SOP Library

Hire, pay and let go 7 of 16 in this group

SOP 174

Standardize selection and the founder's final interview

What this page is for. Use it at 100 to 249 people, when the people recruiting for the business no longer hold one standard, so the wrong candidates get too far. It carries why the founder should not step back from hiring yet, the same questions for every candidate, interview panels, the founder's short final interview, a simple scoring rubric and how the founder reads its scores, and one personal habit of asking about the one red flag. The size this belongs to, and every other function at it, are on SOP 168.

SOP-174-Standardize-selection-and-the-founders-final-interview.md

1. The constraint and what graduates it

Recruiting is how you find and hire people for your company.

Item At this stage (100 to 249 people)
The constraint Recruiters hold inconsistent standards and the culture suffers; wrong candidates are getting too far along in the process
To graduate Standardize selection, and add a founder fit check

2. Keep the founder in the hiring

This cannot be hammered in enough, it is said: how often founders get uninvolved with recruiting too soon. In some of the best companies that have been built, the founder did the final check, even if it was only 15 minutes, and went on doing it for people joining the team, especially anyone in a management position.

What follows, in the order given: structured interviews, panels, the founder's final interview and a scoring system.

3. Ask every candidate the same questions

The first is structured interviews: the same questions for all the candidates for a given role, so you can assess the key competencies. Take each of the four kinds of fit and check that every interviewer asks the same five to 10 questions. If they do not, each interviewer ends up biased by different answers to different questions, and you need a process that protects you from your own emotions. So write standard questions for each type of interview, for every role that exists.

4. Interview in panels

The second, to lower the risk, is panels: several interviewers, each with an assigned role. For example:

  • two or three people conduct a culture interview together;
  • two people conduct a technical interview together.

The reason is to remove as much individual bias as you can. A lot of people like to hire people like themselves. Fill your panels instead with people who have an incentive to hire someone competent. If you are hiring into a sales manager role, bring in the customer service manager and the IT manager. They also want someone very proficient in that seat, but they might not be as desperate as the director of sales to fill it, so they protect you from lowering your standards.

That matters at this size, because people are desperate to fill roles: the recruiters want to show some wins, and people badly need some help. Put people on the panel who have no incentive to fill the role that fast, so that they hold the standard for the entire company.

5. The founder's final interview

The third: the founder does a final interview of at least 20 minutes, to make sure nothing is missed. When standards are wavering, the reason to get back in the room is to find out where the ball is being dropped:

  • Is it culture, or is it skill?
  • Are the interviewers not asking the same questions?
  • Is something not yet standardized that you are missing?
  • Are the recruiters missing it? Are the specialized recruiters less engaged than at the beginning, when they were trying to win your business?

Taking the final interviews is how you get that feedback. This does not mean a two-hour interview for every role. It means 20 minutes. One way, and usually the practice described: set aside one day a week and stack five or six of them back to back in the morning, as a final check.

The reason given: at this size, nothing matters more than hiring and firing the right people. If allocating resources is your job, then bringing in people who either take or give more resources to the company is probably one of the most useful moves you can make. So until the structured interviews are in place and humming, keep doing the founder fit check as well.

6. Score every candidate on one simple rubric

The last, and arguably the favorite in one practice here, is a scoring system, so that you move from hiring people subjectively to hiring them objectively: an interview rubric that scores answers on a consistent scale.

There are so many fancy ways to do this, and in the view here they are hateful. A fancy quiz gets built, and people honestly do not want to fill it out properly, so they do not think about it and do not do it. You need something simple enough that the people on your team will actually take the time to fill it out. The favorite way here:

  • Skill: five scores. Take the top five skills the role needs and score the candidate on each.
  • Culture fit: one score per core value. It need not be five. However many values you have, that is how many culture scores you give.
  • The scale: 1 to 5. How does the person rate on this skill, from one to five? How do they rate on this value?

Then you can look at the scores of each candidate in the final round. In one practice here this is done all the time, even as a check on one's own view: many times a candidate was really liked, then the scores were written down, and who came out higher was a surprise.

7. Liking a candidate is not the same as their skill

A lot of the times, what you will see at this point, and your team might not know it yet, is that liking someone blinds you to their actual skill and to what they will really bring. Some people are very likable and charismatic, and very good at interviewing and at making you like them, but not very good at doing the job. Interview skill does not equal the ability to do the job; in one practice here, interviewing is thought of as a skill entirely separate from the job you are hiring for.

Some of the best hires in one practice here had honestly horrendous interviews. The question asked was an objective one: does this person have this skill, and this one, and what experience do they have, rather than whether the interviewer simply loved the person. You will know the ones you loved in the interview who came in and were not that good at the job. Getting there has been an evolution, in one practice here. A lot of times you enjoy talking to someone who does not actually bring a lot of value to the business, and that is why a scoring system matters so much at this point.

8. Read the gaps between the scores

With structured interviews, panels and the founder's check in place, score candidates after each interview. Because the process has several steps, several people score each candidate, and you can look at where their scores disagree. That serves the founder at the final check: if one score was the lowest across all three of the interviews, push on that one really hard. It also lets you spend your time on the things that give you the best return.

9. A personal habit: ask about the red flag

One person here loves to watch all the interviews before interviewing someone, work out the one thing that is a big red flag, and ask about it directly. The example given: why does it seem you are so defensive about feedback? The point is to see, by asking, whether the person actually is defensive about feedback. It is called a great use of time. It is a personal habit, not a step in the process above.

10. Where this sits

  • The stage as a whole: SOP 168 — Specialize the business at 100 to 249 people.
  • The four kinds of fit (expectations, culture, skill and potential) are set out, as a short screen and one interview, in section 9 of SOP 149, at five to nine people. There skill is one part of that interview and covers usually the top three to five skills a job breaks into; the rubric here scores the top five.
  • At an earlier size, SOP 163 — Head-hunt and interview for higher-level roles splits the four fits across four interviewers, one fit each, often with the founder or chief executive on potential. This page puts two or three people together on a culture interview and two on a technical one. The counts differ, and each is carried as stated. The founder doing the final interview is advised there too, for that size; here it carries on at 100 to 249 people.
  • Whether to hire a skill in or train it, and the interview test for a senior hire: SOP 44 — Buy or build talent.

11. What this page does not decide for you

  • The questions themselves. No list of the five to 10 questions is given. Not established on this page.
  • Which three interviews the founder compares. The final check refers to all three; they are not named. Not established on this page.
  • A passing score. This page gives no figure for a passing score, or for how skill weighs against culture.
  • Which roles get the founder's final interview. The check is described for people joining the team, especially anyone in a management position; no list of roles is given. Not established on this page.

12. What this page does not cover

Reference checks, job offers and employment law are not covered on this page.

Terms defined on this page

Interview panel
Several interviewers with set roles, such as two or three on culture and two on technical skill, including people in no hurry to fill the seat, so standards hold.
Interview rubric
A simple score sheet: the top five skills and each core value, each rated 1 to 5, so hiring turns objective.
Structured interview
Every candidate for a role gets the same five to 10 questions for each kind of fit, so answers can be compared without bias.