The resume is not the person. The interview is not the job. — David Siegel
A friend recently pointed me to Paul Petrone’s article, “We are terrible at interviewing. Here’s how to fix it.” Petrone, who writes for LinkedIn, points out that a combination of interview methods works better than one single approach. He explains that unstructured interviews are worse than no interviews at all, and that the structured interview has a far better chance of finding people who will be suitable for a given job.
In this essay, I want to emphasize that the kind of structured interview matters, and that most people are doing the less-effective kind for building a resilient culture. I will also show how to hire people with no resumes or interviews at all. I will end with a suggestion to LinkedIn and to employers for experimenting with a new method of screening candidates.
Note: Petrone’s article is short and worth reading now if you’re not familiar with structured interviews.
The Hiring Process is Broken
We’ve known for a long time that unstructured interviews are no better than using astrological signs to pick good hires. Yet, in a recent Deloitte survey, HR professionals themselves recently gave themselves a score of 1.65 on a scale from 0 to 5. The hiring process is still broken. People are not resources. The resume is not the person. The interview is not the job. Plenty of people ace the interview and don’t last more than six months. We have a long way to go to learn what really correlates with happy, productive employees.
Much of what we have learned, we’ve learned from Laszlo Bock, the SVP of People Operations at Google, and his team of workforce scientists. Bock’s new book, Work Rules!, is a must-read for anyone who wants to learn how to hire people and how not to hire people. Much of what I’m talking about relates directly to chapter 5: “Don’t Trust Your Gut: Why our instincts keep us from being good interviewers, and what you can do to hire better”
I have tried to explain much of this in my series on employee satisfaction. To sum up: people are going after the wrong jobs, gaming the hiring process, and getting the very jobs that will make them least happy. As Bock himself says: just memorize the most likely structured-interview questions, and then while giving your pre-scripted answers, look around the person’s office and try to find any points you may have in common with that person, because people hire people who are like them. In other words, to win the game, turn a structured interview back into an unstructured one. There are plenty of consultants who will help you prepare for a structured interview, so you can tell interviewers the stories they unconsciously want to hear. The trick is to build a system that’s immune to these effects.
At Google, they hire a lot of technical people, and those people spend plenty of time showing their code, doing coding problems, and discussing their code with the Googlers who interview them. This is better than asking about the past. For non-technical hires, they use a series of interviews. For many hires, they use a double-blind approach, where one team interviews and a second team makes the decision. As Bock says:
The goal of our interview process is to predict how candidates will perform once they join the team. We achieve that goal by doing what the science says: combining behavioral and situational structured interviews with assessments of cognitive ability, conscientiousness, and leadership.
Behavioral interviews ask about the past. Situational interviews ask people to imagine real work scenarios and say how they would solve them. (I note that Bock’s concept of leadership is the same as mine, and that they don’t put much faith in background checks, which aren’t mentioned in his book at all.)
In my limited experience, companies do far more behavioral interviews than situational interviews, possibly because situational interviews require more work to create, or possibly because behavioral interviews are closer to the old, unstructured interviews. If you’re hiring for skills, you might want to use structured behavioral interviews. But if you’re building an agile company and hiring for culture fit first, or if, like Google, you’re looking for “curious generalists” over specialists, I believe you should try some experiments to see what combination works best for various positions. To do this, I have two suggestions: 1) try before you buy, and 2) use a situational question as a screening tool right up front. I’ll evaluate them a bit more after I describe them.
Try Before You Buy
Rather than interviewing, you can try people and see whether you enjoy working with them and they enjoy working with you. Richard Sheridan, in his amazing book,Joy Inc.: How We Built a Workplace People Love, explains that his company, Menlo Innovations, simply puts a day on the calendar where people can show up and get to work. The first day, they introduce the company in a short talk, then people do mock exercises alongside Menlo’s people, who are looking for what they call “kindergarten skills.” At Menlo, they hire programming generalists — people who work well with others and can learn new skills and systems quickly. They don’t want specialists. They want cooperative, supportive problem solvers. After this first day, a certain number are invited to spend two weeks working at the company, paid, as an evaluation. After this period, they decide whether to hire, offer people contract work, or thank them and say goodbye.
There are many creative ways to try before you buy. Orchestras conduct auditions behind a screen, so they can only hear people play and aren’t biased by other factors. You can conduct mock scenarios — some companies have current employees role play to interact with candidates. You can give them take-home assignments. You can brainstorm on ways to put people into similar situations that they might find in their new jobs. But my suggestion here is really to work with them for a period of time and see how it turns out. Not just for technical people, but at all levels. How many times have we seen the resume/interview approach fail at the executive level?
One thing they have learned at Menlo Innovations: if they find two new people working well together, one helping the other succeed, they have just found two good employees. Compare that to your average hiring process.
Ask yourself this question: if you had to hire people with no resumes and no interviews, what would you do? How can you try people and see whether they are a good fit for your culture? I ask you to force yourself to answer this question, even if you never do anything about it. You’ll probably learn something valuable just from doing the thought experiment.
Looking in the Rear-View Mirror
As Bock confirms, much of what people did in the past is less relevant than we think. If you’re looking for an air traffic controller or a heart surgeon, then of course you need people who have experience and skill. But in many cases, hiring the most skilled or the most successful person doesn’t lead to an agile culture. Teams and teamwork are more important than ever. Bock again:
By far the least important attribute we screen for is whether someone actually knows anything about the job they are taking on. Our reasoning and experience is that someone who has done the same task — successfully — for many years is likely to see a situation at Google and replicate the same solution that has worked for them. … In contrast, our experience is that curious people who are open to learning will figure out the right answers in almost all cases,
Focusing on previous outcomes is particularly misleading, since luck plays a huge role in outcomes. Plenty of executives get hired based on their resume and interviews, and it doesn’t always work out as planned. This is the track-record bias: being smitten by people with successful outcomes overlooks great people who simply didn’t get lucky and may have more potential.
Example: a friend of mine put a team together to try to help Washington, DC, win their bid for hosting the US Olympics in 2024 and lost to Boston. Is my friend and his team responsible for the USOC’s arbitrary decision? Of course not. They did a stunning job. From a hiring perspective, he’s every bit as capable as the person who headed up Boston’s bid. This kind of thing happens all the time in business, yet we are more impressed by the winners. Don’t succumb to the track-record fallacy — people who have tried hard and failed have usually learned more than those who have succeeded.
Many times, I have seen this senior consultant job description: “You have an MBA from a leading school and have made a quick succession up the ranks at a well known consulting company.” These companies — and their clients — get the people they deserve.
I once hired a writer who had a Pulitzer Prize. She had been on a team of journalists that had covered a large earthquake, and everyone on the team got the prize. To my great misfortune, she couldn’t write to save her life. My bad. I should have asked her to write something before hiring her.
As Petrone points out in his article:
In an exhaustive study that analyzed 85 years of research, college professors Frank Schmidt and John Hunter found that the quality of a work sample was the best indicator of how well a person will do at an advertised job.
While Petrone acknowledges this, he also emphasizes behavioral interviews, and that’s where we disagree. I believe it’s better to ask forward-looking questions rather than backward-looking questions. While there is a lot of active investigation in this area, it’s legitimate to disagree. This is very difficult research to do. There are many complicating factors. It may turn out, for example, that behavioral (past-looking) questions are more predictive of success in a traditional, command-and-control work environment, and that situational interviews are more appropriate for lean/agile companies that value autonomy in younger workers.
I want companies to match the interview techniques to their cultures and the kind of people they are hiring. In the companies I work with, I believe situational interviews are more predictive than behivoral ones. Rather than asking questions about the past, ask about scenarios they might encounter working at your company (or with your customers) and get them to explain what they would do, or give them an assignment to complete. As Petrone acknowledges, this is the most important information a candidate gives, and I completely agree.
The situational question doesn’t have to be about a particular skill or tactical decision. It can be more cultural and broad. It can be a dilemma. It can be about helping others or helping customers. It can be about coaching, listening, or creativity. These softer questions may give you more insights into how people will fit into your company. A Google search for “situational interviews” will turn up much more advice on how to do it.
The answers each person gives by themselves aren’t going to tell you whom to hire. But if you have a group of good candidates, situational questions can help you find the one who will be the best fit for your company. To do it right, you must write down their answers, or get them to turn in their work. Ask all candidates the same situational questions, get the interviewer or the applicant to summarize their answers in writing, and then compare the answers horizontally. I expect after comparing a few questions, the right candidate will jump out at you. If the answers are very similar, either several candidates were very good, or most of the questions were not.
Here’s my number-one reason for emphasizing future-looking questions rather than looking at the past: people can’t prepare for them. While you can easily get a list of common structured behavioral interview questions and prepare your answers in advance, you can’t prepare for forward-looking questions.
Many web sites advise candidates to take a situational question and relate it first to something they did in the past, telling a story of a successful outcome. This turns a situational question into a behavioral one, which can easily benefit a clever candidate. Don’t fall for it! In real business situations, you don’t want people who tell stories of similar problems they solved in the past — you want people to look at the problem in front of them and explore the solution space.
A Situational Screening Tool
Now let’s take this one step further and use the situational question as a screening tool. LinkedIn can help. As I have said in a previous post, Linkedin is an important ecosystem, but it is not a healthy, functional one. There is one thing they could do to improve immediately: allow employers to put a situational question on their job listings and give them a checkbox for whether this is a required part of the application.
A situational question gives an applicant a chance to say how he/she might approach a particular challenge in the new role. This gives each applicant time to put her best foot forward and make a good impression with her thoughts on a relevant topic. It also gives her a glimpse into the kind of work expected of her. The goal is not to write perfect prose, but the applicant should put some effort into showing how she thinks and how she might be different from others. If a prospect is willing to respond to the challenge, you know you have someone who is sincerely interested in the job. Be sure to give a deadline, so they know how much time they have to compose their answer, and tell them they can’t modify their answers once they submit.
If you’re going to use situational questions as a screening tool, be sure to strip out name, sex, age, race, and anything else that may bias people reading the answers. The goal is to let people shine, regardless of any of these factors, to see if they might be a good fit for the particular job.
But beware — not everyone writes confidently. If writing is an important part of the job, use a written scenario. If not, you may want to make it verbal or at least tell people they won’t be judged on their writing but rather just their general approach and thinking. Video can also be an option.
[Employers: you can put this question right into your job description without waiting for LinkedIn to add this feature.]
A Science-Based Approach
This is workforce science: stop believing what everyone else says and learn from your own empirical data what works and what doesn’t. What works for Google may not work for your company.
Here’s an example of an experiment: put your situational question into your job request and tell applicants they must answer the question to be considered. Then have one group of people screen applicants according to their answers to the challenge question, without looking at resumes, and another group look only at resumes and cover letters. This will result in two very different rankings of the incoming applicant pool. Now you can to a randomly controlled experiment: take the half the applicants from the top of each list and run them through your interview process blind — without telling interviewers which list they came from. Look for the recommendations. Measure the outcomes. If you do this a few times, you may learn something. If this provides a signal, then you can start to experiment with a more forward-looking interview approach.
I’ve explained why I think asking forward-looking scenario questions are better than asking people to tell stories of the past. I believe most companies should screen candidates according to a written question before looking at resumes. To do it right, you must ask every candidate the same question and compare their written answers without knowing anything else about them. I’ve also challenged you to think about trying people before hiring them. While I hope that many companies will get rid of resumes and interviews altogether, I understand that most won’t. Because people and jobs are so complex and matching is so difficult, I recommend doing experiments to determine what works best.
Read next: Triple your Productivity with Kanban!