Hire

It breaks by hiring to feel like a real company. Michael Seibel's read, from years of YC office hours: most early startups hire too early, because headcount feels like progress and an org chart feels like adulthood. Every pre-product-market-fit hire raises burn and lowers speed, and the trap is that it feels like the opposite in the moment: relief now, invoice forever. The identity currency buys the headcount; the money currency pays for it monthly.

The senior version is the scaling executive hired early, the classic self-inflicted wound in Horowitz's documentation. The post-raise "get me a VP" reflex buys the dream org years before the company can feed it. She arrives, asks for a team and a budget, because that's what VPs are for, and bills $220k fully loaded while the actual work sits untouched, because the actual work was execution and you hired a strategist. Reid Hoffman's correction holds the whole call: hire for the next 12 to 18 months, and good enough for this stage beats perfect for a stage you may never reach.

1. Write the pain sentence. One sentence, present tense, a number inside. What breaks, how often, what it costs. That sentence is the real spec; everything else in the req is decoration.

2. Sort recurring from one-time. One-time or spiky pain (a migration, a launch, a rebrand) gets a contractor or an agency, never a salary. Recurring pain moves to the next rung.

3. Climb the deletion ladder. Before staffing the pain, try to delete it, in order:

• Automation first. Can a tool, a script, or a process change kill this outright? A robot doesn't onboard, churn, or need managing. The best hire is often a script.
• Systematize second. Write the runbook. If you can't describe the job, you can't delegate the job, and hiring into undocumented chaos employs someone to be confused full time.
• Contractor third. Rent the skill, test whether the volume is real.
• Hire last. Recurring pain that survived the ladder, with proven volume. Now it's a hiring question, and now the req writes itself, because the ladder wrote the spec.

4. Check the window. Scope the role to the next 12 to 18 months of actual work. If the pitch for the role needs the word "eventually," you're hiring for the dream org. The dream org doesn't exist yet; its executives bill now.

5. Barrels before resumes. Keith Rabois's frame from PayPal and Square: output is gated by barrels, the rare people who take an idea from conception all the way to shipped. Everyone else is ammunition, useful exactly in proportion to the barrels they load. Two consequences. First, barrels are found empirically, by expanding responsibility until someone hits their ceiling, which means your next barrel may already be on the team, underscoped: check before you shop. Second, when you do shop, hire for extreme strength, never for absence of weakness. The candidate with no flaws and no spike is ammunition with good interview manners.

6. Raise the average, then price it honestly. The test: does this person raise the team's average on the dimension the next 18 months actually turn on? A maybe is a no. Then the math: fully loaded cost runs far past salary (taxes, tools, and the management hours you stop spending elsewhere), so budget 6 to 12 months of runway per early hire. The ruin check is two questions: does this hire flip you from default alive to default dead, and can you afford to be wrong in 90 days? Set the 90-day output bar in writing before the offer goes out. If being wrong means a bridge round, the answer was a contractor.

• Hiring to feel like a real company. Named above, and it's the most common way this breaks.
• The post-raise shopping spree. The bank balance changed. The stage didn't.
• Buying a VP to escape work only founders can do. Pre-product-market-fit selling is a founder job; nobody else can hear the market flinch at your pitch and change the product by Thursday.
• Staffing undocumented chaos. Systematize first, or the second hire will be managing the confusion of the first.
• The flawless candidate. No weaknesses usually means no spike.
• Title inflation to close. Titles cost nothing today and everything in 18 months, when the real VP needs to arrive above the courtesy one.

B2B SaaS at $70k MRR, fresh $2M seed, founder wants a VP of Marketing. The pain sentence takes three tries and finally lands as: "We published two posts in six months and pipeline is 90% my founder-led sales." That's an execution gap, not a strategy gap; strategy is currently the founder, and stays the founder for 18 more months. The ladder runs: reporting automates into a dashboard, design goes to a contractor, and the surviving recurring pain is weekly demand-gen execution. The window check kills the VP outright. The barrel check inside the six-person team finds no marketing spike. So the req becomes a full-stack marketer, extreme strength in content and lifecycle, $150k fully loaded, roughly 7% of the raise per year. Wrong-in-90-days is affordable and defined in writing: 12 campaigns shipped and measured pipeline contribution by day 90.

She clears the bar by day 60. Eighteen months later she owns demand gen end to end and still hasn't hit a ceiling, so her scope expands again: the barrel test running live, on payroll. The VP conversation returns at $250k MRR, with a playbook to hand over. Which is what VPs are actually for.