Years ago, my mother bought a trinket while on holiday. It was a typical impulse purchase, jewellery you see and buy when you are in vacation mode. She was shocked to see how much it was worth when she sold the gold recently.

I own a few bitcoins and am bombarded on social media with predictions that it will rise to $1 million. Maybe so, but if it happens it is because one million will buy far less than today. I am not in a rush to live in a world where bitcoin buys so many dollars.

In timing terms, my Mum has done a bit better than simply matching inflation. Gold has been on a tear thanks to central bank purchases. At over $5,000 an ounce, even a holiday trinket fetches several hundred pounds.

This is a reminder that impulse decisions are not always regretted. Most of The Profit Elevator focuses on process improvements in search of enhanced outcomes. Yet there are situations where a CEO has to commit before the spreadsheets can justify the move. New markets can look unattractive on paper because the signals are weak and historical data is missing. Years later, forging ahead may prove to have been the best decision you made.

We have a number of construction industry clients at MSBC. These range from contractors and subcontractors, to building materials manufacturers and logistics companies. Most of the services we offer are relevant to other industries. Yet this niche developed because of one person’s family history in the glass and glazing business.

A construction client of ours faced a decision on whether to start retrofitting energy work five or six years ago. It was hard to model demand. Regulations were uncertain, the potential pricing unclear and the supply chain was immature. A cautious analysis would have concluded the opportunity was too vague.

Yet the CEO decided to go ahead. Governments were tightening carbon rules, energy costs were headed upwards and there were plenty of buildings in need of upgrading. The gut feeling was strong.

As it happens, waiting for proof would have meant arriving after competitors had secured relationships and built expertise and reputation. In these situations the real decision is timing. By the time the numbers are clear, the opportunity is largely gone.

Customers are often poor at articulating what they will value in the future. This is particularly true for new products or new combinations of services. You might see something that early customer feedback does not fully validate.

A while back, we took the decision to bundle advisory, implementation and operational support for certain types of automation. Some prospects said they only needed one service, and sales wanted them to remain unbundled to make deals easier. The rational move appeared to be giving customers what they wanted.

Yet having sold these services, we understood an integrated solution solved a deeper problem of fragmented accountability. The bundled model felt right even though it complicated sales in the short term. While we missed out on some smaller opportunities, we ended up owning the entire client relationship with larger and more profitable clients.

This kind of decision is difficult to prove in advance because the value only becomes clear once the model exists.

In both of these cases, the leader is operating in a situation where the outcome cannot be calculated in advance. There are markets that are not yet formed, customer needs that are not fully expressed, or capabilities whose value depends on future strategy. In these situations, excessive analysis creates a false sense of precision.

The best way to approach similar opportunities is through scenario planning. When accurate predictions are impossible, your aim is to understand which assumptions matter most. You are testing the resilience of your strategy should those assumptions prove wrong.

Most scenario exercises fail because they test irrelevant variables. Focus on two or three forces that could reshape the market. These might include regulation, technology adoption, cost curves for key inputs, customer behaviour shifts and competitive structure. For example, in AI-driven services the uncertainty is how quickly customers trust automation with important decisions. The technology itself is rarely the problem.

Scenarios are not stories. They are only useful if they have actionable insights. If a situation does not change a decision, then it is not useful. Your aim is to make decisions that work across multiple possible futures.

There are many other considerations when scenario testing. The end purpose, however, remains unchanged. You discover the assumptions your strategy depends upon and the signals that will tell you if it is working. You cannot remove uncertainty but should be comfortable acting despite it.

These business decisions are not pure impulse in the manner of a holiday souvenir purchase. They are, however, instinctive. The CEO’s judgement is in play and informs decisions that humans will be making for the foreseeable future.