Build a story bank of seven
You do not need an answer for every question. You need seven stories that can be pointed in several directions.
There are hundreds of behavioral questions and perhaps a dozen things they are trying to find out. Preparing per question is impossible and produces brittle, memorised answers. Preparing seven strong stories and learning to angle them covers almost any panel you will sit in.
The seven
- Something you delivered that mattered — with a number attached, and a clear account of your part in it.
- A time you disagreed with someone more senior, and what happened afterwards.
- A failure that was genuinely yours, including what you now do differently.
- A time you led without authority — you had no reporting line and it happened anyway.
- A hard technical or analytical problem, told so a non-specialist could follow it.
- A time you changed your mind because of evidence, or someone changed it for you.
- A time you had to say no, or cut scope, or disappoint someone on purpose.
Angle, do not recite
One story serves several questions depending on which beat you emphasise. The launch story is a delivery story if you emphasise the outcome, a prioritisation story if you emphasise the triage, a leadership story if you emphasise how you got two engineers reassigned, and a conflict story if you emphasise the argument with the product manager. Prepare the material; choose the angle in the room.
Write each one as five lines
- One line of context — where, when, what was at stake.
- Three lines of what you did, phrased as decisions rather than activities.
- One line of result, with a number or a concrete change.
- Then, separately, the number itself: know how it was measured and what your contribution to it was, because you will be asked.
Range matters
Check that your seven do not all come from one job, one year, or one type of problem. A panel that hears four stories from the same six-month project starts to wonder what the other three years looked like. Spread them across employers and across kinds of difficulty — technical, interpersonal, organisational.
Say them out loud once
Not to memorise — to find the parts that do not survive being spoken. Stories that read well on paper often contain a chunk of context that is unnecessary out loud, or a result you cannot actually justify when you hear yourself claim it. Ten minutes of speaking finds both.
Seven prepared stories, angled well, beat forty memorised answers — and they survive the question you did not anticipate.