12 May 2026
The Practice of Deciding
Strong opinions, loosely held. Here's what that actually means.
Strong opinions, loosely held. Here’s what that actually means.
The architects I’ve worked with who make the best decisions aren’t the most confident. They’re the most practiced.
That distinction is I think important. Confidence is a feeling. Practice is a capability. And in the work of building, whether that’s an enterprise architecture, a product, or a team, the ability to decide well under uncertainty is the thing that separates people who ship from people who deliberate.
I’ve been on both sides of that.
The paralysis isn’t a knowledge problem
When a decision stalls, the instinct is to reach for more information. Another stakeholder to talk to. Another analysis to run. Another week to let things settle or see what changes.
Most of the time, that instinct is wrong.
The information you’re waiting for probably won’t arrive, likely won’t be conclusive, or will arrive too late to matter. What’s actually happening is a confidence-in-process problem. You don’t trust your method for making the call, so you keep delaying the moment when you have to use it.
More data is sometimes the answer. But if you find yourself reaching for it every time, the problem isn’t the data.
What “strong opinions, loosely held” actually means
The phrase gets misquoted constantly. People hear it as permission to hold weak opinions, to stay deliberately non-committal so you can’t be proven wrong. That’s not it. That’s just hedging.
The original intent is harder than it sounds. You form a genuine view. You commit to it fully, clearly enough that people know where you stand, strongly enough that it can actually guide a decision. And then you hold it in a way that’s genuinely open to revision when better information arrives. Not performatively open. Actually open.
Most people do one or the other. They either commit so hard that new information can’t get through, or they stay so loose that they never actually lead. The practice is learning to do both at once: full commitment, real openness. It’s not a personality type. It’s a skill, and it’s developed through repetition.
The architectural frame
One of the more useful things enterprise architecture teaches you, and it took me longer than it should have to internalise this, is that you’re not optimising for the perfect decision. You’re optimising for the right decision at this moment, with the information available, in a way that keeps future options open where possible.
The frame I come back to: is this a one-way door or a two-way door?
Two-way doors should be opened fast. You can walk back through if you’re wrong. The cost of a bad call is low and you course-correct. The cost of delay is real: momentum lost, teams blocked, windows closed. Treat two-way door decisions as experiments. Make the call, watch what happens, adjust.
One-way doors deserve more time. Not infinite time, but genuine deliberation, the right people in the room, explicit consideration of what you’re foreclosing. Architecture choices, team structures, client commitments, technology bets. These compound. Get them wrong and you carry the cost for years.
The failure mode I see most often isn’t treating one-way doors carelessly. It’s treating two-way doors as if they were one-way, applying the same agonising process to a low-stakes call that you’d apply to a structural commitment. That’s where the paralysis lives. Knowing what you need to care about is key.
Deciding in the open
There’s a version of decision-making that happens inside someone’s head and gets announced as a conclusion. And there’s a version that happens transparently, where the reasoning is visible, the assumptions are named, and the conditions for reversal are stated.
The second version is harder in the moment. It exposes your thinking before it’s been validated. But that exposure is actually the point.
Share a design early, before you’ve spent weeks on it, before you’ve defended it in three meetings, before you’re emotionally invested in it being right and something useful happens. The people you share it with aren’t threatening your work. They’re improving it. And because you haven’t committed everything to this version, you can actually hear them. The opinion is loosely held because it hasn’t calcified yet.
Wait too long and the dynamic inverts. The thinking hardens. The time invested becomes something to protect. Feedback that would have been genuinely useful at week one feels like an attack at week six. You’re no longer deciding, you’re defending.
This is as true for a product decision as it is for an architecture diagram. The earlier the conversation, the more honest it is. The more honest it is, the better the outcome.
Deciding in the open means briefing people on the direction before it’s final, not announcing it after. It means your reasoning is visible, your assumptions are named, and the conditions for reversal are stated. It builds trust with the people who have to implement your call. And it means that when someone challenges it, you’re not defending your ego, you’re referring to a reasoned position you can actually update.
The practice
Deciding well isn’t a talent. It’s forming a habit.
Know whether the decision is a one-way or two-way door, and calibrate your process accordingly. Form a genuine view and commit to it clearly. Share it early, before you’re invested in being right. State your assumptions. Name what would change your mind. Make the call. Watch what happens. Update without drama.
Repeat that enough times and something changes. Not that you stop being uncertain, uncertainty is permanent and anyone who tells you otherwise is selling something. What changes is your relationship to it. You get comfortable making calls in conditions that used to feel like they needed more information, more time, more consensus.
That comfort is the practice. It’s what separates the people who lead through ambiguity from the people who manage it from a safe distance.
The best decision-makers I’ve worked with aren’t waiting for certainty. They’ve stopped expecting it.
John Murphy is co-founder and principal at ailiniu, an enterprise architecture and AI consultancy based in Ireland. ailiniu helps regulated financial services and insurance firms navigate architectural complexity and build the organisational capability to move at pace.