Worse before better: why the right fix rarely feels right at first

The previous article in this series ended on a claim that's easy to read past: that the best solution to a deeply rooted problem often makes things visibly worse before it makes them better, while the quick fix often feels good right away – and then reverses. That deserves its own article, because it's one of the most underrated reasons good decisions get abandoned too early, and bad ones get to run too long.

Two curves, same decision

Call them Worse Before Better (WBB) and Better Before Worse (BBW). Both describe what happens after a decision, over time, but in opposite order:

A WBB fix addresses a root cause. Because it disrupts a structure the organization has already adapted to – processes, behaviors, even skills built up around the old way of doing things – the cost shows up immediately, while the benefit arrives with a delay. Paying down technical debt in a software product is a clear example: development speed drops first, since time goes into cleanup instead of building new features, before it rises to a level that would otherwise never have been possible.

A BBW fix does the exact opposite. It relieves a symptom quickly, often by borrowing resources or trust from the future. Heavy discounting to hit a quarterly target is a textbook example: sales rise immediately, but customers learn to wait for the next discount, margins erode, and the brand's price integrity is devalued – a decline that shows up far later than the initial rise.

What makes these curves dangerous isn't that they exist. It's that they look identical at the moment a decision is made. Neither curve has a direction yet; it only has an intention. The difference only becomes visible afterward, if you know what to look for.

Why the WBB curve gets abandoned too early

This is the more acute risk for most organizations: an action that addresses the root cause, and therefore inevitably makes things look worse for a while, gets cancelled before it has time to pay off. The dip gets read as proof the decision was wrong, rather than as an expected part of it. The organization reverts to the old, symptom-relieving path – which happens to be the same path the "Fixes that Fail" archetype from the previous article describes.

It's no accident this happens so often. Decision-makers are typically rewarded on quarterly or annual results, while a genuine structural improvement often has a much longer time horizon than that. Whoever holds the line through the dip is rarely the one who gets credit once the curve turns, which makes it rational in the short term – for the individual decision-maker – to choose the BBW path even when they know better.

Why the BBW curve doesn't get caught in time

The reverse risk is at least as costly: an action that appears to work perfectly at first gets assumed solved, and the organization stops looking for the underlying cause. Because the decline arrives with a delay, and often in a different part of the organization than the one that made the original decision, the connection between cause and effect is hard to see once it becomes visible. By that point, more has often been built on the fragile foundation, making the eventual correction more expensive than if the shortcut had never been taken.

The worst case: when the decision-maker has already moved on

There's a variant of the BBW curve that's especially hard to guard against, because it exploits not just a delay in the system but a delay in who's still around to own the consequence. Maintenance is a recurring example. A CEO who cuts deeply into maintenance spending on facilities, fleet, or infrastructure sees the result immediately as lower costs and higher margin – numbers that often trigger a bonus, and that make the person attractive for the next CEO role or board seat. The assets keep running for a while, because most systems have built-in slack that absorbs a period of neglected maintenance without it showing. But the slack runs out eventually, and when it does, the bill – in the form of sudden breakdowns, expensive emergency repairs, production stoppages, or in the worst case safety incidents – often arrives years later, and almost always after the person who made the decision has already moved into a new role.

This isn't just an anecdote. The phenomenon is common enough to have its own name in the research literature: the horizon problem, which describes how executives whose remaining time in a role is short or uncertain systematically tend to cut maintenance, R&D, and other long-payback investments in favor of actions that show up in the results before they themselves leave.

What makes this variant especially serious is that it breaks the natural correction mechanism otherwise built into most systems: that whoever makes a bad decision is normally also the one who has to deal with its consequences, which over time creates a built-in incentive not to make it. When the decision-maker and the person who inherits the bill are two different people, in two different roles, sometimes at two different companies, that feedback disappears entirely. The next employer sees the same thing the previous board once saw: an executive with a strong track record of fast results improvements – without necessarily seeing what those improvements cost down the line.

That's also why the question of where the cost lands, raised above, needs to be paired with a question of who carries it: will the person making this decision still be here, in this role, when the bill comes due? If the answer is probably no, good intentions aren't enough – what's needed instead is longer bonus vesting periods, measuring the underlying condition of assets rather than just the results they currently generate, and follow-up that outlasts any individual executive's time in the role.

Telling the two curves apart in advance

In practice, you rarely get to wait and see which curve you're on – by then the decision is already made and the effects are already in motion. What you can do in advance is ask a simple question about the action itself: does this fix the cause, or does it just buy time by moving the cost to a later point in the system? If the answer is the latter, a BBW curve is the reasonable expectation, no matter how good the numbers look next quarter.

Another tell is to look at where the cost and the benefit land within the organization. If the benefit shows up in the budget that made the decision, but the cost lands somewhere else – a different department, a different year, a different manager – that's a watch-closely sign rather than a green light.

Knowing which curve you're on is rarely enough to make the right call by itself. What actually determines whether an organization chooses WBB over BBW, and has the staying power to hold through the dip, is whether there are concrete methods for identifying root causes and driving through fixes that hold. That's the subject of the next and final methods article in this series.