Interview guide for startup roles: a loop for early employees when there is no hiring team
On this page
- What is different about early hires
- The loop at a glance
- Stage 1: founder screen (30 minutes)
- Stage 2 and 3: the real-problem work sample and debrief
- Stage 4: ownership interview (45 minutes)
- Scorecard competencies and weights
- Who interviews when the team is five people
- References and backchannels
- Legal points that change as you grow
- The decision rule and the offer
- Mistakes that weaken startup loops
- Questions people ask
Early startup hiring breaks most interview advice. There is no recruiting team, no interviewer pool and no leveling guide; the founders interview between investor calls, and every hire changes the company. The usual failure is a loop made of friendly conversations, where the decision rests on whether a founder "clicked" with the candidate. The opposite failure is copying a large company's six-stage loop and losing every good candidate to a faster offer. This guide gives founders and early managers a short, structured loop for the first 5 to 50 employees, whether the role is engineering, sales, operations or design, with the founder's part, a paid real-problem work sample and the legal thresholds that change as you grow.
If you are hiring your first executive, such as a VP of sales or a CFO, use the interview guide for executive roles, which includes the board and a first-90-days presentation. For how hiring fits the runway, see the hiring plan example for startups.
What is different about early hires
- The job will change. The person you hire for one problem will own three others within a year. Test learning and range, not only the current task.
- Nobody will manage them closely. Founders are busy. Test whether the candidate can set their own priorities and say what they need.
- Every hire sets a precedent. Early pay, titles and process decisions become the norm. Write them down.
- The candidate is taking a risk. They are interviewing you too. Be honest about runway, stage and what could go wrong.
The loop at a glance
| Stage | Interviewer | Owns | Length | Pass rule |
|---|---|---|---|---|
| 1. Founder screen | A founder | Motivation; stage fit; pay and equity expectations | 30 min | Wants this stage, with open eyes |
| 2. Real-problem work sample | The person closest to the work | Core skill; judgment under ambiguity | 2–4 hours, paid | At least 3 on core skill |
| 3. Work sample debrief | Same person plus a founder | Reasoning; response to challenge | 45 min | At least 3 on reasoning |
| 4. Ownership interview | A founder | Ownership; prioritizing; learning speed | 45 min | Nothing below 2 |
| 5. Team conversation | One or two current employees | Collaboration; communication | 30 min | Nothing below 2 |
| References | A founder | Ownership and range in past roles | 2 × 15 min | No unresolved concerns |
Aim to finish in ten business days. Without an applicant tracking system, keep scorecards in a shared folder and collect them before anyone discusses the candidate; see running structured interviews without an ATS.
Stage 1: founder screen (30 minutes)
- "Why a company at our stage, and why now?" Listen for: an informed reason. Candidates who want stability or a defined ladder deserve to hear that you cannot offer it yet.
- "Tell me about a job where your role changed a lot while you were in it. What did you do?"
- "Here is our stage, runway in months and the main risks as we see them. What questions do you have?" Share what you can; good candidates ask about it anyway.
- "The range for this role is [salary] plus [equity range]. Does that work?" Settle it now.
Stage 2 and 3: the real-problem work sample and debrief
Take a problem the company actually has, scale it down, and remove anything confidential. Examples: for an early engineer, extend a small version of your codebase with a feature that has an unclear requirement; for a first marketer, a launch plan for a real upcoming release with a stated budget; for an operations hire, untangle a messy spreadsheet of vendor contracts and propose a process.
Leave one important thing unspecified on purpose, and say that questions are welcome. Score:
- Core skill: the quality of the work at the level you need now.
- Judgment: what they chose to do first and what they left out, with reasons.
- Handling ambiguity: asked about the gap, or made a stated assumption.
Pay for the exercise if it runs beyond a couple of hours or uses real company material; the job trial day guide covers paid trials. For scoring take-home work consistently, see how to evaluate a take-home assignment.
In the debrief, ask them to walk through the work, then push back once on a key decision. Score whether they defend it with reasons or change it with reasons; both can be good. Changing their mind without reasons, or refusing to engage, is not.
Stage 4: ownership interview (45 minutes)
- "Tell me about something you built or fixed that nobody asked you to. How did you decide it mattered?" Listen for: a real problem and a cost-aware choice, not busywork.
- "Tell me about a time you had more work than time and no manager to ask. What did you drop?" Listen for: explicit trade-offs and telling people.
- "What is something you learned from scratch in the last year? How did you learn it?" Listen for: a method and a timeline.
- "Tell me about a decision you disagreed with at work. What did you do?" Listen for: saying so directly, then committing.
Scorecard competencies and weights
Example weights for an early generalist hire. Lock yours before the first candidate.
| Competency | Owned by | Example weight |
|---|---|---|
| Core skill for the current need | Work sample | 30% |
| Judgment under ambiguity | Work sample; debrief | 20% |
| Ownership without close management | Ownership interview; references | 20% |
| Learning speed and range | Ownership interview | 15% |
| Collaboration and directness | Team conversation; debrief | 15% |
"Culture fit" is not a row. At this size it usually means "similar to the founders", which narrows the team when you most need range.
Who interviews when the team is five people
Small teams run out of interviewers fast, and the same two founders end up in every stage. Split the competencies so each person scores only their rows, and write scores before talking to each other. When nobody on the team has the skill you are hiring for, such as a first designer or first finance hire, ask an adviser or an experienced contact to review the work sample against your scorecard. Pay or thank them, and give them the same rubric the team uses, so their opinion is evidence rather than a verdict.
References and backchannels
Founders often call mutual contacts the candidate did not name. If you do, apply the same questions to every candidate, weigh one secondhand opinion lightly, and never contact a current employer without the candidate's permission. Ask referees about ownership and range: "What did they take on that was outside their job?"
Legal points that change as you grow
Checked against the linked primary sources as of October 2026. Not legal advice.
- Federal thresholds. Title VII applies to employers with fifteen or more employees for each working day in twenty or more calendar weeks (42 U.S.C. 2000e(b)), and Title I of the ADA uses the same 15-employee test (42 U.S.C. 12111(5)); the Age Discrimination in Employment Act applies at twenty (29 U.S.C. 630(b)). Count employees in the current or preceding calendar year.
- State laws reach smaller teams. California's Civil Rights Department says employers of five or more are subject to FEHA's prohibition on discrimination, and harassment is prohibited in all workplaces. Run a structured, documented process from your first hire.
- Form I-9 from hire one. USCIS states that all U.S. employers must complete Form I-9 for every individual they hire for employment in the United States.
- Pay ranges. Posting rules depend on where the role and the company are; see pay transparency laws by state.
The decision rule and the offer
- Floor: at least 3 on core skill. Potential does not cover a gap the company needs closed now.
- Weighted total: in this example, 3.0 or higher on a 1–4 scale; early hires warrant a higher bar.
- Founder veto, written: a founder may decline a candidate who passed, but must write the job-related reason on the scorecard.
- Equity in writing: share count, vesting schedule and cliff, exercise price and total shares outstanding, with no predicted value. The offer letter template has a section for it.
Mistakes that weaken startup loops
- Hiring friends of friends without a work sample. Familiarity is not evidence.
- Overselling. A candidate who joins on a promise you cannot keep leaves when they notice.
- Long unpaid take-homes. Strong candidates with jobs decline them.
- No written scorecard. At ten people, inconsistent decisions are noticed and remembered.
Questions people ask
How should a startup interview early employees?
Keep the loop short but structured: a founder screen, a work sample on a scaled-down version of a real problem, a structured interview about working with ambiguity, and a conversation with one or two current team members. Agree on the scorecard before the first candidate, because a small team's informal impressions drift quickly.
Should startups pay candidates for work samples?
Pay for anything longer than a couple of hours or anything that produces work you might use. It is fairer to candidates who cannot work for free, it widens your pool, and it removes any doubt about whether you are getting free work. Keep shorter exercises on invented problems.
Do anti-discrimination laws apply to small startups?
Federal thresholds depend on headcount: Title VII and the ADA apply at 15 employees and the ADEA at 20, but many state laws reach smaller employers. California's FEHA, for example, applies to employers of five or more, and its harassment rules apply to all. Form I-9 applies to every hire. As of October 2026; not legal advice.
How should equity be explained in a startup offer?
Give the number of shares or options, the vesting schedule, the cliff, the exercise price if any, and the total shares outstanding so the candidate can calculate their percentage. Do not predict a value. Put it in writing and let the candidate ask questions of someone who understands the plan.