"We need AI" is a bad strategy
There is a lot of noise around AI, and executives often set a goal to "adopt AI" so as not to fall behind. But that is strategy backwards. The right approach is to start with business tasks: where we lose money, time, customers, and can AI help here. AI is a tool for solving tasks, not an end in itself.
The difference in approaches
- Bad: "adopt AI because everyone does" to projects with no goal, money down the drain.
- Good: "support is overloaded to where can AI relieve the load?" to concrete value.
What it looks like
The board demands an "AI strategy". Where do I start so it is not for show?
Start with the business, not the technology. Ask: 1) Where are our most expensive bottlenecks — routine eating expensive peoples time? Slow processes? Overloaded departments? 2) Where do errors and delays cost money? 3) Where do we fail to scale due to manual labor? For each such point ask: can AI give measurable benefit here — save time, reduce errors, speed up? So the strategy becomes a list of concrete tasks with expected value rather than "we are for all things good with AI". That is what the board will both appreciate and be able to control.
Value comes in different forms
AI creates value in several ways: time savings (automating routine), cost reduction, faster processes, better quality, new capabilities (what was previously impossible). A good strategy is concrete: what value, where, for whom, and how to measure it. A vague "increase efficiency" is not a strategy.
Course rule: start with business tasks, not technology. Adopt AI where it gives measurable value, not "so we have it".
🧠 Where to start an AI strategy?