Problem-solving methods: the tools that actually address what we've covered

The previous four articles in this series have been diagnostic. We've covered why an action can backfire against its own purpose (the cobra effect), why that happens (systems thinking, loops, and delays), and why the best fix often feels worst at first (worse before better). What's been missing is what to actually do about it. This article collects four methods – not as a complete toolkit, but as four targeted answers to four specific traps we've already identified.

Against the cobra effect: root cause analysis

The cobra effect arises, as article 2 showed, when you measure or fix a proxy instead of the actual cause. The countermeasure is simple to describe, if not always simple to do: ask why, repeatedly, until you reach something you can actually act on.

5 Whys, a technique central to the Toyota Production System, works by asking "why did this happen" five times in a row – each answer becoming the basis for the next question – until you land beyond the symptom and in the structural cause. The fishbone diagram, developed by quality engineer Kaoru Ishikawa, does the same thing visually: possible causes of a problem get sorted into categories (method, people, materials, machines, and so on) to avoid settling on the first, most obvious explanation.

The point of both methods is the same: they force in an extra step before acting, a step where you actively look for alternatives to the explanation that happens to be easiest to measure.

Against system blindness: map the loop, find the leverage point

Article 3 introduced loops and delays as the explanation for why cause and effect are rarely a straight line. Knowing that isn't enough on its own – the next step is to actually draw out the loop for the specific problem at hand: which parts affect each other, in which direction, and with how much delay. Even a simple whiteboard sketch, with arrows between the factors involved, makes invisible connections visible in a way words rarely do.

The next step is to ask where in that loop an intervention actually makes a difference. Donella Meadows argued in her classic essay on leverage points that organizations almost always intervene in the wrong place – adjusting a number or a rule (low leverage) instead of changing the information flows, incentives, or underlying goals of the system (high leverage). Mapping the loop isn't just an exercise in understanding, then; it reveals where in the system an action actually bites, instead of just shifting the problem one step sideways.

Against unpleasant surprises: the pre-mortem

Article 1 was about thinking through several scenarios before acting, rather than just following one straight plan. A concrete method for doing that systematically, not just as a general stance, is the pre-mortem, developed by psychologist Gary Klein.

The technique is simple: before a decision or project is set in motion, the team gathers and imagines that it has already failed – catastrophically, a year down the line. Each person then writes down, individually, why they think it went wrong. By starting from the assumption that the failure has already happened, rather than asking "what could go wrong," it becomes both easier and more socially acceptable to raise concerns people would otherwise keep to themselves – and research has shown the method substantially increases the number of risks identified compared to a standard risk review.

Against fixing the wrong thing: find the actual constraint

Article 4 was about telling a genuine fix (worse before better) apart from a shortcut that just moves the cost further down the timeline (better before worse). One way to make that distinction concrete, not just a question you ask yourself, is the Theory of Constraints, developed by physicist and management thinker Eliyahu Goldratt.

The core idea is that any system – a production line, a sales team, an organization – has, at any given moment, exactly one actual constraint that determines total capacity, no matter how many other parts get improved. Optimizing anything other than that constraint creates a feeling of improvement without actually increasing overall capability – a variant of the better-before-worse pattern described in the previous article, just aimed at resources rather than trust or brand. The method forces the question: what's the actual constraint right now, and does this action fix it, or does it just optimize something that already has enough capacity?

Four methods, one common thread

What ties these four methods together isn't that they belong to the same school, or even the same field – they come from quality engineering, system dynamics, cognitive psychology, and production management. What ties them together is that each one forces a question that's otherwise easy to skip: is this the real cause, the real connection, the real risk, the real constraint – or just the most available one?

All four methods also share something worth noticing: they're aimed at what went wrong. That's no accident – it's how most organizations reflexively respond to a problem. But think about it for a moment and it's an odd choice. In most operations, the probability of something going right, against the odds and under the same constraints as the rest of the organization, is considerably rarer – and therefore at least as informative – as one more analysis of what went wrong. That imbalance, and what to do about it, is the subject of the next article in the series.