May 20, 2026 May 20, 2026 2026-05-20

In April 2011 I wrote seven things to do when the shit hits the fan, directly from inside a disaster. An iPhone app had gone catastrophically wrong — the kind of week where every solution you try spawns two new problems, and you start wondering whether you should just burn the whole project down and start over. I published the list as a way of processing what was happening. It was practical, written under pressure, and based on exactly one type of crisis.

I revisit this now, because fifteen years is a long enough timespan to have significantly more data on this subject.

The Unpredictability of the “Problem”

The original post observed that “unpredictable” crises are often less unpredictable than they feel at the time. Looking back now at everything between 2011 and 2026 — a significant business failure, some months of 2020 that nobody wants to relive, and the quieter ones that don’t make good stories but take just as long to get through — the observation still holds. Most of these crisis had signals. The signals were either too faint, or too inconvenient, or I’d decided that reading them clearly wasn’t useful right in that specific moment.

This is worth saying at the start because it changes how you relate to the seven steps below. If you understand that the “unforeseeable” disaster often had a prologue you didn’t acknowledge, you need to process beyond “what do I do right now”. It also involves a longer, quieter question: what am I not reading right now, in whatever seems stable? But let’s keep this for a future post. When the shit actually hits the fan you have more urgent things to do.

What the Seven Steps Look Like in 2026

Assess, don’t stress. Still the right starting point. With a tiny touch: the assessment has to be also honest, not just productive. In 2011 I was channeling stress into problem-solving mode, which was useful, but the assessment was sometimes selective — I’d look carefully at the parts I could act on, and skim over the parts I’d have to sit with. Real assessment means looking at the entire problem spectrum, including the parts where the answer is “this is going to hurt for a while and there’s nothing to do about that right now.” That’s harder than converting stress into action, and more necessary.

Decide so it won’t collide. The original was about choosing priorities and acting on them, which is right. What fifteen years added is: sometimes the decision is to not decide yet. In a technical crisis, waiting is usually expensive. In a personal or health crisis, premature deciding can close doors that needed to stay open. The skill isn’t only in making decisions — it’s in recognizing which decisions are genuinely urgent and which ones are tempting you to act just to feel less helpless. Those two feel identical from the inside, but they are really different situations.

Communicate. This held across every type of crisis without exception. The 2020 version was: call the people you love more often than feels necessary. The professional version: tell stakeholders early, even when what you’re telling them is uncertain. In my experience, silence reads as worse news than the truth almost every time. People can handle hard information better than they can handle their own silence filling stories.

Rebuild carefully. The word “carefully” does different work depending on what’s broken. In a code disaster, carefully means methodical — one change at a time, check each one. In a relationship crisis, carefully means slowly — some repairs can’t be rushed without just applying a superficial patch. In a health scare, it means trusting a process that doesn’t feel like progress. The instruction is the same across all three. But the application has to match the domain, and the temptation is always to apply the logic of the fastest category to whichever one you’re actually in.

Leverage. In professional crises, leverage is relatively clean — use your tools, use your network, use your platform. In personal ones, the same impulse can go sideways. Not every crisis benefits from being turned into content, or a learning opportunity, or a networking moment. Some things just need to be lived through without being optimized. Over time I’ve learned to ask which kind of situation I’m in before reaching for that tool, because applying leverage to something that needs space is its own kind of mistake.

Keep your fingers crossed. What this became, over time, is an acceptance practice. There’s a line between what you control and what you don’t, and the step that 2011-me called “keep your fingers crossed” is really about finding and honoring that line. Not giving up. Not pretending it doesn’t matter. Just stopping at the place where your effort ends and something else begins. The Stoics had a name for this. So does anyone who has been through a long medical crisis, or a year of trying to build something that kept not working, or a global situation nobody could fix by working harder.

Learn. Still valid, but timing matters more than the 2011 version acknowledged. The urge to extract the lesson while you’re still inside the event is understandable — it’s part of the same impulse as “assess, don’t stress.” But it often produces premature conclusions, because you don’t have the full picture yet. The real learning tends to land six months to a year later, once you can see the whole shape of what happened. The note-taking during the crisis is still worth doing. Not to learn from immediately, but to give yourself accurate material to work with once the dust settles and you can finally see clearly.

The Thing That Wasn’t on the List

In a single-week crisis — the kind the 2011 post was written from — your body is basically fine. You sleep less, you skip the morning exercise, you eat whatever’s near the keyboard. Not a problem for a week. In a crisis that goes on for months, which the more serious ones tend to do, the base becomes everything. Sleep, movement, regular meals. Not because you’ll perform better (you obviously will, but that’s not the point), but because the base is what makes the rest of the list possible. You can’t assess clearly when you’ve been running on four hours of sleep for three weeks. You can’t communicate well when you’re deep in a cortisol spiral. You can’t even tell whether you’re making decisions or just reacting.

The base isn’t self-care in the lifestyle sense. It’s the actual infrastructure the other seven steps run on. Protect it earlier than feels necessary, because by the time you notice you’ve lost it, chances are that you’ve already been operating without it for a while.

What Actually Held

The framework is more useful as guidance than a checklist. In the middle of a real disaster, you don’t run through seven items in order. You reach for whichever one is most relevant to what’s happening right now — and some of them, like “keep your fingers crossed” and “learn,” do their work quietly in the background whether you invoke them consciously or not.

What held across fifteen years and several different kinds of falling apart: the combination of honest assessment, early communication, and accepting the boundary of your control. Those three, done consistently, cover more ground than the rest of the list combined. Everything else is support structure built around them.

The iPhone app crisis of 2011 eventually resolved. I don’t remember now exactly how. What I remember is that the week felt permanent while it was happening, and then it wasn’t. Most of them are like that — they feel like the new normal right up until they stop.