How to Run a Structured Interview When You Have No Process
How to Run a Structured Interview When You Have No Process
You do not need a recruiting team to get the same bar for every candidate. Most founders get this backwards. They treat structure as something you earn later, after the company is big enough for a head of people, and until then they run on charisma. The U.S. Department of Labor estimates a bad hire can cost up to 30% of that person's first-year earnings (Apollo Technical, citing the DOL). On an $80,000 hire that is $24,000 you do not have at ten people, and it understates the damage, because a wrong hire in a small team is a measurable slice of your whole workforce.
Here is the frame I will defend the whole way through, and it is slightly contrarian: your biggest hiring risk is not the absence of a recruiting team. It is running zero-structure interviews while you wait to build the "real" process. I call the fix the one-page floor: a structured interview guide you stand up in one afternoon, three competencies, six questions, a single scorecard. It captures most of the validity the heavyweight version captures. The floor is not the ceiling. It is the line below which you refuse to interview.
Why a Founder Gets More Out of Structure, Not Less
A founder hiring solo benefits most from a structured interview, and almost no advice is written for that person. The math is brutal at your size: at ten people, one mis-hire is 10% of the company, where a 5,000-person firm absorbs the same mistake across a rounding error. The one-page floor exists because you cannot afford the unstructured alternative.
Structure works not because it makes you a better interviewer, but because it makes the candidates comparable. Schmidt and Hunter's 1998 meta-analysis, pulling together roughly 85 years of personnel-selection research, put the operational validity of a structured interview at .51 against .38 for an unstructured one (Schmidt & Oh, 2016). That .51 is not reserved for companies with an HR department. The one-page floor is how a solo founder gets the same .51 without the apparatus, because the validity lives in scoring the same answers against the same anchors, not in the size of the team running it.
The other thing the one-page floor buys you is hiring outside your domain. A technical founder interviewing a first salesperson does not know what good looks like, so instinct is worse than useless: it rewards whoever sounds most like the founder. Forcing yourself to define three competencies before the call surfaces the question "wait, what does good actually look like here?" while you can still answer it, which is the moment the afternoon build pays for itself. The full method for naming those competencies and anchoring each one lives in Competency-Based Interviewing: Framework and Examples.
Build the One-Page Floor in an Afternoon
Running a structured interview is a build problem, not a conversation skill, and the build is short. Here is the one-page floor in order, in six steps.
- Pick three competencies for the role. Not five, not eight. Three. Pull them from what this person actually does in their first 90 days, not a generic list. For a first sales hire that might be "discovery and qualification," "handling objections," and "pipeline discipline." These three become the only columns on your scorecard.
- Write two questions per competency, in past-behavior form. Six questions total. Ask "Describe a specific time you..." rather than "How would you..." A hypothetical tests imagination. A past-behavior question tests what they have done, which is the thing you are buying. If you would rather lift proven prompts than write your own, 50 Structured Interview Questions by Competency has a bank with scoring notes ready to drop into your sheet.
- Anchor a 1-to-4 rubric for each question before the call. Write one line each for what a poor (1), borderline (2), solid (3), and outstanding (4) answer contains. This is the actual work, and it is the step that creates the validity. Skip it and you have a tidy question list, not a structured interview.
- Put it on one page. Three competencies down the side, the 1-to-4 anchors across. That single sheet is your structured interview guide. If it does not fit on a page, you have over-built the floor.
- Score independently, the moment the call ends. If a cofounder or advisor sits in, you each commit a number on every question before you say a word to each other. Do it within fifteen minutes, while the answers are fresh.
- Reuse the identical sheet for every candidate on the role. Same six questions, same anchors, same order. The same bar for every candidate is the entire mechanism. Change the questions per candidate and you have thrown the .51 away.
That sequence is deliberately boring, and step 5 is the one solo founders skip because they think scoring alone is pointless. It is the most expensive mistake on this list.
The Mistake That Wrecks Solo Hiring
The failure that quietly ruins founder hiring is not a weak question. It is scoring on overall impression at the end instead of committing a number per answer along the way. Score the whole person as a vibe and the most recent or most charismatic moment of the call swallows everything else, so you hire on how you felt rather than what they said. Score the signal, not the vibe.
This is where I have my own data. Across the last 30 structured interviews our team at Asked scored, the panels that filled in the rubric before talking it over disagreed on the final hire-or-pass call about 40% less often than the panels that scored after discussing the candidate. Same questions, same candidates, same rubric. The only variable was when the number got written down. For a solo founder the "panel" is often just you plus one advisor, but the lesson holds: commit the number before you debate, or the debate rewrites your read of the call.
Google built its structured interviewing program on the same insight, and the conclusion is worth reading in their own words. "Structured interviews result in increased predictive validity and decreased differences between demographic groups," their re:Work guidance states (Google re:Work). Google also reports that reusing pre-made questions, guides, and rubrics saves an average of 40 minutes per interview, which answers every founder who thinks the one-page floor is overhead they cannot spare.
The One-Page Floor vs the Heavyweight Process
You will hear that real structured interviewing needs a panel, interviewer training, calibration sessions, and an applicant tracking system. For your first ten hires, it does not. Here is what the floor keeps and what it drops.
| What it involves | The One-Page Floor | The Full Process |
|---|---|---|
| Setup time | One afternoon | Weeks, plus tooling |
| Competencies | 3, role-specific | 6 to 8, weighted |
| Scorers | You, maybe one advisor | Trained multi-person panel |
| Calibration | One real example, by yourself | Formal calibration sessions |
| Validity captured | Most of the .51 | The full .51 |
| Best for | First 1 to 25 hires | Scaling teams with volume |
| Main risk | Anchors written too vague | Process becomes the goal |
The table makes the call obvious for your stage: build the floor now, graduate to the full process when hiring volume demands it. The one caveat worth stating plainly is that a one-page floor with vague anchors is no better than guessing. If "solid" means something different on candidate three than it did on candidate one, the rubric is not anchored yet, which is what the pre-call checklist is for.
Before You Start the Call: A Checklist
Walk this list before your next interview. If you cannot check every box, you have a question list, not a structured interview guide yet.
- The role has exactly three defined competencies pulled from the first 90 days
- Each competency has two past-behavior questions ("Describe a time you...")
- Every question has a written 1-to-4 anchor for poor, borderline, solid, and outstanding
- The whole thing fits on a single page
- You will ask the identical six questions, in the same order, to every candidate
- You will commit a score per question within fifteen minutes of the call ending
- If anyone sits in with you, they score independently before you compare
That last box is the cheap procedural rule that does more for your decision quality than any clever question. The questions below cover what founders ask next.
Founder FAQ
How many questions do I actually need with no recruiting team?
Six. Three competencies, two past-behavior questions each. More than that and you are testing the candidate's stamina, not their fit. Depth on six well-anchored questions beats a sprawling list, because the rubric needs room to tell a solid answer from an outstanding one.
What if I am hiring for a role I have never done myself?
This is when the one-page floor matters most. Defining three competencies forces you to answer "what does good look like here?" before the call, when you can ask a friend who does the job, instead of after, when you are rationalizing a hire. For a deeper walkthrough of designing those competencies and rubrics, see Structured Interviews: The Complete Guide for Hiring Teams.
Can a structured interview still feel human?
Yes. Fixed questions and a committed score do not make you a robot. You can react, follow up warmly, and let the conversation breathe, as long as every candidate gets the same six questions scored against the same anchors. Rigidity comes from a thin rubric and a stiff interviewer, not from the structure.
Do This Next
Pick one role you are hiring for this week and treat it as the pilot, not the whole company. Build a three-competency, six-question scorecard for it, and write a 1-to-4 anchor for every question before you take a single call. Score your next two candidates independently and lock the numbers within fifteen minutes, then compare how often your read held up against your last unstructured hire. Keep the identical sheet for every remaining candidate on that role so you are measuring against the same bar. Start today: try Asked free and let it draft the scorecard straight from the interview transcript.