Good at emergencies, bad at prevention

Every organisation we have worked inside is good at emergencies.

Something breaks on a Friday afternoon. A system goes down. A regulator asks a question nobody expected. A large client says, quietly, that they are reviewing their options. Within the hour there is a room. The right people are in it. The hierarchy suspends itself without anyone announcing that it has. Decisions that would normally take three weeks are made in twenty minutes, and by Monday morning the thing is fixed. Afterwards people say, with some justice, that the team really pulled together.

They did. That is not the interesting part.

The interesting part is the damp patch on the wall that somebody noticed eleven months earlier. It was mentioned. It went into a report. It was assigned to a working group. The working group met twice, produced a paper, and the paper went to a committee that had four other papers that quarter and a budget that was already committed.

Nobody was negligent. Every individual step was reasonable. And the wall came down anyway.

That is the pattern we want to spend this episode on. Organisations are superb in a crisis and poor at heading one off. The reason is not that leaders are short-sighted. Most of the leaders we know can describe the threat to their business with real precision. They can tell you what is coming. They can often tell you roughly when.

What they cannot easily do is act on it. Acting on it means spending certain money, certain political capital and certain attention on an uncertain problem. The machinery that makes a well-run organisation reliable is built to refuse exactly that trade. It is the same machinery. That is what makes this hard, and it is why willpower alone does not fix it.

The story most people have heard about Kodak is that it missed digital photography. That version is comfortable, and it is wrong, and the true version is far more useful to anybody running an organisation today.

In nineteen seventy-five, an engineer at Kodak named Steven Sasson built a working digital camera. It was the size of a toaster, it recorded to cassette tape, and it took twenty-three seconds to capture a single black and white image. He showed it to management. The reaction was not blindness. Sasson has described it as a version of: that is cute, but do not tell anyone about it.

Kodak went on to spend heavily on digital imaging over the following two decades. It held foundational patents. At various points it sold digital cameras in serious volume. This is not a company that failed to see the future. It is a company that saw the future clearly and could not bring itself to fund the destruction of its own present.

Because the present was extraordinary. Film was one of the most profitable products in industrial history. Kodak did not simply sell cameras; it sold the film, and then it sold the processing, and then it sold the paper the photographs were printed on. Every photograph anybody took anywhere in the world was, in effect, a small annuity. Digital photography ended all of that, and every honest analysis inside the company would have shown exactly that.

So imagine you are a Kodak executive in the early nineteen nineties. You can fund a digital programme that, if it works, replaces a high-margin consumable business with a low-margin hardware business. Or you can fund the film business, which is performing, which employs your colleagues, and which the market expects you to defend.

Every quarter, the second option is correct. Every quarter. And then one quarter it is catastrophically wrong, and by then the decision is no longer yours to make. Kodak filed for bankruptcy protection in January two thousand and twelve.

We think the useful lesson here is not about vision. It is that a series of individually rational decisions produced an irrational outcome, and no single meeting was the mistake. If you are waiting to catch your own version of this in a single bad decision, you will not catch it, because it will not arrive as one.

The second case is closer in time, and it removes the last excuse.

In two thousand and seven, Nokia was not a struggling company. It sold roughly half the smartphones on earth. It had the best hardware manufacturing operation in the industry, the deepest distribution, and enormous strength in exactly the markets that were about to grow fastest, including much of Africa. If you had asked anybody that year which company was best placed for the coming decade of mobile, the answer was obvious.

The iPhone launched that same year. Android followed shortly after.

Nokia's problem was not that it lacked engineers, or money, or warning. Internally, people knew. Nokia had touchscreen prototypes years before the iPhone shipped. What it had was Symbian: an operating system that worked, that shipped in enormous volume, that an entire organisation understood, and that was becoming the wrong answer.

In February two thousand and eleven, the incoming chief executive Stephen Elop wrote a memo to staff that leaked almost immediately. He described a man standing on a burning oil platform who has to choose between the fire and the freezing water below. His point was that Nokia was on a burning platform and had to jump.

We want to be careful here, because this is the part people get wrong. That memo is widely quoted as a moment of clarity. In our reading it is a moment of confession. By the time you need a burning platform metaphor to move an organisation, you have already lost the years in which moving would have been cheap. Elop was not early. He was accurate and late, and accurate and late is a description of most failed transformations we have watched.

There is a further cost that is rarely counted. Announcing that your platform is on fire tells your customers, your developers and your best staff that your current product is dead before the replacement exists. Sales of the existing devices fell hard afterwards. The urgency that finally unlocked the change also destroyed value while the change was happening. That is what waiting buys you: you do not merely act later, you act on worse terms.

Nokia sold its devices business to Microsoft in two thousand and fourteen.

Now the other side, because a diet of failure teaches very little.

Around two thousand and eleven, Adobe sold Creative Suite the way software had always been sold. A customer paid a large sum once, every eighteen months or two years, for a box or a licence key. The business was profitable and the product was dominant. Designers had no serious alternative.

Adobe decided to stop selling it that way, and to move to a subscription. Creative Cloud.

Consider what that decision actually required. Under the old model, revenue arrived in a lump at the moment of purchase. Under the new one, the same customer pays a small amount each month. In the short term your reported revenue falls, not because you are losing customers but because you have changed when the money arrives. You are asking your board, your shareholders and your staff to accept a visible decline in the numbers that everyone is judged on, in exchange for a business that is more durable later.

Customers were angry. There were petitions. There was a genuine argument, which we think was fair, that people who had bought software outright were being moved to renting it. The share price took a knock. Analysts questioned it publicly.

Adobe held. And the business that came out the other side was substantially larger and considerably more predictable than the one that went in.

The detail we keep returning to is the timing. Adobe did not do this because Creative Suite was failing. It did it while Creative Suite was winning. That is the whole point. The strength of the old business was what paid for the transition, and that strength was a wasting asset. Spend it while you have it, and you are choosing. Wait until it is gone, and you are reacting, and you will be doing the same work with less money, fewer people who believe you, and a shorter clock.

The example closest to home is also the one most often told wrongly, and the true version is better.

M-Pesa launched in Kenya in two thousand and seven. It is now so embedded in East African life that it is easy to assume it arrived as a grand strategic vision, drawn up in a boardroom, executed as designed.

It did not. It began as a much narrower idea, supported by British development funding and worked on with Vodafone: a way for microfinance customers to repay loans by phone. A sensible, modest, useful project.

What happened next is the thing worth studying. During the pilot, users began doing something the designers had not intended. They started using the airtime and the transfer mechanism to send value to each other. To family upcountry. To a spouse. To settle a debt between two people who had nothing to do with any microfinance institution.

That is the moment. In most organisations, that behaviour is a defect. It is off-specification, it complicates the reporting, and there is a project plan with a signed scope that says something different. There will be somebody in the room, usually a competent and conscientious somebody, whose job is to keep the project inside its scope.

Safaricom, under Michael Joseph, let the users redefine the product. They shipped the thing people were actually doing rather than the thing that had been approved.

And then they did the harder part, which nobody puts on a slide. They built agent networks. They dealt with cash management for thousands of small agents. They worked through regulation with the central bank in a country where nobody had done this before. Ordinary, expensive, unglamorous execution, sustained for years.

Two things travel from this. The first is that the strategic insight came from watching customers, not from a strategy offsite. The second is that Safaricom was not in trouble when it did this. Voice revenue was healthy. They built the next business on top of a business that was working, which is the only comfortable time to build anything.

We should be honest about why the preventive decision is hard, because pretending it is easy is how leaders end up feeling personally inadequate about a structural problem.

When you act early, you act on a threat that has not yet materialised. So you spend real money on a hypothetical, and you can never prove the counterfactual. If the crisis never comes, nobody thanks you. There is no photograph of the fire that did not happen. The colleague who fixed the Friday outage is visible. The colleague whose work meant there was no outage is invisible, and at budget time invisible work is the first to be cut.

There is a second problem, which is that the returns arrive after the leader does. Most senior tenures run three to five years. Serious transformation runs longer. So the person who bears the cost is frequently not the person who collects the benefit, and we should not be surprised that this affects behaviour. It is a structural incentive, not a character flaw, and it does not go away because everybody in the room is well intentioned.

The third is the quiet one. In a strong organisation, the argument for waiting is always well made. It is made by intelligent, experienced people with real evidence and no bad motive. The case for delay is not stupid. If it were stupid, none of this would be difficult.

We want to slow down and look at the actual room, because the refusal never announces itself as a refusal.

Here is how it goes. Someone brings a proposal to change something that is currently working. They have done the analysis. They believe it.

The first response is not opposition. It is a question, and it is a good question: what is the return, and over what period. This is a reasonable thing to ask. It is also, quietly, fatal, because the honest answer is that the return is uncertain and the period is longer than the planning cycle. The proposal is now competing against initiatives with clean numbers, and clean numbers win, because they are legible and because the people presenting them can be held to them.

The second response is a request for more work. Come back with a fuller business case. Again, entirely reasonable. But more work takes a quarter, and a quarter later the market has moved slightly, and the paper has to be redone, and the person carrying it has learned something about how much of their credibility this is going to cost.

The third response is the one that looks most like agreement. The proposal is approved in principle, and it is given a small pilot budget, and it is assigned to somebody who already has a full job. It has now been neither accepted nor rejected. It has been placed somewhere it can be pointed at when the question comes up, and where it cannot possibly succeed.

We have all sat in this meeting. Several of us have been the person asking the good questions. Nobody in it is behaving badly. Every one of those three responses is defensible, and would be correct if the proposal were ordinary.

Which is why we think the practical fix is procedural rather than attitudinal. If a proposal is genuinely about relevance in three years, it should not be assessed on the same terms as a proposal about efficiency next quarter. Different clock, different measure, different budget, and someone senior enough that requesting more work is not a costless move.

Organisations that do this well tend to have a small, explicit, protected line for it, with a named owner, ring-fenced from the annual squeeze. Not because the money is large. Because ring-fencing is the only known defence against a machine whose default answer, delivered politely and with good evidence, is later.

So what do we actually do with this.

The first thing is to stop treating early action as caution. It is not caution. Moving while you are strong is the most ambitious thing a leadership team can do, because it spends today's strength on tomorrow's relevance, and strength is the one asset that quietly expires.

The second is to put a name on it. Not a committee, not a workstream, not a steering group. One person who will answer for the outcome, who has access to the resources, and who has enough cover to be wrong once. We have never seen a change that worked without this, and we have seen many changes fail while everybody involved was working hard.

The third is a question we would put on the agenda of a good quarter rather than a bad one. Not what is going wrong. In a strong quarter, nothing is going wrong, which is precisely why the question is useless. Ask instead: what are we currently protecting that we would not build today. What in this business exists because it once worked, and is now defended because it is ours.

If you can answer that honestly in a room, you have found your damp patch. What you do next is a choice about whether you would rather be excellent in the emergency, or not have it.

We know which one gets the applause. We also know which one is the job.