"How teams actually decide"

Watch a team try to decide something that matters and you learn more about its health than any survey would tell you. Some teams move from a hard question to a committed choice cleanly — the real disagreement surfaces, the right people weigh in, a decision is made, and the team actually commits to it. Others cannot get there: they circle, defer, relitigate, decide by whoever is loudest, or reach a conclusion no one truly commits to that quietly comes undone weeks later. The difference is a distinct capability, and it is one of the quietest and most consequential determinants of how fast and how well a team executes.

Decision-making is one of the eight dimensions of a healthy team, and it is often the one whose weakness is felt most and named least. Teams experience the symptoms — the stalled choices, the endless revisiting, the decisions that never quite hold — without recognising them as a decision-making problem, because they look like separate frustrations rather than one underlying capability. But how a team decides is a specific, buildable thing, distinct from how any individual on it decides, and a team of strong individual decision-makers can be a poor collective one. Understanding it as its own dimension is the first step to fixing the thing that quietly slows so much execution.

Deciding is a team capability, not an individual one

The first thing to see is that team decision-making is not the same as the decision-making of the people on the team. A group of individually excellent decision-makers can decide badly together, because collective deciding involves things individual deciding does not: surfacing disagreement across people, combining different knowledge, managing the social dynamics of who defers to whom, and producing shared commitment rather than a single person's conclusion. Those are group capabilities, and they can be weak even when every individual is strong, which is why a team's decision quality has to be understood on its own terms.

This is why "we have smart people, so we must decide well" is a false inference. The smartness of the individuals is necessary and not sufficient; what determines the team's decision quality is how those individuals decide together, which depends on the group's dynamics rather than the members' intelligence. A team can be full of excellent judgment and still decide poorly, because the excellent judgment never makes it into the decision — suppressed by hierarchy, lost in poor process, or overridden by the loudest voice. The decision-making dimension is about whether the team's collective intelligence actually reaches its decisions, which is a different and often much weaker thing than the intelligence itself.

The three ways teams decide badly

Team decision failures tend to take three shapes. The first is paralysis: the team cannot reach a decision at all, circling the question, gathering more information, deferring, and relitigating without ever committing. The choice stays open long past the point where it should have been made, and the cost of the delay often exceeds the cost of any of the options. This is frequently a safety problem — the team cannot surface and resolve the real disagreement, so it avoids deciding rather than confront the conflict — or an ownership problem, where no one is clearly responsible for making the call.

The second is deciding by the wrong input: the decision gets made by whoever is loudest, most senior, or most insistent, rather than by whoever has the best knowledge of the question. The team's actual expertise is overridden by its hierarchy or its volume, so the decision reflects who spoke most forcefully rather than what was true. The third is deciding without commitment: the team reaches a conclusion, but the conclusion is compliance rather than genuine buy-in, so it quietly comes undone — people execute their own real views, or the question gets reopened the moment the decision meets friction. Each of these is a different failure, and each has a different fix, but all three trace back to the same roots: a lack of safety to surface real disagreement, and a lack of clarity about how the decision gets made and by whom.

Paralysis is usually avoided conflict

The team that cannot decide is often not short of information or analysis; it is avoiding the disagreement that deciding would force into the open. Somewhere under the endless circling is a genuine conflict — two people who want different things, a real tension between options that reflects a real difference in the team — and deciding would require surfacing and resolving that conflict, which does not feel safe. So the team gathers more information it does not need, defers to a later meeting, and relitigates settled points, all as ways of avoiding the conflict that a decision would expose. The paralysis is not indecision; it is conflict-avoidance wearing indecision's clothes.

This is why more information and more analysis so rarely cure a paralysed team, and can make it worse by providing more material to circle. The team does not lack data; it lacks the safety to have the disagreement that deciding requires, which is a trust problem, not an information one. A team that feels safe to surface and work through genuine conflict can decide, because the disagreement gets aired and resolved rather than avoided; a team that does not will keep circling, because the alternative to circling is a confrontation it is not safe enough to have. Fix the safety and the paralysed team can suddenly decide; leave it broken and no amount of additional analysis will produce a choice, because analysis was never what was missing.

The loudest voice is not the best input

The team that decides by its hierarchy or its volume is wasting exactly the collective intelligence a team is for. When the decision goes to whoever is most senior or most insistent, the team's actual knowledge of the question — which may sit with a quieter or more junior person — never reaches the decision, so the team decides with less than it knows. This is one of the most common and most costly decision failures, because it looks like decisiveness: someone strong made the call, the team moved. But the call was made on the wrong input, and the team's real expertise was silently overridden.

The fix is the same safety that cures paralysis, aimed slightly differently: making it safe for the person with the best knowledge to have their input actually count, regardless of their seniority or volume. This requires both the safety to speak and a genuine willingness at the top to be overridden by better information from below, which is a leadership behaviour as much as a team one. A leader who reliably defers to the best input, even when it contradicts them, teaches the team that knowledge beats hierarchy in decisions; one who reliably decides by their own preference teaches the opposite, and the team stops surfacing the knowledge that would override them. The best-deciding teams are the ones where the right input wins, which requires deliberately overriding the natural tendency for the loudest or most senior voice to win by default.

The decision that will not stay decided

The third failure — deciding without commitment — is the one that wastes the most effort downstream, because the decision looks made and is not. The team reaches a conclusion, but the conclusion is compliance rather than genuine buy-in, so it does not hold: people quietly execute their own views, the question gets reopened at the first sign of friction, and the team finds itself deciding the same thing repeatedly. This is the decision-making face of the alignment problem — agreement in the room that is not alignment in practice — and it comes from the same source: the decision was reached without surfacing the real disagreement, so the disagreement is still there, unresolved, waiting to undo the choice.

A decision that produces genuine commitment is one where the disagreement was actually worked through, not skipped — where the people who did not initially agree were heard, their objections addressed or genuinely considered, and they arrived at a choice they can commit to even if it was not their first preference. That is more expensive to produce than fast consensus and far more durable, because it does not come undone the moment it meets resistance. The teams whose decisions stick are not the ones that decide fastest; they are the ones that surface and resolve the real disagreement on the way to the decision, so that what emerges is commitment rather than compliance. Skipping that work produces a decision that will not stay decided, which is often worse than no decision at all, because it costs the team the same question multiple times.

Speed and quality are not opposites here

There is a false trade-off that shapes how teams think about deciding: the belief that they must choose between deciding fast and deciding well. In practice, the teams that decide well also tend to decide fast, and the teams that decide badly are often the slowest, because the things that make a decision good — surfacing the real disagreement, using the right input, producing genuine commitment — are also the things that keep it from being relitigated later. A team that decides slowly and badly, circling and then reaching a conclusion that comes undone, is slower overall than a team that decides cleanly and commits, even though any single meeting might look faster in the paralysed team.

So the goal is not to trade quality for speed or speed for quality; it is to build the capability that produces both, which is the capability to surface and resolve real disagreement efficiently. A team that can do that decides faster and better than one that cannot, because it neither circles in avoidance nor produces decisions that unravel. The apparent trade-off between speed and quality is usually a symptom of weak decision-making — a team that has to choose because it lacks the capability to do both — and the fix is not to pick a side but to build the capability. The best-deciding teams are fast and durable at once, which is what good decision-making, as a dimension, actually looks like.

Decision-making, trust, and alignment

Decision-making sits between trust and alignment in the web of dimensions, depending on the first and producing the second. It depends on trust because every one of its failures traces back to safety: the paralysed team cannot safely surface conflict, the loudest-voice team cannot safely override hierarchy, the no-commitment team cannot safely disagree on the way to the choice. Without the safety to have real disagreement, a team cannot decide well, because good deciding is fundamentally about working through disagreement rather than avoiding it — which is exactly what trust makes possible.

And it produces alignment, because a decision made well — with the real disagreement surfaced and resolved — is a decision people are genuinely aligned behind, while a decision made badly produces the compliance that fails to hold. So decision-making is the mechanism that converts trust into alignment: the safety to disagree, applied to a specific choice, produces a committed shared direction. This is why the dimensions cannot be read in isolation — a decision-making problem is usually a trust problem upstream and an alignment problem downstream, three views of the same underlying thing. Reading them together, as the eight dimensions are meant to be read, reveals that improving a team's safety to disagree improves its deciding and its alignment at once, because they are all the same work.

Who decides is itself a decision

A quiet but common cause of decision failure is ambiguity about who actually owns the call. When it is unclear whether a decision is the team's to make collectively, the leader's to make alone, or one person's to own within a boundary, the team gets stuck in a different way — not on the substance of the choice but on the question of who gets to make it. People wait for someone else to decide, or several people assume it is theirs and pull in different directions, or a decision gets made and then challenged on the grounds that the wrong person made it. The substance was decidable; the ownership was not clear, and the ambiguity did the stalling.

Getting clear on who decides — before the decision, not after — removes a whole category of decision failure. It does not mean every decision is one person's; some genuinely should be collective, some genuinely should be a single owner's, and the healthy thing is that the team knows which is which for a given choice. What causes trouble is the unspoken ambiguity, where no one has said whether this is a consult-and-decide, a collective call, or a delegated one, so the team improvises and often improvises inconsistently. A team that is explicit about decision ownership decides faster and cleaner, because it never has to resolve the meta-question of who decides in the middle of resolving the actual question. Clarity about the how of a decision is part of the capability to make it.

Disagree and commit

One of the most useful decision disciplines a team can build is the ability to commit genuinely to a decision one argued against. In a healthy team, a person can push hard for their view, lose the argument, and then commit fully to the chosen direction anyway — not the sullen compliance that undermines execution, but genuine commitment to making the team's choice work despite having preferred another. This is what lets a team both surface real disagreement and still move: people can fight for their position knowing that if they lose, the loss is honourable and the commitment expected, rather than a defeat to be quietly relitigated in execution.

This discipline depends on the disagreement having been genuine — people commit to a decision they lost when they feel they were actually heard, their objection genuinely weighed, and the choice made on the merits rather than by hierarchy or volume. A person overridden without being heard complies resentfully and reopens the question later; a person who lost a fair argument commits, because the process was legitimate even though the outcome was not what they wanted. So disagree-and-commit is not a rule you can impose; it is a capability that grows from good decision-making — from decisions made through genuine, heard disagreement rather than around it. A team that has it can move fast on decisions that had real conflict underneath, because the conflict was resolved in the deciding rather than deferred into the executing, which is exactly what separates a team that decides well from one that only appears to.

Reading how a team decides

Decision-making can be read, and it is read not by outcomes but by the quality of the process — because a team can reach a good outcome through a bad process by luck, and a bad outcome through a good process through bad luck, so the outcome alone is a noisy signal. What can be read reliably is whether the team surfaces real disagreement, decides with the right input, and produces genuine commitment, or whether it stalls, defers to the loudest voice, and reopens settled questions. That reading captures the capability, which is what predicts future decisions, rather than the outcome of any one decision, which is contaminated by luck.

And it can be read again over time, to see whether the capability improved — the before-and-after discipline in what Day 14, 30 and 60 tell you. Because decision-making rests on trust and produces alignment, reading it alongside those dimensions reveals the causes and consequences that reading it alone would miss: a team deciding badly usually has a safety problem under it and an alignment problem after it. This is why we read decision-making as part of the whole picture in why measurement changes the conversation, not as a standalone metric. Reading the outcomes tells you whether the team got lucky; reading the process tells you whether the team can decide.

A worked example

A team's execution is slow, and the leader concludes the team lacks urgency — people are not moving fast enough. So the leader applies pressure: tighter deadlines, more push, a demand for speed. Execution does not get faster, because a reading finds the real bottleneck is not urgency but decision-making. The team is paralysed on the choices that gate the execution, circling them endlessly because it cannot safely surface the real disagreement underneath — so the work waits on decisions that never get cleanly made, and no amount of urgency downstream speeds up a decision upstream that the team cannot bring itself to make.

Applying pressure made it worse, adding stress to a team already stuck, without touching the reason it was stuck. Reading first would have found the decision paralysis and its cause — a lack of safety to have the conflict that deciding required — and aimed the work at making it safe to surface and resolve that disagreement, so the gating decisions could actually get made. Same team, redirected from demanding speed to enabling decisions, and execution speeds up not because people were pushed harder but because the choices that were blocking them finally got made and stuck. The measurable result is faster execution flowing from cleaner deciding — which no amount of pressure would have produced, because the team was never slow from a lack of urgency; it was slow from an inability to decide.

Build the capability to decide

Decision-making is one of the eight dimensions, and it is the one whose failures are felt most and diagnosed least — experienced as stalled choices, endless revisiting, and decisions that will not hold, rather than recognised as a single underlying capability. A team decides well when it surfaces the real disagreement, uses the right input rather than the loudest voice, and produces genuine commitment rather than compliance — and every one of those rests on the safety to have real disagreement, which is why decision-making, like the other dimensions, keeps leading back to trust.

Build a team's decision-making by building the safety to surface conflict, the clarity about who decides, and the willingness to let the best input win over the loudest — and read the process, not just the outcomes, to know whether the capability is real. Most decision-making problems are trust problems upstream and alignment problems downstream, which is why they are best addressed as part of the whole picture rather than in isolation. Build the capability to decide, and a team stops circling, stops deferring to volume, and stops reopening settled questions — which removes one of the quietest and most persistent drags on how fast and how well it executes.