How to Prepare for an Interview: A Real Method

You have almost certainly already read a list of fifty interview questions. Possibly several. And you still feel unprepared, which is the correct reaction, because reading a list of questions is not preparation.

The lists fail for a simple reason. An interviewer does not ask “what is your greatest weakness?” in those words. They ask “what’s the part of the job you’d find hardest?” — or they never ask it at all and instead follow up on something you said forty seconds earlier. The memorised answer is keyed to a phrasing that does not arrive, so you either force it in where it does not fit, or you freeze because the thing you rehearsed has nowhere to go.

What survives contact with a real interview is material. Not scripts — material. A small stock of real, specific things you have done, held loosely enough that you can point any of them at whatever question turns up. That is what this article builds, along with the research worth doing, the logistics that quietly wreck good interviews, how video changes things, and what to say when you genuinely do not know.

One honest thing first: you cannot prepare your way out of not being right for a role. Some interviews are decided before you sit down — an internal candidate, a budget that shifted, a requirement that was never in the advert. Preparation does not win those. It wins the ones that are genuinely open, and there are more of those than the rejections make it feel like.

Why does preparing from a list of questions make you worse, not better?

Memorised answers are audible. A candidate reciting a prepared paragraph sounds different from a candidate thinking — the rhythm flattens, the eyes drift up and sideways during retrieval, the sentences are suspiciously well-formed, and the detail is generic because it was written to fit any employer. It does not read as prepared. It reads as evasive.

And the follow-up is where interviews are decided. The first answer makes a claim; the follow-up tests whether the claim is real. Someone describing a project they genuinely ran can go three or four questions deep — who objected, what it cost, what they would do differently — because they were there. Someone reciting cannot go one question deep. You do not need to be articulate to survive that. You need to have been there and to remember the details.

Prepare material you can draw on, not answers you have to retrieve. Material bends to the question. A script only fits the question it was written for.

What is the one thing worth preparing, and how do you build it?

Your stock of stories. Six to eight concrete episodes from your working life, each specific enough to be believable and flexible enough to answer several different questions. If you do nothing else in this article, do this.

Finding them

Do not start from the questions. Start from your own timeline, and write down moments rather than responsibilities — episodes with a beginning and an end. Prompts that reliably surface them:

Something that went wrong and you were involved in fixing it
Something you changed — a process, a document, a way of working that annoyed you enough to alter
Someone you persuaded who did not initially agree with you
A deadline that was going to be missed and was not, or was missed and you handled the fallout
A mistake that was genuinely yours, and what you did next
A time you worked with someone difficult
Something you learned quickly because you had no choice
A piece of work you are quietly proud of, even if nobody noticed

Fewer than six and you will be visibly recycling one. Many more and you will not know them well enough to use under pressure. If you are early in your career, returning after a break, or changing field, they need not come from paid employment — volunteering, study, caring responsibilities, a side project, running something for a community group all contain the same raw material: a problem, a decision, a consequence. If the episodes are older than you would like, use them anyway. Say when it was, plainly, and carry on.

How much detail, and what kind

Specificity is the whole difference between a story and a claim. Not “I improved the reporting process” but “the monthly report took two people a day and a half because three systems disagreed about what a customer was.” Numbers only if you actually have them — do not invent one to sound impressive, because the follow-up will find it.

For each episode, know these well enough to say them without notes:

The situation, in about two sentences. Enough for a stranger to picture it, and what was at stake if it went badly.
The specific problem or decision. Not the general area — the actual fork in the road.
What you did, personally. Say “I” where it was you and “we” where it was the team, scrupulously. Claiming a team’s work is the easiest thing in the world for an interviewer to catch, because they will ask what your part was.
Why you did that rather than the obvious alternative. The part most people skip and the part interviewers most want. Reasoning travels further than outcomes.
What happened, including if it was mixed. A partial success you can discuss honestly beats a triumph nobody believes.
What you would do differently. One real thing, not a humble-brag about caring too much.

You may know this framing as STAR — situation, task, action, result. It is a fair checklist for testing whether a story has a shape. But followed literally it is a main cause of the robotic interview answer: people announce the sections aloud, front-load a minute of scene-setting nobody asked for, and land on a “result” polished into meaninglessness. Use it before the interview to check a story is complete. Do not use it as a template while you are speaking.

Keeping them flexible

The point of six to eight is that each answers several questions. A project that went sideways can serve a question about pressure, about a mistake, about working across teams, about a difficult stakeholder, about prioritising — depending on which part you lead with. So learn each episode as a whole event, not as an answer to one question.

Two questions are worth rehearsing as questions, because they open almost every interview: some version of “tell me about yourself,” and some version of “why do you want this job?” Do not memorise paragraphs. Know where you want to start, the three things you want to have said, and how you will land it — then let the words vary. Each deserves more work than a paragraph here can give.

What research actually helps, and what only sounds researched?

Reciting the founding year, the office count and a line from the values page sounds researched and is worth nothing. Anyone can read a homepage. What is worth knowing is anything that changes what you would say or ask:

What they sell or do, well enough to explain it to a friend. If you cannot, everything you say about contributing will be vague.
Who their customers or service users are, and what those people want. Talking about the work from the user’s side is unusual and lands well.
What has changed recently — a new product, a restructure, a new market, a funding round, a regulatory shift. Recent change is usually why the role exists.
Who is in the room, and what they would care about. Your future manager wants different things from the person there to check you can work across departments.

Two or three real things beat twenty shallow ones. The test is whether a piece of research lets you say something the other candidates cannot.

Reading the job description as evidence

A job description is not a neutral list of duties. It is a document written by people with a problem, and it leaks. Read it a second time asking what they are worried about.

Repetition is the loudest signal. If “stakeholder management” appears in the summary, the responsibilities and the criteria, someone previously struggled with stakeholders and it hurt. “Able to work with minimal supervision” means the last person needed a lot, or the manager has no time. “Comfortable with ambiguity” means the role is undefined and they know it. “Fast-paced” often means under-resourced, and it is fair to ask what that looks like day to day.

Unusual specificity is the second signal. Most of the document is boilerplate; the one oddly precise requirement — a named system, a particular report — is there because a real person put it there for a real reason. What is missing matters too: a senior role with no mention of budget or line management tells you where the authority sits.

How do you work out what the role is actually for?

Almost every open role exists for one of a few reasons, and identifying which puts you ahead of nearly everyone in the process. Nobody teaches this.

Something is broken. A backlog out of control, a system only one departing person understands. They want the bleeding stopped.
Something is growing. The work is fine; there is too much of it. They want throughput and someone who needs little shaping.
Someone left and the work is uncovered. The most common and the most boring. They want continuity and low risk.
They are building something new. They want appetite for ambiguity and evidence you have started things before.
Someone senior wants a capability the team lacks entirely. They want proof you have really done it, not adjacent to it.

You can usually tell from the wording, how urgent the hiring feels, whether the role is new or a replacement, and what has recently changed. If you cannot, ask early — “can I ask what prompted the role being opened?” — then adjust for the rest of the conversation.

Once you know, your material sorts itself. Broken means you lead with the episode where you stabilised something. Growing means you lead with volume and pace. New build means you lead with the thing you started from nothing. Same stock, different order — and the interviewer experiences someone who understands their situation, which is exactly what it is.

Why does rehearsing in your head not work?

Because thinking a story is a different task from saying it. In your head it is coherent and takes four seconds. Out loud it takes two and a half minutes, and somewhere in the middle you find you cannot explain the bit that made sense internally, that you have talked for ninety seconds without reaching the point, and that a sentence you were sure of collapses halfway through. Speaking uses different machinery from thinking, and only the speaking machinery gets used in the interview.

So say them out loud, alone in a room. Aim for two or three minutes each. Listen for where you stall, where you over-explain the setup, where you cannot summarise the outcome. Do each two or three times, not twenty — you want the shape familiar, not the wording fixed. If it starts sounding like a recital, deliberately tell it differently next time.

Recording yourself is unpleasant and useful: you will hear filler, the point arriving late, and whether you answered. If someone will run a mock interview, ask them to add follow-ups and to phrase a question badly now and then, because real interviews do.

Which logistics quietly ruin interviews?

Interviews are lost by things unrelated to your ability, all preventable in twenty minutes.

The route. Know the actual door, not the postcode. Business parks, hospitals and large sites have entrances nowhere near the address.
The time. Be in the area twenty minutes early; go in five to ten minutes before. Arriving forty minutes early burdens whoever has to deal with you.
Time zones. If they are in a different one, confirm in writing in both.
The link. Open it the day before. Some platforms need an install, an account, or a browser you do not have.
Names. Know who you are meeting and how to say their names. If unsure, ask at the start rather than avoiding the name for an hour.
What to bring. CV copies, notebook and pen, your written questions, anything they asked for, water. Notes are fine; reading from them is not.
A phone number. Being late is survivable. Being late in silence is not.
Access needs. If you need an adjustment — a break, materials in advance, a quiet room — ask beforehand. A good employer handles it without comment.

What is different about a video interview?

Most advice still assumes a room. The conversational skills transfer; the physical setup does not, and it does more work than people realise, because on video the frame is all the interviewer has.

Light matters most. It should come from in front of you — a window you face, or a lamp behind your screen. A window behind you makes you a silhouette for an hour. If that is your only window, turn the desk around.

Camera height. Roughly eye level, which for a laptop means a stack of books. Below eye level is unflattering and makes you look like you are leaning away. Frame from mid-chest up with a little space above your head, so your hands show when you gesture.

Eyeline. You cannot watch their face and the lens at once. Look at their face while they speak; look at the camera for the first seconds of your answers and when you make a point that matters. Dragging the video window to the top of the screen narrows the gap. Do not stare into the lens throughout.

Background. Tidy and uninteresting — a wall, a shelf, a plant. Virtual backgrounds smear the edges of your head when you move. If you are interviewing from a shared house or a room that is also a bedroom, do not apologise for it. Nobody sensible is judging your housing.

Sound beats picture if you must choose. Wired earphones with a microphone usually beat laptop speakers and avoid echo. Test inside the actual platform, not just system settings.

Interruptions. Silence phone and desktop notifications, close the mail client, tell whoever else is in the building. If a child or a pet appears, deal with it briefly and move on.

When the connection fails — assume it might, and agree the fallback at the start: “if we drop, shall I rejoin, or would a phone call be easier?” That one sentence removes the panic. If your video degrades, say so and switch the camera off rather than freezing mid-sentence. If you lose them, rejoin once, then ring the number you made sure you had. Handling a technical failure calmly is a live demonstration of composure.

How should you handle a panel, or a second-round task?

Panels feel harder than they are. Get the names at the start and write them in seating order. Answer to the person who asked, then bring your eyes across the others — you are talking to the panel, not holding a private conversation. If someone stays quiet or expressionless, read nothing into it; they are probably scoring against a form.

Panels usually run a structured script with each person owning a section, which is why questions can feel abrupt or repetitive. If you are asked something close to what you already covered, say “I touched on this earlier, but the part I’d add is —” and go somewhere new.

Task stages test different things: a presentation, a written exercise, a technical problem, a half-day with the team. The common errors are answering a different question from the one set, and going far beyond the brief — a task asking for fifteen minutes that gets forty-five signals you cannot scope work. Ask what format, who will be present, and how long. If the task needs assumptions, state them aloud; handling incomplete information is usually part of what is assessed.

What do you do about nerves?

Not “relax” — that has never worked for anyone. Nerves before an interview are appropriate. Some visible nervousness reads as caring, and interviewers tolerate it far better than candidates believe; almost nobody is rejected for a shaky voice in the first two minutes. What helps is practical:

Slow your breathing out — a longer exhale than inhale, for a minute or two, before you go in.
Eat something. Nerves plus low blood sugar is shakiness, and it is avoidable.
Walk beforehand. It burns off adrenaline and stops you rereading notes for the twelfth time.
Do not cram in the last hour. New material does not stick and it crowds out what you knew.
Use the water. A pause to drink is three legitimate seconds of thinking time.
Let the first answer be imperfect. It settles almost everyone.

If your nerves are disabling rather than uncomfortable — panic symptoms, going blank entirely — that is a different problem from ordinary interview anxiety, and worth getting proper support for rather than gritting through repeatedly.

What do you say when you do not know the answer?

This is the most useful interview skill there is and almost nobody writes about it. You will be asked something you cannot answer. Everyone is. What separates candidates is entirely the next fifteen seconds.

Bluffing is generating plausible sentences around a gap; anyone who knows the subject spots it, and it converts “does not know this one thing” into “will pretend to know things,” which is disqualifying in a way ignorance is not. Collapsing — apologising, going quiet — wastes a question you could have scored on. The good option has three parts.

Say it plainly, once. “I haven’t come across that.” No apology tour. Two seconds, and it buys credibility for everything else you say.
Then reason out loud. “I haven’t done it, but I’d start by finding out X, because what would worry me is Y — and I’d check with whoever owns Z before committing.” You have just shown how you approach unfamiliar problems, which predicts your performance far better than one fact.
Say how you would close the gap. Where you would look, who you would ask, how long it would take. If you have learned something comparable before, that is the moment for the relevant episode from your stock.

The same applies to gaps in your history — a redundancy, a health period, caring for someone, a stretch of not finding work. Say what it was in one plain sentence, without dressing it up, then move to what you did next. Long explanations signal shame and invite digging; short, level answers close the subject. Most interviewers care far less than you fear, and the ones who make a meal of it have told you something useful about the employer.

It is also fine to ask for a moment. Three seconds of silence feels enormous to you and negligible to them.

What should you ask them, and what happens after?

Prepare your own questions from what you actually want to know — what would make you accept or refuse, what you would need to succeed in the first six months, what the last person found hard — rather than from a list designed to impress, which is transparent and wastes your only chance to gather real information; there is more worth saying about the questions you ask at the end and what the answers mean.

Afterwards, write down what you were asked and what you wish you had said, while it is fresh, then send a short note — there is a right way to follow up after an interview. And if it goes well, there is a separate skill in negotiating the offer before you accept it.

What does the preparation actually look like, day by day?

Assuming a week. Compress it if you have two days; the order matters more than the timescale.

As soon as it is booked: confirm time, format, duration and who you are meeting. Check the platform or the address. Ask about any task and get the brief in writing.
Day one: reread the job description twice — once for what the job is, once for what they are worried about. Write down the three or four things they most need, and what you think the role is for.
Day two: build the stock. Six to eight episodes as rough notes, not scripts. Check them against what they need; fill the gaps.
Day three: research. Two or three substantive things you could not have got from the homepage. Look up the people you will meet.
Day four: say every episode out loud, twice. Record one. Cut the rambling.
Day five: write six questions of your own, expecting to ask three or four. Reread the CV and cover letter you sent, which you may have written weeks ago.
Day before: test link, camera, microphone and light, or check the route and the door. Lay out clothes, print CVs, save the phone number. Read the episode list once, then stop and do something else entirely.
The hour before: eat, walk, breathe slowly. Skim your questions, not your answers. Arrive twenty minutes spare, go in five to ten minutes before, be pleasant to whoever is on reception, phone silenced and away.

One question opens almost every interview and deserves separate work: how to answer “tell me about yourself” without reciting your CV. And once the job is yours, the first ninety days have their own trap.

Two questions open almost every interview and each rewards separate work: why do you want this job is a test of comprehension, not enthusiasm.

If the interview goes well, the leverage moment arrives immediately after it — an offer is the one point where you can still change the terms.

The official job-search guidance covers the surrounding process without trying to sell you a course.

Prospects keeps a large interview library that is free and sells nothing.

Frequently asked questions

How long should my answers be?

Most behavioural answers land at ninety seconds to two and a half minutes. Under thirty seconds feels thin; over three minutes and you can usually watch attention going. If unsure, finish the point and ask whether they want more detail — that is a normal thing to do and it hands the pacing to them.

Is it acceptable to take notes in?

Yes. Bring a notebook with your episodes in brief, your questions, and a page for notes. What does not work is reading answers off the page, or losing the thread because you are transcribing. Note-taking is more visible on video, so say once “I’m just making a note” and the pauses stop looking like distraction.

What if I have no relevant experience for something they ask about?

Say so, then reason from the nearest thing you have. “I haven’t managed a team, but I’ve coordinated three people across departments with no authority over any of them, and what I found was —” That is a genuine answer and far stronger than pretending or shutting down. Interviewers hiring for potential listen for transferable judgement, not a checklist match.

Should I ask about salary in a first interview?

You can ask about the range if it was not advertised, and it saves you both time if the gap is unbridgeable. Phrase it as a factual check — “can I check the range for this role?” The negotiating comes later, after an offer, when you have leverage you do not have yet.

How do I prepare when I am interviewing for several roles at once?

Build the story stock once; it is the same stock every time. Then do only the role-specific work: reread that description, work out what that role is for, choose which episodes to lead with, write questions for that employer. That is an hour or two per interview, not a week. Keep one document per application so a second round three weeks later does not start from zero.

Does preparation matter if the decision seems already made?

Sometimes it is, and nothing changes that. But you rarely know from the inside which kind of interview you are in. The cost of preparing for one that was never open is a few hours; the cost of not preparing for one that was open is the job. Prepare, accept that some outcomes were never yours to influence, and do not read a rejection as a verdict on your work.

How to Prepare for an Interview Without Memorising Answers