In 2017, a Fortune 500 retailer spent eleven weeks deciding whether to move their e-commerce platform to a new hosting provider. Eleven weeks of meetings, stakeholder reviews, risk assessments, and alignment sessions. The actual migration took four weeks. The delay cost them an estimated $2.3M in lost uptime improvements and deferred feature launches. Nobody was fired for the slow decision. Nobody was even criticised. Because the process had produced consensus, and consensus feels like due diligence.
It wasn’t.
What Slow Decisions Actually Cost
A decision that takes two weeks when it could take two days doesn’t just cost you the twelve extra days. The real expense is everything that happens – or doesn’t happen – while the organisation waits.
Thirty people can’t start their work until someone above them decides. Meetings proliferate to “align stakeholders” and “gather input.” Follow-up emails go back and forth. Slide decks get created to present options that were already obvious. Calendar invites multiply. The people closest to the problem – the ones who could actually execute – sit idle or work on lower-priority tasks because the green light hasn’t come.
Amazon estimated internally that slow decisions cost them more than wrong decisions. That insight led to “disagree and commit” becoming one of their 14 leadership principles. Jeff Bezos wrote about this directly in his 2016 shareholder letter: most decisions are reversible, two-way doors. Treating them like irreversible, one-way doors slows everything down for no real risk reduction.
He was right. And most leadership teams haven’t absorbed the lesson.
Why Consensus Feels Safe
The appeal of consensus is obvious. If everyone agrees, nobody can be blamed when it goes wrong. The decision becomes collective property, and collective property means diffused accountability. “We all agreed this was the right move” is a powerful shield in a post-mortem.
But that safety is an illusion.
By the time you’ve watered down a strategy to get twelve people to sign off on it, you’ve produced something so inoffensive it’s also ineffective. The bold moves – entering a new market, killing an underperforming product line, restructuring a department – require someone to make a call that not everyone will agree with. When you optimise for agreement, you optimise for mediocrity.
I watched a SaaS company spend three months building consensus around a pricing change. By the time every stakeholder had weighed in, the original 30% price increase had been negotiated down to 8%, the rollout plan had been delayed by a quarter, and the expected revenue impact had shrunk from $4M to $600K. Everyone was comfortable. The company left $3.4M on the table because comfort was more important than conviction.
The RACI Problem
RACI matrices were supposed to fix this. Responsible, Accountable, Consulted, Informed – assign roles clearly, and decisions will flow smoothly. In theory.
In practice, RACI charts clarify who does what but they don’t change culture. A RACI matrix will tell you that the VP of Product is the decision-maker for feature prioritisation. It won’t stop the VP of Sales, the VP of Marketing, and the CTO from all wanting to weigh in. And in most organisations, “Consulted” is interpreted as “has veto power” and “Informed” is interpreted as “gets to reopen the discussion.”
The problem isn’t structural. It’s behavioural. People weigh in on decisions because they can, because the culture rewards input over speed, and because staying quiet feels like abdicating responsibility. No framework diagram on a wiki page changes that.
What does change it is a leader who says: “I’ve heard everyone’s input. Here’s what we’re doing. If you disagree, tell me now, and then commit to executing this as if it were your own decision.” That takes nerve. Most leaders would rather schedule another meeting.
What “Disagree and Commit” Actually Looks Like
The phrase gets tossed around a lot, usually by people who’ve read one blog post about Amazon’s culture. Implementing it is harder than quoting it.
Here’s what it looks like when it works. The decision-maker – and there is one decision-maker, not a committee – gathers input from relevant people. This isn’t a week-long process. It’s a 30-minute conversation, a Slack thread, or a one-page memo with a 24-hour comment window. The decision-maker reads the input, makes a call, and communicates it clearly: “We’re going with Option B. I know some of you preferred Option A. I’ve considered your arguments and I’ve decided Option B better serves our goals for these reasons. I need everyone to execute on this fully.”
The commitment part is the hard part. Most organisations do “disagree and then passively undermine” instead. The person who wanted Option A slow-walks their deliverables. They flag every minor problem as evidence that the wrong choice was made. They bring up their preferred option in tangentially related meetings. This is corrosive, and it happens everywhere.
Stopping it requires two things: the decision-maker has to genuinely listen before deciding (people commit to decisions when they feel heard, even if they disagree), and the organisation has to treat undermining a committed decision as a serious performance issue. Not a difference of opinion. A performance issue.
The 70% Rule
Bezos popularised another idea that pairs well with disagree-and-commit: if you have 70% of the information you wish you had, make the decision. Waiting for 90% means you’re almost always too slow.
This makes intuitive sense in technology organisations where the cost of reversing a decision is often low. Picked the wrong A/B test variant? Run another test. Chose the wrong microservice architecture? Refactor in the next quarter. Hired for the wrong role? That’s a harder reversal, which is why hiring decisions deserve more deliberation.
The 70% rule doesn’t mean being reckless. It means being honest about what additional information will actually change your decision versus what information you’re gathering to feel more comfortable. In my experience, about half of all “information gathering” in corporate settings is comfort-seeking disguised as diligence. The team already knows what they want to do. They’re looking for data that confirms it.
Just make the call.
For technology leaders, Farnam Street’s decision-making frameworks offer practical mental models for distinguishing between decisions that need deliberation and decisions that need speed. Worth bookmarking.
When Consensus Is the Right Call
Not every decision should be fast. Certain categories genuinely benefit from broader input and slower deliberation:
- Irreversible decisions – M&A, major restructures, market exits, significant capital expenditure. If you can’t undo it cheaply, invest the time to get it right.
- People decisions with cultural impact – layoffs, executive appointments, changes to compensation philosophy. These affect trust, and trust is hard to rebuild.
- Regulatory and compliance matters – getting these wrong has consequences that dwarf the cost of delay.
The mistake isn’t deliberating on these. The mistake is treating every decision like it belongs in this category. Most decisions in a given week are reversible, low-to-medium stakes, and would benefit from a two-day turnaround rather than a two-week one. When a CEO applies the same decision process to “should we redesign the homepage” as to “should we acquire this company,” something has gone wrong.
C-Suite Dynamics and Decision Speed
The CEO who tries to build consensus among their entire leadership team on every strategic decision will move at the speed of their most cautious executive. There’s always one person on the leadership team who needs more data, wants another scenario modelled, or isn’t comfortable with the risk profile. If that person has implicit veto power – and in consensus cultures, everyone does – they become the bottleneck for the entire organisation.
Strong CEOs know which decisions belong to which leaders and give them room to make the call. The best CEO programs teach this: your job isn’t to make every decision. It’s to make the decisions only you can make, delegate the rest clearly, and intervene only when the decision-maker is stuck or the stakes are existential.
This extends to board interactions as well. A technology leader presenting to the board who asks for consensus on technical strategy is inviting twelve non-technical opinions into a domain where they’ll slow things down without improving the outcome. Present a recommendation. Explain the reasoning. Answer questions. Move forward.
For COOs specifically, decision speed is operational oxygen. Every day a decision sits unresolved, operational efficiency degrades. The COO’s calendar should not be filled with alignment meetings – it should be filled with decision-making moments.
A Pattern Worth Noticing
The best executives I’ve worked with share a common trait that cuts across industries, company sizes, and functional backgrounds. They gather input quickly, decide quickly, and course-correct quickly. They’d rather be 80% right today than 95% right next month.
They also share a comfort with being wrong. Not recklessly wrong, but wrong in the way that comes from acting on imperfect information and adjusting as you learn. They treat decisions as hypotheses, not pronouncements. “We’re going to try this. If the data says we’re wrong in 90 days, we’ll adjust.”
Compare that to the executive who needs to be right before they act. Who commissions the extra analysis. Who schedules the third alignment meeting. Who asks for “just one more week” to think about it. That executive feels responsible. They look responsible. But they’re costing the organisation far more than the occasional wrong decision ever would.
Changing the Default
If your organisation defaults to consensus, changing that default is a multi-year culture project. But you can start with small moves:
- Name the decision-maker explicitly – for every significant decision, one person has the final call. Not a committee. Write their name down.
- Set decision deadlines – “We will decide by Friday” is more powerful than most process improvements. An imperfect decision on Friday beats a perfect decision three weeks from now.
- Distinguish decision types – create a simple taxonomy. Type 1: irreversible, high-stakes, take your time. Type 2: reversible, moderate stakes, decide in 48 hours. Most decisions are Type 2.
- Track decision velocity – measure the average time from “decision needed” to “decision made” for key decisions. Make it visible. What gets measured gets faster.
- Reward speed, not just outcomes – publicly praise leaders who make fast calls, even when they have to course-correct later. Especially when they course-correct, because that proves the system works.
Leaders with strong executive presence tend to be the ones who pull this off. Not because they’re louder or more forceful, but because they’ve earned enough trust that people follow their lead even through uncertainty.
Consensus has its place. But when it becomes the default mode for every decision, it stops being collaboration and starts being paralysis with better optics. The organisations that outperform their competitors aren’t smarter. They’re faster. And speed starts with someone willing to say: “We’ve talked enough. Here’s what we’re doing.”
Ben is a full-time data leadership professional and a part-time blogger.
When he’s not writing articles for Data Driven Daily, Ben is a Head of Data Strategy at a large financial institution.
He has over 14 years’ experience in Banking and Financial Services, during which he has led large data engineering and business intelligence teams, managed cloud migration programs, and spearheaded regulatory change initiatives.