Barrels and Ammunition
Adding headcount doesn't add output — only adding people who drive initiatives end-to-end does.
- Difficulty
- Moderate
- Time to result
- ~months to results
- Steps
- 5
- Confidence
- 95%
Companies contain a tiny, fixed number of people who can independently take an initiative from inception to success (barrels). Everyone else is ammunition — valuable support, but stacking more of them behind existing barrels just adds coordination tax, not output. To do more in parallel you must add barrels, not headcount. The barrel-to-ammunition ratio dictates how many important initiatives a company can pursue at once.
Origin
Keith Rabois's own framework, developed after years of dinners with frustrated CEOs whose burn rose while output stalled; he references discussing it in his 2014 YC lecture and cites a conversation with Jack Altman about the barrel count at Lattice.
Core principles
- 01Output scales with barrels, not headcount
- 02Most companies have between 2 and ~15 barrels; PayPal at acquisition (~254 people) had only 12–17
- 03Hiring more ammunition without more barrels increases coordination/collaboration tax and reduces output
- 04Ammunition is real and necessary — match its type and amount to the specific problem
- 05A barrel is defined by outcome ownership: 'fire and forget' delivery, or proactive escalation with root-cause diagnosis and time to intervene
How to run it
- 1
Count your actual barrels
Honestly tally how many people in the org can take an initiative from inception all the way to success with no hand-holding. The number is almost always far smaller than leadership assumes — often just two.
Pro tip Benchmark against reality: even PayPal, one of the densest talent networks ever, had only 12–17 barrels among ~254 people.
- 2
Diagnose the drag before hiring
When output stalls despite rising headcount, resist the reflex to hire more. Recognise that stacking people behind the same initiatives is what creates the drag coefficient.
Watch out Adding ammunition to a barrel-constrained org makes you do LESS, not more, because of coordination tax.
- 3
Test candidates for barrel ability
Assess whether a person can take an idea and make it happen: point at a hill and see if they motivate people, accumulate resources, and measure what's needed to get the company over it. Anything less is ammunition.
Pro tip A true barrel who hits a wall comes back proactively with the issues, what they already tried, root-cause diagnosis, and enough lead time for you to help.
Watch out Don't mistake competence or seniority for barrel ability — the test is independent end-to-end ownership of an outcome.
- 4
Expand scope on proven barrels
Start a new barrel on a narrow, lower-difficulty outcome, then progressively entrust them with more complexity and ambiguity as they deliver. Give barrels more surface area rather than hiring around them.
Pro tip Watch for barrels in unexpected places — Rabois found one in a two-day intern who solved a problem his entire office team had failed at.
- 5
Set the ammunition ratio per project
For each initiative, decide the right amount and type of ammunition (designer, engineers, PM, data analyst) the barrel needs. Some projects need almost none; others need a full team.
Pro tip Being deliberate about ratios makes team construction intentional instead of reactive headcount growth.
In the wild
Rabois wanted cold, healthy smoothies delivered to engineers at 9pm. His substantial office/EA team repeatedly failed — smoothies were late or warm, so no one ate them. A two-day intern, Taylor Francis, said 'I'll solve it.' The next day cold, great-tasting smoothies arrived at 9pm on the engineers' standing desk.
→ Rabois concluded 'I found a barrel' and later handed Francis almost everything to do.
Rabois asked Jack Altman on a podcast how many barrels a company as large as Lattice had. The answer was two — which Rabois notes is a common answer even for a very good company.
→ Illustrates that even successful mid-to-large companies operate on a handful of barrels, capping how many initiatives run in parallel.
Common mistakes
Solving a barrel shortage by hiring ammunition
Raising a big round and hiring aggressively without expanding barrel count just stacks people behind existing initiatives, raising burn and coordination tax while output stays flat or falls.
Treating 'agency' as self-explanatory
Rabois warns that words like agency and strategy 'go in one ear and out the other' — leaders nod without operationalising the barrel behaviour, so they never actually screen or build for it.
Is it for you?
Best for
Founders and CEOs post-Series A/B who have hired aggressively but feel output per unit time has stalled or dropped
Not ideal for
Very early pre-traction teams where the founders are the only barrels and headcount isn't yet the bottleneck
From the transcript
“the number of people that can independently drive an initiative from beginning from inception to success is very limited”
“the ratio of barrels to ammunition is what dictates the number of important initiatives that can be pursued simultaneously”
“and they're going to get your company across that hill. That's a barrel.”
“Oh my”
“the answer was two.”
From the episode
Hard truths about building in the AI era
Keith Rabois (Khosla Ventures)