How Behavioral Interviews Work and What They Test

Many developers spend months on algorithms and ten minutes on the "soft" round. Then the soft round decides the result. A strong coder who sounds vague, blames others, or cannot explain what they personally did will often lose to a slightly weaker coder who tells clear, honest stories.

This series is about that round. There is no code to run. Instead, every lesson gives you complete spoken answers you can read aloud, a good version next to a weak version with a short reason, a worksheet to fill with your own stories, and a self-check list. All the people, teams and products in the examples are made up. The numbers in them are made up too. Use them as shapes, never as facts about your own life.

This is lesson 0 of nine. It explains what interviewers score and how stories are built. The next lesson, Tell Me About Yourself, Why This Company and Why Are You Leaving, covers the three questions you will hear in almost every interview. After that, failure, conflict and feedback and ownership, leadership and impact fill your story bank with the stories that carry most rounds.

Two other series on this site help here. Career Lesson 0 (resume) shows how to write the bullets that your stories will expand. Career Lesson 2 (mock interviews) shows how to practise out loud with feedback, which is the single most useful habit for this round.

Words used in this lesson

• Behavioral question is a question about how you acted in a real past situation, such as "Tell me about a time you disagreed with a teammate."
• Signal is the thing an interviewer is trying to learn about you from your answer, such as "takes ownership" or "listens to others."
• STAR is a story shape: Situation, Task, Action, Result.
• Reflection is an optional last step: what you learned or would do differently. Some people write the shape as STAR(R).
• Story bank is your short list of true stories, each tagged with the signals it can show.
• Scorecard is the sheet an interviewer fills in after the interview, often with a few skills rated one by one.
• Follow-up question is a deeper question about your story, such as "What did you personally do?" It tests whether the story is real.
• "I" versus "we" is the habit of saying what you did yourself, as against what the team did.
• Seniority signal is evidence in your answer of the level you work at: bigger scope, more people, harder trade-offs.
• Company values are the traits a company says it hires for. Some publish them as a list of principles.

What this round is really testing

The technical rounds ask: can this person solve the problem? The behavioral round asks different things:

1. Will this person work well with others, including when they disagree?
2. Will they own a problem, or wait to be told?
3. Do they learn from mistakes, or repeat them?
4. Can they communicate clearly to people who do not know the details?
5. At what level do they work: a single task, a feature, a whole system, a team?

The tech interview handbook says the same thing in other words. A company wants to know that you have a history of acting in the way that would make you succeed there. It also wants to place you at a level, such as junior, senior or staff. Both come from your stories, not from what you promise.

A company also does not want a very talented person who makes the whole team slower. The handbook points to the idea of a "brilliant jerk": someone whose skill does not pay for the damage they do. Behavioral rounds are one of the few tools a company has to spot that before hiring.

The three kinds of question

The handbook splits behavioral questions into three types. It is a useful way to think, because each type needs a slightly different answer.

1. Preferences and plans. "What are you looking for next?" "Where do you want to be in five years?" "Why this company?" The answer is a short, honest opinion. There is no STAR here. See lesson 1.
2. Resume deep dives. "Walk me through the payments project on your resume." The interviewer picks a line and pulls on it. Your story must have depth: why, how, trade-offs, result.
3. "Tell me about a time when..." The classic form. This is where STAR helps most.

Most people only prepare for type 3, then get surprised by type 2. If something is on your resume, expect to talk about it for ten minutes.

A map from question to hidden signal

Behind every question is a signal. If you know the signal, you know what to show.

The handbook's list of common questions (collected from public interview reports) has many of these. A few I took from it as real examples of how they are asked: "Tell me about a time you had a conflict with a co-worker", "What is the most constructive feedback you have received", "Tell me about a time you met a tight deadline", "How do you handle disagreements with co-workers", and "In your professional experience have you worked on something without getting approval from your manager?" You will see all of these in the later lessons.

The STAR shape and what it hides

STAR stands for:

• Situation: the background. One or two sentences.
• Task: what you needed to achieve, or what was your job in this. One sentence.
• Action: what you did, step by step. This is the longest part.
• Result: what happened, with a number if you can.

The handbook adds a point I agree with: the Actions are the real target of the interview. The company is not hiring you to repeat your old project. It is hiring you to repeat your old behaviours on its projects. So the actions have to show a behaviour that can be copied: "I asked each person for their real worry before proposing anything", not "we had a meeting."

It also suggests a second R, Reflection: end by saying what you learned. That is a cheap way to look more senior, because juniors tell what happened and seniors tell what it taught them.

Time split for a 90-second story

A spoken story at a calm pace is about 130 to 150 words per minute. So a 90-second story is about 200 words. A good split:

The most common error is the reverse: 60 seconds of background and 15 seconds of action. The interviewer does not care about the background. They care about you.

Where STAR goes wrong

STAR is a good start, but it has limits. Know them.

• It sounds like a script. If you recite S, T, A, R in a flat voice, it feels like a memorised speech. Use the shape inside, but speak like a person.
• It hides the "why". Interviewers often want your reasoning at each decision, not only the steps. Add a short "because" to your main actions.
• It rewards happy endings only. Some of your best stories end badly or in the middle. STAR can still work: the Result is "what happened" and the Reflection is "what I changed."
• It does not teach you what to leave out. A story with six actions and four side plots is a mess. Choose the two or three actions that show the signal and cut the rest.
• It does not fix "we". You can fill all four letters and still never say what you did.

Good versus weak: a full side-by-side

The best way to learn the difference is to see two full answers to the same question. Here is one question and two spoken answers. Both are fictional and about the same story.

Question: "Tell me about a time you had a tight deadline."

The weak answer (about 45 seconds)

Why it is weak:

• No specifics. What was the release? What was tight, and by how much?
• No "I". "We" appears many times, so we do not know what the speaker did.
• No decision. The only strategy is "work more", which is not a skill.
• Long hours as a badge. Many interviewers read this as poor planning or a risk of burnout.
• The result is "the manager was happy". There is no number and no cost.
• It ends with a claim ("I work well under pressure") instead of showing it.

The good answer (about 95 seconds, about 230 words)

Why it is good:

• Starts with real size: six weeks versus ten, and a date that could not move.
• "My job was" is clear. The listener knows the role.
• Three actions, each with a reason. Scope list, risk first, weekly updates.
• It shows behaviour a company wants: cutting scope openly, warning early.
• A result with a number and a cost (saved cards later), which is honest.
• It ends with a reflection that sounds like a lesson, not a boast.

Notice that the good answer is not louder or more heroic. It is more specific and it shows choices.

Company values: a public example

Some companies publish a list of the traits they hire for. Amazon's "leadership principles" are the best-known public example. I could not open the page from my machine, so I rely on the general, widely known fact that the list exists and that it includes ideas such as customer focus, ownership, diving deep into details, and disagreeing openly and then committing to the decision. If you are interviewing at a company with such a list, behavioral questions are often built from it. The handbook says the same: top companies use their values to score candidates, so you should find the values and prepare stories that show them.

Here is a method that works for any company, with or without a published list:

1. Open the careers page and the engineering blog. Write down the five words that repeat.
2. Next to each word, write one story from your bank that shows it.
3. If a value has no story, you have a gap. Find a story, or be honest and talk about a smaller example.

Do not fake a value. If a company says "we move fast and break things" and you are a careful person, do not pretend. Talk about how you move fast safely. Interviewers can smell a stretched story with two follow-up questions.

Here is a tiny example (fictional company, fictional values) of the mapping:

Build your story bank

This is the core of the whole series. A story bank is a short list of true stories from your work, each of which can answer many questions.

The handbook's advice is: do not try to memorise answers for every question. Pick three to five big projects with high impact, high difficulty and high involvement by you. Most questions can be answered from those, because a big project contains conflict, ambiguity, pressure, a mistake and a win, all inside one project.

I suggest six stories for a person with at least a couple of years of work. If you are a student or a fresher, use college projects, internships, open source, hackathons and part-time jobs. They count.

Step 1: collect

Open your resume, your old pull requests, your chat history and your last performance review. Write one line for each of these. Do not judge yet.

• A project you are proud of
• A time something broke in production
• A time you disagreed with someone
• A time you got feedback that hurt
• A time you helped someone grow
• A time you decided without enough information
• A time you took on something nobody owned
• A time you failed to meet a deadline
• A time you learned something fast
• A time you improved a process or a tool

Step 2: pick six and tag them

For each of your best six, fill in this card. A card is only a page of bullets. It is not a script.

STORY CARDName (3 words):           e.g. "checkout deadline"When / where:             e.g. "last year, online shop team of 5"My role:                  e.g. "main developer for checkout"Situation (1 line):Task (1 line):Actions (2-4 bullets, each with a "because"):   1.   2.   3.Result (with a number, or an honest estimate):Cost or what went badly:Reflection (1 line):What I would do differently:Signals this can show (tick):  [ ] ownership  [ ] conflict  [ ] failure   [ ] deadline  [ ] leadership  [ ] mentoring  [ ] ambiguity   [ ] feedback  [ ] impact  [ ] learning  [ ] customer focusPeople involved (first names are fine, no real private details):

Step 3: build the matrix

Put your six stories in rows and the common signals in columns. Put a tick where a story fits well. Here is a filled example for a fictional mid-level developer, Dev:

Look at the columns. Does any column have no tick? That is a gap. "Failure" has two ticks, good. "Conflict" has two. If a column is empty, go back to step 1 and find a story to fill it.

Look at the rows too. Story A fits four columns, so you will use it a lot. That is fine, but do not use the same story for every answer. If the interviewer asks three questions in a row and you give the same story, you look thin. Aim to have two stories ready for each common signal, so you can switch.

Step 4: say each one out loud

Reading a card is not the same as saying a story. Say each story aloud with a timer. The first try usually runs three or four minutes. Cut to about 90 seconds. Record yourself on your phone. Listen once. You will hear "um", you will hear "we", and you will hear where it gets boring. Fix those.

Do not memorise word for word. Memorised answers sound flat and break when the interviewer interrupts. Memorise the bullets and the first sentence. The handbook gives the same advice: write bullets, then practise saying them near the interview date so they come out naturally.

How to handle follow-up questions

Good interviewers do not stop at your first answer. They probe, to see if the story is real. Common follow-ups:

• "What exactly did you do?"
• "Why did you choose that option and not another?"
• "What would you do differently?"
• "How did the other person react?"
• "How do you know it worked?"
• "What was the hardest part?"

A real story survives these. An invented or inflated one does not. This is the main reason to use true stories and to prepare the "alternatives" and the "what I would do differently" lines. The question "what would you do differently" is a gift. A good answer is calm and specific: "I would have asked for a rollback plan before the migration, not after."

A short model for "What would you do differently?" (about 20 seconds):

A second side-by-side: "Tell me about a project you are proud of"

This is a very common question. Weak versus good, again fictional.

Weak (about 40 seconds):

Why weak: it lists tools instead of problems. "Complex" is a claim with no evidence. We do not know who the users were, why the dashboard mattered, or what the speaker decided.

Good (about 100 seconds, about 250 words):

Why good: a real problem with a measurable cost, a clear role, a decision with a reason, a technical detail the interviewer can dig into, and a result in two numbers. The tools appear only where they matter.

Mistakes that cost offers

A short list of things that come up again and again in interview feedback:

1. No answer to the actual question. You told a good story, but about something else. Listen to the question and tag the signal first.
2. Too long. Over three minutes without a pause. The interviewer cannot interrupt politely. End at around 90 seconds and ask "Would you like more detail on any part?"
3. Blaming. Even when someone else was wrong, a story that makes another person the villain makes you look hard to work with. See lesson 2.
4. Only wins. If every story is a triumph you look rehearsed or unaware. At least two of your six should be about mistakes or hard moments.
5. Hypotheticals. "I would usually..." is not a story. If you are asked about a time, give a time. If you have no matching story, say so briefly and give the closest one.
6. Lying or inflating. Interviewers cross-check, call references, and ask follow-ups. A small stretch can become a big problem.
7. Vague numbers or none. "Improved performance" says nothing. "Cut page load from eight seconds to three" does. If you do not know the number, give an honest range or say "I estimate".
8. Confidential details. Do not share secrets, customer names or internal numbers that your old employer would not want public. You can say "a mid-sized online retailer" and use rounded numbers. Interviewers respect this.

Reading the room

Different interviewers have different styles. A few things you can do in the first minute:

• If they say "I'd like to hear about a time...", they want a story. Use STAR.
• If they say "How do you usually handle...", they want a short principle, then an example. Give your rule in one sentence, then a quick example.
• If they interrupt you early, they have heard enough of the background. Skip to the action.
• If they nod and say nothing, finish the story and ask if they want more.
• If you do not understand the question, ask. "Do you mean a disagreement about design, or about priorities?" is fine, and it buys thinking time.

It is also fine to pause for five to ten seconds. "Let me think of the best example" is a strong sentence. Silence feels long to you but looks calm to them.

Try it

1. Write your six story cards using the template above. Do not try to make them perfect.
2. Build the matrix with the six signals: failure, conflict, deadline, leadership, ambiguity, impact. Find your gaps.
3. Pick the story you are least sure about. Say it aloud for 90 seconds and record it. Count how many times you said "we" and how many times you said "I".
4. Take the weak answer to "tight deadline" above. Rewrite it for a story of your own, with a size (days or weeks), a role, three actions with reasons, and one number.
5. Write your "what I would do differently" line for each story.

Answers

There is no single right answer. Check your work against these points:

1. Each card should have a result and a cost. If you have only good news, look again for what it cost you or what went wrong.
2. A healthy matrix has at least two stories per column for the most common signals (conflict, failure, leadership). If a column is empty, it is the first thing to fix, not the last.
3. A normal first count is "we" at least twice as often as "I". Aim to have your own actions clearly in "I" and the team result in "we".
4. A good rewrite has a number for the time, a clear first-person job, and a choice that cut or ordered the work. If your rewrite still says "we worked hard", it is still the weak version.
5. "What I would do differently" should be small and concrete, such as "ask the question earlier", not "nothing" and not "everything."

Self-check list

Before any behavioral interview, tick these:

• I have six stories on cards, each with a role, result, cost and reflection.
• At least two of them are about mistakes, conflict or hard feedback.
• Each story is under 100 seconds when said aloud.
• I can say what I did in each in one sentence that starts with "I".
• Each has at least one number or an honest estimate.
• I know the signals a company values and which story matches each.
• I have said my stories aloud, not only read them.
• My stories are true and I could defend each against five follow-up questions.
• I removed confidential names and exact private numbers.
• I have a one-line answer to "What would you do differently?" for each story.

Interview questions

Below are real behavioral questions, with first-person model answers. Each is short here (about 40 to 60 seconds). In an interview you can extend each to 90 seconds using the shape from this lesson. All details are fictional.

Key takeaways

• Behavioral rounds test past behaviour as a guide to future behaviour. Signals like ownership, conflict handling, learning and communication matter most.
• There are three question types: plans and preferences, resume deep dives, and "tell me about a time." Prepare for all three.
• Use STAR, but spend most words on your actions and reasons. Add a short reflection.
• Keep six true stories in a story bank, tagged by signal, with at least two about mistakes or hard moments.
• Say the story aloud, aim for about 90 seconds, and keep the bullets, not the exact words.
• Prepare "what I would do differently" for every story. Follow-up questions are where interviews are won.
• Use "I" for your actions and "we" for team results. Do not blame, do not inflate, do not reveal secrets.

What's next

The next lesson is Tell Me About Yourself, Why This Company and Why Are You Leaving. These three questions open most interviews, so you should have them ready before anything else. After that comes failure, conflict and feedback. Later, lesson 8 puts everything into a full mock interview. If you lead people already, Team Lead Lesson 8 (1:1s, feedback, delegation and conflict) shows the same skills from the manager's side, which helps you understand what interviewers are listening for.