Enterprise AI has consumed hundreds of billions in investment over the past two years. The technology has advanced dramatically. Yet according to McKinsey’s global survey, fewer than one in three AI initiatives are delivering meaningful results. In our conversations with data leaders across the DACH region, we hear the same frustration: the platforms are capable, but the rollouts are failing.
The issue sits squarely with talent strategy. Most organisations have hired data scientists and ML engineers, ticked the headcount box, and assumed the rest would follow. It hasn’t. What we see repeatedly is a shortage of people who can sit between the technical teams and the business units, translating AI capability into operational change. These roles require a rare combination: deep technical fluency, commercial instinct, and the patience to manage cultural resistance. They’re difficult to hire because most candidates skew heavily to one side or the other.
The DACH market has a specific version of this problem. German enterprises tend to move cautiously, preferring proven approaches over experimentation. That mindset works well for traditional IT transformation but creates friction with AI, where iteration and tolerance for early-stage imperfection are essential. We’re seeing demand spike for candidates who have already led AI implementations through the messy middle phase, rather than those who have only worked on greenfield pilots.
For hiring managers, this shifts the brief considerably. The question is no longer whether someone can build a model. It’s whether they can navigate a sceptical organisation, secure buy-in from operations teams, and design processes that allow AI to improve over time. We’re advising clients to prioritise candidates with change management backgrounds alongside technical credentials. The organisations treating AI adoption as a pure engineering problem will keep hitting the same walls.
Prompted by reporting from Datanami. Read the original article.