolted on. Not tested in a pilot that goes nowhere. Integrated. > "Choosing the right technology matters far more than choosing the newest one." — Nick Allaert That line stuck with me. Because the pressure to adopt the shiniest tool is real. But the newest thing isn't always the thing your team can use, trust, or maintain. ### Bridging the Business-IT Divide One of the biggest culprits? The gap between business and IT. They often speak different languages, chase different metrics, and blame each other when things stall. Nick's approach at SANDT is to close that gap early. That means getting both sides in the same room from day one, not after the budget's been spent. - Define what success actually looks like - Agree on who owns each decision - Build in feedback loops before launch, not after ### Change Management Isn't a Buzzword You can have the best system in the world. If your people don't embrace it, it's just an expensive tab nobody opens. Successful change management starts with listening. Why are people resistant? Usually it's fear, not stubbornness. Fear of looking incompetent. Fear of extra work. Fear of being replaced. Address those fears directly and adoption stops being a fight. ### What Actually Works So what separates the wins from the wreckage? A few things show up again and again. - Start small and prove value fast - Involve the people who'll use the tool daily - Pick technology that fits your reality, not a demo - Measure return, not activity None of this is glamorous. But it works. For entrepreneurs and business leaders eyeing digital innovation, the lesson is clear. The goal isn't to adopt AI first. It's to adopt it right. Get the people part right, and the technology finally has a chance to do its job.