The first thing most companies do after committing to AI transformation is build a “company-wide AI strategy roadmap.” They hold executive workshops, commission consulting reports, and draw up three-year plans. A recent YouTube video (AXPERT) argues that this top-down approach can itself be a cause of AX failure. The point appears to be that most attempts that begin with a strategy document never carry through to execution on the ground.
Why “start at the edge” is persuasive
If the claim holds, the reason is not hard to guess. A company-wide strategy captures abstract direction, but the places where AI actually creates value are the repetitive steps inside a specific task and the bottleneck inside a specific team. Unless it starts from the concrete friction that frontline staff experience, even a finely crafted strategy document is unlikely to answer “what do we do first?” at the execution stage. What cases or data this video draws on for its argument, however, needs to be verified in the original video.
From what TECH2030 has observed across a range of AX journeys, company-wide strategy and frontline execution are closer to complements than opposites. A bottom-up approach that builds conviction through small wins (pilots) and then expands them into organization-wide direction often carries less risk than a top-down approach that tries to draw a perfect blueprint from the start. That said, bottom-up experiments need a minimum of principles and governance if they are not to fragment. “Do not write the strategy first” and “you do not need a strategy” are different claims, and we recommend watching the video directly to see exactly where it strikes that balance.
Sources: Why starting with a company-wide AI strategy fails | Enterprise AX starts at the edge (AXPERT)