Skip to content

The two-minute project pitch

A two-minute opening for a project walkthrough: the setting and the problem, your part, the one decision that shaped it, and the outcome. It gives the interviewer a map of the project before they choose where to dig.

On this page

What it is

The two-minute project pitch is the opening of a project walkthrough: a short, complete account of one piece of work that lets the interviewer pick where to go deeper. The round it opens goes by several names, among them a project deep dive, a retrospective, an experience discussion and a technical presentation. Whatever the name, the pitch has four pieces: the setting and the problem, your part in it (with the people around you), the one decision that shaped the work, and what came of it. Two minutes is enough to say all four if each gets a few sentences, and it matches the length Amazon's interview guide suggests for a behavioral answer, two to three minutes.

PieceWhat it answersRough share

Setting and problem

Where and when it happened, who was around it, what was wrong or wanted, and why that mattered to the business or its users.

About a quarter

Your part

What you owned, and who else was on it with what role. One or two sentences.

About a sixth

The one decision

The choice that shaped the project, the alternative you rejected and the cost you accepted.

About a third

Outcome

How it ended, in one sentence, with a baseline and a result if you measured them.

About a sixth

The shares are our suggestion rather than a rule. The point is the proportion: the setting and the team are there so the decision makes sense, and the decision gets the most time because it is where your judgment shows.

Why a map comes first. You carry months of context on the project; the interviewer has a line or two from your resume. Ashby's published guide to its past-projects interview says the candidate leads while the interviewers stop them with questions along the way, going deep into whatever they find interesting, and it names clear communication as one of the skills being assessed: you are expected to judge where context is needed and where it is not. A pitch that covers all four pieces up front lets the interviewer choose a thread with the whole story in view, instead of asking you to back up and explain what the project was for.

Why the problem carries its stakes. Saying what was wrong is not enough on its own; say what it cost, or who was waiting for it. Employers who run this round ask how customer or business needs shaped the technical decisions, and how the problem was scoped and success defined. A problem stated with its stakes also sets up the outcome: if you open with failed invoices, the closing number is the invoice failure rate.

Why the people are in it. Naming who else was on the project, and what each did, is what makes your own part credible. Amazon's guide asks candidates to use "I" statements when describing collaborative work so that personal contributions and decisions are clear. The precise version names both: the team built the service, a colleague designed the reconciliation check, and you chose the rollout order and took its cost to the lead. Managers do the same with the goal they set, what they delegated and where they stepped in. There is more on this in Saying "I", not "we".

Why it ends on the outcome, then stops. Give the result in one sentence, then hand the floor back and wait. Our advice is to resist filling the silence with a second project or a tour of everything else you did. The pitch has done its work once the interviewer knows what happened and can choose a direction. For how the follow-ups go from there, see A level deeper on your project.

Formats differ, so ask. Ashby describes walking through a project twice: 10 to 20 minutes in an introductory call and 30 or 45 minutes in a final round, with slides and diagrams optional and no credit for using or skipping them. OpenAI's interview guide says its formats vary by team. Use your recruiter's instructions for the time, the audience and the materials; the two-minute opening fits any of them.

There is more on the round in our blog guide.

When to use it

Use it whenever one piece of work is the subject: "walk me through a project", "tell me about the project on your resume", or the first minutes of a dedicated project round. One recruiting-platform guide for developers suggests opening with a brief summary of what the project was and why, whether it was solo or group work and which parts were yours, then letting the follow-up questions dig in. The pitch is that brief summary with the decision and the outcome added.

Use it in preparation too, before you build the longer version. Writing the two-minute opening first forces the choice of which decision and which outcome lead; the longer walkthrough then expands the team, the constraints and the sequence without hiding the result until the end. If you can't fill a piece (you don't remember the baseline, or someone else made the decision), that is a sign to pick a different project or to say plainly what you do and don't know.

When not to

Some prompts look like a project question but want a different shape.

  • "Tell me about a time when..." asks for one event and spends most of its time on your actions. Use the STAR method for that; a project pitch spreads the time across the setting, the team and the outcome instead.
  • A narrow follow-up. When the interviewer asks how the cutover stayed consistent, answer that. Restarting the four pieces spends the time they set aside for depth.
  • An interruption mid-pitch. Finish the sentence, answer the question, and reconnect with one line ("so that was the main call; after it came the first regional rollout") rather than defending the order you had planned.
  • A prepared presentation. If the recruiter asked for a 15-minute talk with slides, the two-minute version is your first slide, not the whole talk.
  • Tell me about yourself. That wants a short tour of your path and why this role, not one project.

The pitch also does less for you when the project was not yours to shape: you joined late, executed a plan someone else made, or left before release. You can still describe the decisions you owned inside it, but say so, and expect the interviewer to weigh it accordingly. It is hard for anyone to judge the outcomes of your decisions on work you did not see through, and at least one employer says so in its own guide to the round.

Compared

PieceThin versionFull versionWhen the thin version is right

Setting

"I worked on onboarding at a SaaS company."

"In 2024 I was one of five engineers on the onboarding team at a B2B SaaS company. New customers were taking about three weeks to send their first live data."

When the interviewer already knows your team from the screen: one clause, then move on.

Problem and stakes

"The flow was clunky."

"Sales said a third of signed deals stalled in setup, so the delay was costing renewals, not just goodwill."

When the stakes are obvious from the setting, such as a live outage; don't belabor them.

Your part and the team

"We rebuilt it."

"I owned the import path. A designer reworked the flow, and a second engineer built the validation UI."

When it was a solo project: say so in a clause and skip the roll call.

The one decision

A list of everything you built, in order.

"I chose to accept messy uploads and clean them server-side instead of rejecting them with errors. That meant a bigger parsing job for us, and fewer stalled customers."

In a ten-minute screen, when asked for the headline: name the choice and let them ask about the alternative.

Outcome

"It went really well."

"Time to first live data fell from about three weeks to four days over the next quarter."

When the project hasn't landed yet: give what you have measured so far and say what you don't know.

Hand back

"...and then the next quarter I moved on to search, which..."

"That's the overview. Where would you like to go deeper?" Then silence.

When the interviewer has already jumped in with a question: answer it, the hand-back has happened.

In a picture

Setting andproblem

Step 1 of 5

Setting and problem

Open with where and when, the team around it, and what was wrong or wanted with its stakes. Two or three sentences.

How to say it in an interview

The common mistake

Quiz

Forty seconds into your pitch, the interviewer asks how your data model handled duplicates. What do you do?
Which closing line gives the interviewer the most to work with?
You were one of six on the project. You led the rollout plan; a colleague designed the architecture. How does the team appear in your pitch?
You have just given the outcome in one sentence. What comes next?

Related

Sources

  1. The Project Deep Dive Interview: How to Walk a Panel Through Your Work (Star Interview blog)
  2. Engineer Past Projects Deep Dive (Ashby)
  3. Your complete guide to the Amazon interview process
  4. How to Walk Through a Project in an Interview (Hatchways)
  5. Interview Guide (OpenAI)
Sign in to report a mistake