Do you know what money can buy? Money can buy a lot of “smart” things. A business friend once gave me a tour of his renovated home. He installed smart lights, smart air conditioning and a smart television. A smart refrigerator that, I suspect, judges his diet. All controlled from his phone.
Then the power went out. And this thoroughly modern man stood in his own driveway, unable to open his own automatic gate, holding a phone as useful as a paperweight. We had a good laugh. But I have thought about that gate many times since, because it is a perfect picture of how most organizations are installing artificial intelligence. They bolt brilliant new gadgets onto an old design and call it transformation.
McKinsey, in its report “Growth favors the bold,” has a wonderfully sharp phrase for this. They call it “activity without architecture.” It means a company busily adds AI tool after AI tool, use case after use case, until it has a proud collection of clever features and almost nothing to show on the bottom line. Plenty of activity. No architecture.
The research is blunt about it. The high performers, the ones actually turning AI into money, are three times more likely to have fundamentally redesigned how their people work, rather than simply speeding up the old way. That is an eye-opening sentence for those of us who love buying tools and hate rethinking systems. And it exposes the next myth, which is my favorite because it lets so many leaders off the hook — the myth that this whole thing is a job for the tech department. Hand it to the CTO. Let the IT people sort it out. Wake me when it works.
The report says no. Growth comes when you treat this as a business transformation with people at the center, not a software rollout with humans in the way. At half of the top-performing companies, business leaders and technology leaders sit down and design the new pathways to growth together. Translation for those of us in the corner office: you cannot outsource the future of your business to the person who fixes your printer.
Then comes the number that made me turn off the sound of my TV and put down my coffee. For every one peso a company spends on an AI initiative, McKinsey says it may need to spend about three more pesos on change management: training people, redesigning workflows, changing incentives and helping employees actually use the technology. One part technology. Three parts people.
McKinsey’s newer work on agentic AI goes even further, describing a 1:3:5 pattern: one part technology, three parts process redesign and five parts capability building and adoption. So what can we derive from this finding? AI transformation is not primarily an IT project. It is an operating-model project. The software may be installed in weeks. The real work is getting people to change how decisions are made, how work moves and who owns what.
Two more myths we have to eradicate. The first is that AI takes years to pay off. McKinsey reports that leading companies often see early indicators within weeks and meaningful economic value within three to six months, provided they focus on high-value workflows instead of spraying AI across the organization like room deodorizer.
The second excuse I hear often in boardrooms is: “Our data is not ready yet.” I have heard this phrase so many times, and after so many meetings, I can tell when it’s coming before the words even arrive. It sounds responsible. Sometimes it is. But sometimes it is procrastination underneath impressive corporate speak. McKinsey says roughly 80 percent of companies do not need to rebuild their entire data architecture before starting. “Good enough” data can be sufficient to begin capturing value.
So here is the business lesson. Think of AI the way serious companies think about a new factory. You do not buy expensive machinery, drop it onto the old shop floor and assume productivity will magically rise. You redesign the line. You retrain the operators. You change maintenance routines. You revise quality controls. You rethink inventory flow. You measure output differently. Otherwise, you have simply purchased a very expensive machine for an outdated process.
The same applies to AI. Putting powerful technology on top of broken workflows does not create transformation. It creates faster dysfunction. That is why the better question is not, “What can this AI tool do?” Ask instead: “How should work in this company now be done?” Then decide where AI belongs.
The flashiest model will not necessarily win. The company with the best-designed system will. So stop collecting gadgets. Start redesigning the business. And prepare your people. Remember McKinsey’s 1:3:5 pattern: one part technology, three parts process redesign and five parts capability building and adoption.
Join and subscribe to Kongversations, the YouTube podcast that reached 10,000 subscribers in just its first six months. You can also catch the podcast “Inspiring Excellence” on Spotify, Apple Podcasts, Google Podcasts and other major platforms.