Stop Calling It an 'AI Strategy.' You Need an Automation Strategy.
I've sat through 14 "AI Strategy" presentations this year. They all look the same.
Slide 1: AI is transforming every industry. Slide 2: Our competitors are using AI. Slide 3: We need to use AI too. Slide 4: Here's a list of AI tools. Slide 5: Questions?
That's not a strategy. That's a panic attack in PowerPoint format.
Here's the thing nobody wants to hear: you probably don't need an AI strategy. You need an automation strategy—and AI happens to be one of the tools in it.
The Difference Matters
An "AI Strategy" starts with technology. What AI tools should we buy? Which vendor do we choose? How do we train people on ChatGPT?
An automation strategy starts with work. What are we doing today that's repetitive? What's slow? What's error-prone? What do our talented people spend time on that doesn't use their talent?
Then you ask: what's the right tool to fix this? Sometimes it's AI. Sometimes it's a simple script. Sometimes it's a Zapier workflow that a non-technical person can build in an afternoon. Sometimes it's just a better process with no technology at all.
I worked with a company that spent €80,000 exploring "AI-powered document processing." Their problem? Invoices came in three different formats and someone manually reconciled them. The solution? A €200/month Zapier workflow with basic OCR. No AI needed. Problem solved in a week.
They felt almost disappointed. "That's it?" Yes. That's it. Not everything needs to be smart. Some things just need to be automatic.
The Framework That Actually Works
Step 1: Audit your time. For two weeks, have every team track what they spend their hours on. I mean really track it, not "what sounds impressive in a meeting." The real stuff. The data entry. The copy-pasting between systems. The report formatting. The email sorting.
Step 2: Categorize each task. Can a human do this better than a machine? If yes, leave it. Can a simple rule handle it? Use basic automation. Does it require judgment but the judgment is predictable? AI might help. Does it require creativity, empathy, or novel problem-solving? Keep humans on it.
Step 3: Start with the things that make people groan. Not the sexy, innovative use cases. The mundane ones. The monthly report that takes someone two days to compile from five data sources. The customer onboarding emails that are 80% identical. The meeting notes that nobody wants to transcribe.
These aren't exciting wins. They won't get you speaking slots at conferences. But they'll save hundreds of hours per year and make your team actually like their jobs more.
Why "AI Strategy" Is Dangerous
When you frame everything as AI, you create pressure to use AI everywhere. That leads to solutions looking for problems. I've seen companies deploy AI chatbots that are worse than their existing FAQ page—because "we need AI touchpoints."
You also create fear. "AI Strategy" sounds like "robot replacement strategy" to employees, even if that's not the intent. "Automation strategy" sounds like what it is: making boring work disappear so interesting work can thrive.
Language shapes perception. Choose words that invite collaboration, not anxiety.
The Punch Line
The most successful AI deployments I've seen weren't called "AI projects." They were called "let's fix this annoying thing." The technology was incidental. The improvement was the point.
Stop writing AI strategies. Start writing lists of things that shouldn't require a human brain. The tools will sort themselves out.