Most companies approach AI backwards. They buy the technology first and then try to find a business case for it. The result is expensive software nobody uses and a team that quietly goes back to doing things the old way. The real ratio is 80 percent business change and 20 percent technology.
The result is a pattern that repeats across industries. A company buys an AI tool, deploys it to a team that was not consulted, measures nothing, and declares six months later that "AI did not work for us." The technology was never the problem.
Why Most AI Investments Fail
The typical approach goes like this. Someone reads about AI. They attend a webinar. They buy a subscription. They hand it to a team and say "use this." There is no clarity on which problem it solves, no baseline measurement to compare against, and no plan for how workflows need to change.
McKinsey's data shows that the companies seeing real returns from AI are doing the opposite. They start with a specific, costly problem. They map the current workflow. They define what success looks like in numbers. Then - and only then - they choose the technology.
How Most Companies Approach AI
- Start with a tool or vendor demo
- Look for problems it might solve
- Deploy to a team without training
- Measure nothing before or after
- Declare success or failure in 90 days
- Move on to the next shiny object
How the Best Companies Approach AI
- Start with the most expensive problem
- Map the current workflow end to end
- Define success metrics before buying
- Involve the team that does the work
- Redesign the workflow around AI
- Measure, adjust, and expand deliberately
What 80% Business Means
When Kutcher says 80% business, he means the work that has nothing to do with code, algorithms, or AI models. He means the organizational change management that determines whether a technology investment produces returns or collects dust.
The 80% includes process mapping - understanding how work actually flows through your business today. It includes team buy-in - the people doing the work need to understand why things are changing and how it affects their role. It includes success metrics - what number needs to move, by how much, in what timeframe.
It also includes workflow redesign. You cannot bolt AI onto a broken process and expect good results. If your current intake process involves four handoffs and two spreadsheets, automating part of it just makes the mess move faster.
The 20% Is the Easy Part
When the business case is clear, the technology selection becomes straightforward. You know what problem you are solving. You know what success looks like. You know which team is affected. The vendor conversation becomes specific instead of speculative.
This is why the best CEOs in McKinsey's research spend less time evaluating AI products and more time understanding their own operations. The technology is a commodity. The business clarity is the competitive advantage.
What Good CEOs Do Differently
McKinsey found that the most effective leaders in the AI transition share a counterintuitive habit. They spend time with junior team members - the people who actually do the day-to-day work. Not in a formal meeting. In the flow of real operations.
This matters because the CEO who understands how the front desk actually handles a new patient call is in a fundamentally different position than the CEO who only sees the quarterly metrics. The first one can identify what to automate. The second one can only guess.
A Four-Step Framework
You do not need a CTO or a consulting firm. You need clarity on which problem costs you the most money. Here is how to apply the 80/20 rule in a business with 5 to 50 employees.
Find Your Most Expensive Problem
List the five activities that consume the most staff hours per week. Calculate the labor cost of each one. The most expensive repetitive task is your automation candidate. This step alone eliminates 90% of bad technology purchases.
Map the Current Workflow
Sit with the person who does the work. Watch them do it. Document every step, every handoff, every workaround. You will discover inefficiencies that have nothing to do with technology - and those should be fixed first. Do not automate waste.
Define Success Before You Buy
What number needs to change? By how much? In what timeframe? "We want to reduce missed calls from 35% to under 10% within 60 days" is a success metric. "We want to be more efficient" is not. Specific targets prevent both overspending and premature disappointment.
Involve the Team First
Before you deploy anything, tell your team what is changing and why. Explain which tasks are being automated and what they will focus on instead. The number one reason AI projects fail is not bad technology. It is people who were not consulted resisting a change they do not understand.
You Do Not Need a CTO
The 80/20 insight is actually good news for small business owners. It means the most important work - the 80% - is something you already know how to do. You understand your operations. You know your team. You see where time gets wasted.
What you need is not a technology expert. What you need is the discipline to start with the problem instead of the product. The best AI investment a 10-person business can make is not a software subscription. It is a clear-eyed understanding of where money and time are being lost.
Final Takeaway
McKinsey's research is clear. The companies getting real value from AI are spending 80% of their effort on business fundamentals - process clarity, team alignment, measurable goals - and 20% on the technology itself. Most companies invert this ratio and wonder why the investment did not pay off.
Start with the problem that costs you the most. Map the workflow. Define success. Talk to your team. Then choose the tool. That sequence is the difference between AI that works and AI that gets abandoned.