Debate, Decide, Unite: Systematic Decision-Making
Name the decider, list the factors, stack-rank them, then unite behind the call — with calories proportional to consequence.
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 6
- Confidence
- 93%
HubSpot's decision operating system, assembled from four parts: data informs but people decide; one named individual owns each decision; the effort spent must be proportional to the consequences; and once the decision is made, everyone unites behind it. Shah adds a personal technique — imagine writing a Python function or Excel formula for the decision, identify the input factors, and stack-rank them, which he claims is 80% of the win even without ever assigning weights.
Origin
HubSpot's synthesis. 'Debate, Decide, Unite' is HubSpot's deliberately warmer twist on Amazon's 'disagree and commit'. The DRI (directly responsible individual) concept is credited to Apple and explicitly not claimed by Shah. The one-way/two-way door heuristic is Jeff Bezos's; Shah's extension is making the calorie spend continuously proportional rather than binary.
Core principles
- 01Data is awesome, but data doesn't make decisions — people do.
- 02Deciding who decides is almost the number one decision to make.
- 03The decider is not necessarily the executive who runs that team.
- 04Calories spent on a decision should be proportional to the consequences of that decision — a continuous version of Bezos's binary one-way/two-way doors.
- 05Merely identifying the factors that should influence a decision is an 80% win, even before weighting them.
- 06Alignment after the decision matters more than winning the debate — at any scale, some people will have argued for the losing option.
How to run it
- 1
Size the decision first
Before spending any effort, ask how consequential and how expensive to reverse this decision is. Big, hard-to-reverse decisions deserve heavy calories; cheap reversible ones deserve almost none.
Pro tip Bezos's one-way vs two-way doors is the elegant binary version; Shah recommends treating it as a continuous scale.
- 2
Designate the decider
Name one position that owns the decision — a DRI. Do not default to the most senior executive in the room, and never designate someone you do not trust to decide.
Watch out Unowned decisions are the default failure mode. 'The team will figure it out' means nobody decides.
- 3
Write the imaginary Python function
Ask: if I could write a function to answer this, what would the variables be? If I were making a spreadsheet, what would the columns be? Collectively enumerate the factors that should influence the decision.
Pro tip Do this as a team — getting the factors out of individual heads and onto a shared list is where most of the value is.
- 4
Stack-rank the factors instead of weighting them
You cannot honestly assign exact weights, so do not fake precision. Just rank which factors matter more than which. The human decider now has dramatically better information.
Watch out Never let the spreadsheet make the call. It informs the decider; it does not replace them.
- 5
Debate openly, then decide
Get all the options on the table with genuine open debate — then have the DRI actually make the call.
- 6
Unite behind the decision
Once the call is made, everyone — including those who argued and lost — aligns behind it. HubSpot deliberately says 'unite' rather than 'commit' because 'disagree and commit' reads as harsher than the culture they want.
Pro tip The wording matters. 'Come together around it' lands differently from 'disagree, then effing commit.'
Watch out This is genuinely hard to do and is the step most organisations skip. It is also where most of the value sits.
In the wild
Shah admits HubSpot, like most organisations, historically made decisions poorly, and the biggest single mistake was failing to get alignment after a decision was made. They adopted Amazon's disagree-and-commit principle but rewrote it, because 'it's fine for you to disagree, but then effing commit' was harsher than HubSpot's culture. Their version: debate, decide, unite.
→ The substance of the principle was preserved while the framing fit their culture, and it became part of the HubSpot cultural vocabulary.
Shah applies the same proportionality logic to his own life. His default answer to any shiny new object is no, and he must force himself to find reasons to say yes. When he does say yes, he forces himself to name what comes out of his schedule to make room — because every yes is by definition a no to something else, and he has been unable to bend the laws of space and time.
→ He wrote a 'Sorry, but no' post on his blog that he describes as life-changing for him, and the mirror of the product's one-in-one-out rule applied to time.
Common mistakes
Letting the chart make the decision
Data-driven companies fall into 'we have a chart that says this, so we should do that'. Data informs; it cannot weigh the factors it does not contain.
Failing to unite after the decision
At any scale, some people argued for the other option and lost. If they do not align, the decision is executed half-heartedly by people quietly hoping to be proved right.
Spending calories out of proportion to consequence
Deliberating a cheap, reversible decision at the same depth as an irreversible one wastes the finite calories the big decisions actually needed.
Is it for you?
Best for
Exec teams and product leaders at scaling companies who suspect decisions are stalling, drifting, or being silently relitigated after the fact.
Not ideal for
Two-person startups where the founders decide everything in a conversation, and crisis situations demanding immediate command.
From the transcript
“we also um have one position that we designate to make the decision and that is not necessarily the executive overranging that team”
“ours is uh debate decide unite that's our kind of phrase that we use in hbot culture”
“with a a decision or thing that I'm trying to do is like okay if I could write python code a python function to come…”
“if I can just identify the factors that I think should impact this decision that's an 80% win right there”
“he calls them like oneway doors versus two-way doors which is a kind of a binary heuristic”
From the episode
Zigging vs. zagging: How HubSpot built a $30B company
Dharmesh Shah (co-founder/CTO)