Paid test as a hiring filter ($100–300 working sample)

Replaces unpaid take-home tests with a 1–2 day paid working sample ($100–300 if no offer is ultimately made), inserted between the founder call and the trial month. The deliverable is a real scoped task from the role's actual workload.

Strong candidates increasingly refuse unpaid take-homes — they're either employed or in demand. A paid test (a) raises response rate from senior candidates, (b) preserves candidate dignity, (c) gives you a real working sample vs a polished generic portfolio, (d) signals the company's compensation discipline before the offer conversation.

1. Define a 1–2 day scoped task related to the role (real backlog item is best — not a contrived exercise).
2. Agree budget — $100–300 range for AI-native markets outside the top-tier metros; adapt for higher-cost markets.
3. Send the brief in writing with a deadline (typically 5–7 days to accommodate their day job).
4. If offer is made, the fee is absorbed into onboarding compensation; if not, candidate is paid for the test on completion.
5. Use the deliverable as Stage-3 evidence in the multi-round process (Initial chat → Founder call → Paid test → optional external advisor → Trial month → Permanent).
6. Trial month follows with explicit KPIs.

• Adapt the dollar range to the target market — $100–300 is calibrated to AI-native rates outside the top-tier tech metros; for Bay Area senior eng the floor is ~$1,000. Pick a number that's friction-free for the candidate (won't make them think "is this worth a day?") but small enough that the company can run 3–5 paid tests per role without budget review.