The slowest to move are not always wrong Japanese businesses are being criticised for dragging their feet on AI. The usual explanations surface: cultural conservatism, risk aversion, consensus-driven decision-making that takes forever. I have been watching this story unfold and thinking about how many of my clients in London show the exact same patterns. They just dress it up differently. The problem is not Japan. The problem is how people weigh decisions when uncertainty is high. And there is a useful framework hiding in this story that most people running their own thing completely miss. Risk aversion is not the same as being careful Let me be direct: avoiding AI because it feels risky is not careful. It is lazy thinking disguised as prudence. Careful means you have examined the options, understood the trade-offs, and made a deliberate choice. Risk aversion often means you have done none of that. You just said no because saying yes required effort. I worked with someone last year who had refused to automate their client onboarding for eighteen months. When we finally mapped it out, the actual risk was about £200 in software costs and four hours of their time to test it. The perceived risk? "What if it breaks and clients get confused?" That is not a risk assessment. That is anxiety wearing a business suit. The decision matrix nobody uses properly Most people have heard of a decision matrix. You list your options, define criteria, score them, and the answer emerges. Simple. Except almost everyone uses it wrong. The mistake is weighting criteria based on what feels important rather than what actually matters to outcomes. Japanese firms reportedly weigh "avoiding mistakes" extremely heavily. This sounds reasonable until you realise that not adopting AI is also a mistake. It just happens later and looks like stagnation rather than a single bad decision. When I run decision matrices with clients using the tool at alira.london, I force them to include "cost of inaction" as a weighted criterion. Most people forget this entirely. They compare Option A versus Option B versus doing nothing, but they score "doing nothing" as zero risk. Doing nothing is never zero risk. It is just risk that compounds quietly. What Japanese hesitation actually teaches us Here is what I take from this story. First, consensus-driven decisions tend to favour the status quo. If you need everyone to agree before moving, the path of least resistance is always to wait. I see this constantly in partnerships and small teams. Three people need to agree, so nothing happens for months. Second, cultural narratives shape risk perception more than data does. Japanese business culture has a strong narrative around avoiding failure. British business culture has a different one, but it still exists. "We have always done it this way" is just as powerful here. The narrative is different; the paralysis is identical. Third, late adopters pay more. The firms that wait until AI tools are "proven" will find themselves buying from vendors who have already raised prices, competing against rivals who have already reduced costs, and training staff who are already behind. There is a 23% cost premium on average for late technology adoption in professional services. I have seen the numbers across multiple sectors. How to actually use a decision matrix If you are weighing whether to adopt any new system, tool, or process, here is how to do it properly. Start by listing your options, including doing nothing. Most people skip this. Doing nothing is always an option, and it needs to be scored honestly. Next, define criteria that actually matter to your business outcomes. Not what sounds professional. Not what your industry expects. What moves money, time, or quality for your specific situation. Then assign weights before you score anything. This is where most people cheat. They score first, then adjust weights until the answer they wanted emerges. That is not decision-making. That is rationalisation. Finally, include "cost of inaction" and "opportunity cost" as explicit criteria. Weight them appropriately. For most decisions involving technology, these should be at least 20% of your total weight. The Decision Matrix tool at alira.london walks you through this structure. It forces the discipline that most people skip when doing it on paper. The real question underneath When someone tells me they are "being cautious" about AI or automation, I ask them one question: what specifically would need to be true for you to move forward? Usually they cannot answer. They have not defined their own criteria. They are just waiting for a feeling of certainty that will never arrive. Certainty is not coming. Not for AI, not for any significant business decision. The firms that win are not the ones who wait for certainty. They are the ones who define acceptable risk, make a decision, and adjust as they learn. Japanese firms will eventually adopt AI. By then, the competitive advantage will have shifted to something else, and they will be behind on that too. The pattern repeats. Do not let it repeat in your business. What to do this week Pick one decision you have been avoiding. Not the biggest one. Something medium-sized that has been sitting in your head for more than a month. Run it through a proper decision matrix. Use the tool at alira.london if you want structure, or do it on paper. Either way, include "cost of inaction" as a weighted criterion. Score it honestly. Then make the decision by Friday. Not "start thinking about it more". Decide. The point is not to get it perfect. The point is to break the pattern of indefinite delay dressed up as caution. You will learn more from one decision made and adjusted than from six months of waiting for certainty.