Preparation, not performance

Interviews are a conversation you can prepare for properly

Not tricks, not scripts, and not a list of "power words". This is how to prepare so that you are the same person under pressure that you are on a good day — and how to handle the parts nobody enjoys: failure questions, gaps, salary and rejection.

An interview is two people trying to answer two questions. Theirs: can this person do the job, and will working with them be good? Yours: is this a job I want, with people I want to work with? Preparation that only serves the first question is half a preparation, and it is why so many interviews feel like an interrogation.

If you only have an evening. Do three things: read the company's own words for an hour (section 1), write out six stories from your CV (section 2), and prepare five questions you actually want answered (section 6). That is most of the value of a week of preparation.

1. Researching a company properly

"Research the company" usually produces someone who has memorised the About page and can recite a founding year. That is not research; it is trivia, and it is obvious. Real preparation gives you three things: a mental model of the business, informed questions, and the ability to make your own experience relevant in the moment.

Where to look, in order

  1. The product itself. Use it. Sign up, take the trial, walk into the shop, read the app store reviews. Nothing else in your preparation will be as useful as twenty minutes of first-hand contact with what they sell.
  2. Their own words. The blog, changelog, engineering or design writing, annual report if they publish one, the careers page as a whole. What they choose to write about is what they care about — and how they write reveals the culture more reliably than any values statement.
  3. The job description, read as evidence. Roles exist because something hurts. A new "first data hire" means decisions are being made blind. A posting that mentions documentation three times has a knowledge-loss problem. Ask yourself what pain created this opening, then check your theory in the interview.
  4. The people. Who will interview you, what did they do before, how long have they been there. Not to flatter them — to understand what they will care about. A founder, a future peer and an HR screener are three different conversations.
  5. The market. Who else does this, and what makes these people different? Recent news, funding, launches, a merger, a regulatory change in their sector.

What to do with what you find

Turn it into three artefacts, on one page:

  • One sentence on how they make money. If you cannot write it, you do not understand the business yet, and it will show.
  • Two observations — something you noticed using the product, or a tension you spotted between what they say and what they ship. Phrase them as curiosity, not critique: "I noticed onboarding asks for a lot up front — is that deliberate?"
  • Three connections — specific moments in your own history that map onto their situation.
The trap of over-research. Reciting their Series B or their founder's previous company can read as sycophantic and slightly unnerving. Research is for making you a better participant in the conversation, not for demonstrating that you did research. Use it when it is relevant, and let the rest sit quietly in the background.

2. Turning your CV into stories

Nearly every interview question is a request for a story, and nearly every weak answer is an opinion offered instead of one. "I'm good at handling difficult stakeholders" is an opinion. "Here's what happened with the finance director in the second quarter" is evidence. Only one of them is memorable.

The six-story kit

Do not prepare an answer for every possible question — there are hundreds, and you will sound rehearsed. Prepare a small library of real stories and learn to angle them. Six is usually enough:

  1. A thing you built or ran end to end — your strongest single piece of work.
  2. A conflict — with a colleague, a stakeholder, a client — that you handled.
  3. A failure — real, owned, with what changed afterwards.
  4. Something you learned quickly under pressure, from nothing.
  5. A judgement call where you had to decide with incomplete information.
  6. Something you improved that nobody asked you to improve.

Pull them straight from the bullets on your CV. That is what the CV is for at this stage: the interviewer is holding it, and every line on it is a question they might ask. A useful exercise the night before is to read your own CV line by line and ask "if they point at this, what is the ninety-second version?" Anything you cannot answer confidently should either be rehearsed or should not have been on there.

STAR, and where it lets people down

The standard structure is Situation, Task, Action, Result. It is a good scaffold: it stops you starting in the middle and forgetting to say how it ended. Use it as scaffolding — then take the scaffolding down. Its failure modes are consistent:

Failure modeWhat it sounds likeFix
Front-loaded setup Ninety seconds of org chart before anything happens Two sentences of context, maximum. Get to the action.
"We" throughout A team story with no discernible you in it Say what you did, and credit the team separately
Audibly rehearsed Fluent, unpausing, and clearly the fortieth delivery Learn the beats, not the sentences
Missing result Trails off after the action; nobody learns how it ended Always land it — including when it landed badly
No reflection Tidy story, no evidence of thought Add one line: what you would do differently now
Mechanical labelling "So, the situation was… the task was… the action was…" Just tell it like a story. They know the structure.
Rehearsed STAR

So the situation was that I was working at a mid-sized firm in the operations department, which had about forty people across two sites, and my task was to improve the reporting process. The action I took was to implement a new system, and the result was that efficiency improved significantly.

Correct structure, no content. Nothing here could not have been said by someone who was not there.

Told as a story

Every Monday, three of us spent the morning rebuilding the same spreadsheet by hand, and it was wrong often enough that the ops lead had stopped trusting it.

I spent a week sitting with the people who used the numbers rather than the people who produced them, which changed what I built — they needed two views, not fourteen. I wrote it against the warehouse database directly so nobody had to key anything in.

It went out on a Thursday, and the following Monday nobody rebuilt anything. What I'd do differently: I shipped it without a way for them to change the date range, and spent a month fielding requests I could have avoided with one dropdown.

Same structure underneath. Specific, human, ends with reflection that is honest without being self-flagellating.

Length and pacing

Aim for ninety seconds to two minutes for a full story. Under thirty seconds and the interviewer has to mine you for detail; over three minutes and they are managing you rather than listening. Finish, then stop. Silence after a good answer is the interviewer thinking, not an invitation to fill it — the most common way people damage a strong answer is by continuing past the end of it.

3. The questions you will almost certainly get

Do not script these. Understand what is actually being asked, and have material ready.

"Tell me about yourself"

Not an autobiography. It is asking: how do you frame your own career, and can you be concise? A two-minute shape that works: where you are now and what you do, two or three moves that explain how you got here and what you got good at along the way, then why this role is the sensible next step. Present, past, future — ending on them.

Do not start at university unless you left it recently. Do not narrate every job. And do not answer with a personality summary ("I'm a really curious person") — they asked about your career.

"Why do you want this job?"

The honest answer is usually a mix of interest, money and timing, and they know it. What they are testing is whether you have thought about this job specifically. Give two reasons: one about the work (a real detail — the product, the problem, the stage the company is at) and one about fit (what you would bring, and what you would learn). A reason that would apply to any company in the sector is not an answer.

"Why are you leaving?"

Be truthful and brief, and do not campaign against your current employer. Even a justified account of a genuinely bad workplace makes the listener wonder how you will describe them one day. Frame it forwards: what you want next that you cannot get where you are. If you were made redundant, say so plainly — it is common, and stating it directly is far better than a vague sentence that sounds like concealment.

"What's your greatest weakness?"

The question is a bad one, but it is asked constantly, and the tell is not the weakness — it is whether you have any self-knowledge. Fake answers ("I'm a perfectionist", "I care too much") fail immediately and are irritating. So do genuinely disqualifying ones ("I'm not reliable with deadlines").

Pick something real, adjacent rather than central to the job, and say what you actively do about it. "I under-communicate when I'm deep in something — I've started writing a Friday summary whether or not anyone's asked for one, because people were finding out about my work too late." That is a real weakness, a real mitigation, and it ends on evidence.

"Where do you see yourself in five years?"

Nobody knows, and pretending certainty is worse than admitting the truth. What they want to know is whether your direction is compatible with the role. Answer in terms of the kind of work and the kind of responsibility you want to be growing into, not job titles. If your honest plan is incompatible with the role — you want to start your own business in two years, and this is a five-year leadership track — that is worth surfacing rather than discovering later.

"Do you have any questions for us?"

Always yes. See section 6 — it matters more than most candidates think.

The unasked question underneath all of them. Most interviewers are, consciously or not, asking one thing: would I want to work with this person on a difficult day? Answers that are specific, honest, appropriately brief and generous about colleagues answer that question. Answers that are polished, evasive or subtly contemptuous of previous employers answer it too — in the wrong direction.

4. Talking about failure, gaps and things that went wrong

These questions are where interviews are actually decided, because they are the only ones where the polished answer is obviously worse than the honest one.

The failure question

"Tell me about a time you failed." The trap is the near-miss disguised as a failure: the project that was late but ultimately triumphed, the disagreement you turned out to be right about. Interviewers hear these constantly, and what they learn is that you will not admit to anything — which is a serious mark against you if the job involves any risk at all.

A good failure answer has four parts: what you were trying to do, what went wrong and your part in it, what it cost, and what you changed as a result. Own your share squarely, without either minimising it or performing excessive contrition. Choose something real but not catastrophic, and never something involving dishonesty.

Non-answer

I'd say my biggest failure was taking on too much because I wanted the project to succeed — I ended up working weekends, but we delivered on time and the client was delighted.

A humblebrag with a failure-shaped opening. It tells the interviewer you will not be candid when something goes wrong.

Real answer

I pushed a supplier switch through on a timeline I'd set myself, and I didn't check how the returns process worked at the new one. About two hundred customers waited three weeks for refunds and support absorbed it.

My part was straightforward: I was excited about the cost saving and I treated the operational detail as somebody else's homework. Now I write down what I'm assuming before a change like that and get one person from each affected team to argue with the list.

Real cost, clear ownership, a concrete practice that came out of it. This answer builds trust rather than spending it.

Gaps in your history

If your CV has a gap, expect it to come up, and prepare a short answer rather than improvising. Short is the operative word — two or three sentences, delivered without discomfort, then move on to something you did or learned.

  • Redundancy: "The team was cut in a restructure in March — twelve of us went. I took two months to look properly rather than take the first thing."
  • Health: you are not obliged to disclose a diagnosis anywhere. "I took time out for health reasons; that's fully resolved and I've been ready to work since spring" is a complete answer, and in many places questions that press further are legally fraught for the employer.
  • Caring or parental leave: state it plainly. It is ordinary and it does not require justification.
  • A job that did not work out: "It wasn't the role that was described to me, and rather than settle in for two unhappy years I left." Do not litigate it.
  • A long search: honest, plus what you did with the time — a course, freelance work, an open-source contribution, anything.

The general principle: the gap is almost never the problem. Visible discomfort about the gap is the problem, because it signals there is more to it. Rehearse the sentence until it is boring to say.

Before the interview, look at your own timeline the way a stranger would. Tessira's AI review flags employment gaps of more than a month between roles, which is a slightly uncomfortable but useful way to find out what a reader will notice before they notice it. The point is not to hide anything — it is to not be surprised.

Being fired

Say it in one sentence, without drama and without blaming: what happened, what your part in it was, and what you took from it. "It wasn't working — my manager wanted someone much more hands-off than I was being, and I didn't adapt fast enough when they told me. I ask about working style in the first month now." Interviewers respect the candour far more than they penalise the event; what they cannot get past is a story that does not add up.

5. Technical stages and take-homes

The advice differs by discipline, but the principles hold across engineering tests, design exercises, case studies, financial models and portfolio reviews.

Live problem-solving

  • Think out loud. They are assessing your reasoning, and a silent candidate who arrives at the right answer often scores worse than one who narrates a wrong turn and corrects it.
  • Clarify before you start. Restate the problem, ask about constraints and edge cases. Jumping straight into a solution is the most common mistake, and in many rubrics the clarifying questions are themselves being scored.
  • Say when you are stuck, and say what you would do about it in real life — look it up, read the docs, ask a colleague. That is the honest answer and it is what the job looks like.
  • Take the hint. If the interviewer nudges you, they are helping, not testing your resolve. Ignoring a hint to prove independence goes badly.
  • Do not bluff. "I haven't used that — here's how I'd approach finding out" is a fine answer. An invented one is discovered within two follow-up questions.

Take-home exercises

  • Respect the stated time box, and say what you did with it. If it says four hours, spend four hours, then write a short note listing what you would do with another day. That note is often read more carefully than the work itself.
  • Include a README or a covering page — the assumptions you made, the trade-offs you chose, how to run or read it. Explaining your reasoning is half the exercise.
  • Do the brief that was set, not the more interesting one you would rather do. Scope discipline is being assessed.
  • Prepare to defend it. Expect the follow-up conversation to poke at your choices. Know why you did each thing, including the parts you dislike.
  • Know when to decline. A reasonable exercise is a few hours of invented work. A multi-day project that closely resembles their actual backlog is not an assessment, and it is legitimate to ask to substitute a portfolio walkthrough or a paid engagement.

Portfolio and case walkthroughs

Show fewer pieces, in more depth. Three projects presented as decisions — what the constraint was, what you tried, what you rejected and why, what it cost — beat twelve screenshots. Be explicit about which parts were yours when the work was collaborative; interviewers ask, and vagueness there is expensive.

6. The questions you ask them

"Any questions for us?" is not a formality, and "no, I think you've covered everything" is the single most avoidable mistake in interviewing. It is your best opportunity to learn whether you want the job, and the interviewer is reading your questions as a signal of what you care about.

Prepare eight, expect to ask four or five, and cross off any that got answered. Good questions are specific, answerable, and about things you genuinely want to know.

About the role

  • What does a good first ninety days look like for whoever takes this?
  • Why is the role open — is it new, or a replacement?
  • What is the thing you are hoping this person fixes or takes off your plate?
  • What would make you say, a year from now, that this hire went really well?

About how the work happens

  • How do decisions get made here when people disagree?
  • What does a normal week look like — meetings, focus time, on-call?
  • How does work reach this team, and who decides what is a priority?
  • What is the last thing the team changed about how it works, and why?

About the people and the reality

  • What do you find frustrating about working here? (Ask a future peer, not the hiring manager. A thoughtful, specific answer is a very good sign; "honestly, nothing" is not.)
  • Who has done well in this team, and what did they do differently?
  • How long have the people I'd be working with been here?
  • What is something you tried in the last year that did not work?

Always ask, at the end

  • What are the next steps and the timeline? It is practical, it is expected, and it tells you when following up is appropriate rather than pushy.
  • Is there anything about my background you would like me to say more about? This gives you one chance to address a doubt you would otherwise never hear about.
Not yet. Leave salary, holiday, remote-working negotiation and promotion timelines out of a first conversation unless the interviewer raises them — with one exception: if a hard constraint of yours (a visa requirement, a location, a start date) would make the job impossible, surface it early. Wasting your own week is worse than an awkward five minutes.

7. The salary conversation

This is a negotiation between people who both want an agreement, not a duel. Two things make it far easier: preparing numbers in advance, and being comfortable with a short silence.

Before you talk about money

  • Know three numbers: the figure you would be pleased with, the figure you would accept, and the figure below which you would decline. Write them down before you are asked. Deciding under pressure is how people accept things they regret.
  • Research the range in your market, for your level, in that location, for that size of company. Published ranges in postings, industry salary surveys and candid conversations with people in the field are all better than a guess.
  • Count the whole package: pension contribution, bonus and how reliably it pays, equity and what it is realistically worth, holiday, healthcare, learning budget, hours, commute, remote flexibility. A lower salary with six extra days of leave and no commute may genuinely be the better offer.

When they ask first

Being asked "what are you looking for?" early is common. Reasonable responses:

  • Deflect once, politely: "I'd rather understand the role a bit better first — do you have a budgeted range for the position?" Often they will simply tell you.
  • Give a range if pressed, and make your target the bottom of it: "I'm looking in the region of X to Y, depending on the overall package." Do not name a range whose bottom you would not accept, because that is the number you will be offered.
  • Note on salary history: some jurisdictions restrict or prohibit employers asking what you currently earn. Where it is asked, you are usually free to redirect to your expectations for this role rather than your history.

When the offer comes

Say thank you, express real enthusiasm, and ask for the full details in writing. Then take a little time — a day or two is normal and nobody sensible objects. If you want to negotiate, do it once, specifically, and with a reason: "I'm really pleased with this. Based on the range for this level and the [specific thing you bring], I was hoping for X — is there room?" Then stop talking and let them answer.

If the salary genuinely cannot move, ask what can: a start date, a review at six months with defined criteria, a learning budget, an extra few days of leave, a title. And if you accept, accept warmly and stop negotiating — the relationship you build here is the one you start the job with.

Never accept on the call. Not because you should play hard to get, but because enthusiasm is a bad state for reading a contract in. "This is great news — can you send the details over? I'll come back to you by Thursday" costs you nothing and has never lost anyone a genuine offer.

8. Video and remote interviews

Everything above still applies; video simply removes a lot of the signal you would normally rely on, so a few mechanical things matter more than they should.

The setup, done once

  • Light in front of you, not behind. A window facing you beats any lamp; a window behind you turns you into a silhouette.
  • Camera at eye level. A laptop on a stack of books fixes the up-the-nose angle that makes everyone look tired.
  • Wired headphones with a microphone if you have them. Audio quality affects how competent you sound, unfairly but reliably.
  • Test the actual platform beforehand — the one in the invitation, not the one you usually use — including screen sharing if you might need it.
  • A plain background and a closed door. Blur is fine. Mute notifications everywhere, including your phone.

Performing on video

  • Look at the camera for the important sentences, at the faces the rest of the time. Constant camera-staring is unsettling; never looking up reads as evasive.
  • Leave a beat before answering. Latency makes people talk over each other, and a half-second pause fixes almost all of it.
  • Be more expressive than feels natural. The camera flattens everything, including enthusiasm.
  • Notes are allowed and invisible — but keep them to bullet prompts beside the camera. Reading from a script is audible.
  • Have a fallback. Know the interviewer's number or email, so a dropped connection becomes a thirty-second interruption rather than a crisis. If it happens, do not apologise for three minutes; rejoin and carry on.

If the job itself is remote

Expect, and prepare for, questions about how you work without supervision: how you communicate progress, how you handle being blocked when nobody is in the room, how you separate work from home. Ask your own, too — where the team is, how much overlap in time zones is expected, whether decisions get made in writing or in calls you might not be in, and how often people are expected to travel.

9. Following up

A short note within a day is worth sending: it is polite, it keeps you in mind, and it is one more chance to say a useful thing. It will not rescue a bad interview, and it should not try to.

Three or four sentences. Thank them for their time; refer to one specific thing from the conversation so it is obviously not a template; add one short, genuinely useful thing — a link to the work you mentioned, or a better version of an answer you fumbled; confirm you are interested. Then stop.

Weak follow-up

Dear Sir/Madam, thank you for the opportunity to interview for the position. I am very excited about the role and believe I would be a great asset to your team. Please let me know if you require any further information. I look forward to hearing from you.

Sent by everybody, remembered by nobody. It contains no evidence that the interview happened.

Better

Hi Priya — thanks for this morning, I enjoyed it. The bit about the reconciliation backlog stayed with me: I dug out the note I mentioned on how we phased ours, in case it is useful (attached). It is a genuinely interesting problem and I'd like to work on it. Happy to answer anything else in the meantime.

Specific, brief, and adds something. It reads like the same person who was in the room.

If the date they gave you passes, one polite check-in is entirely reasonable: reference the timeline they mentioned, restate your interest in one line, ask if there is an update. One. Silence after that is an answer, however unsatisfying; hiring processes stall for reasons that usually have nothing to do with you, and chasing weekly converts a maybe into a no.

10. Rejection, and what to do with it

Most applications end in rejection, including from people who go on to have excellent careers. It is worth saying plainly because the volume of it is genuinely demoralising, and people conclude things about themselves from a process that is far noisier than it looks: internal candidates, a hiring freeze that landed mid-process, a role rescoped after interviews began, someone who did the same thing at a competitor last year.

Ask for feedback — carefully

Ask once, make it easy to answer, and lower the stakes: "Thanks for letting me know. If you have a moment, is there one thing I could have done better? No obligation at all." Many organisations will not give it, for legal reasons rather than unkindness. When you do get it, take it at face value and do not argue — replying to feedback with a rebuttal closes a door that was still slightly open.

Do your own debrief

Within a day, while it is fresh, write down: which questions you fumbled, what you wish you had asked, what you learned about the company, and one thing to change next time. Keep this in one file across a whole search. Patterns become obvious quickly — the same story landing badly, a consistent stumble on the salary question, an over-long answer to "tell me about yourself" — and each one is fixable.

Keep the relationship

A warm reply to a rejection is a low-cost, occasionally high-return act: "Sorry to hear it, but I appreciated the conversation — please keep me in mind if something else opens up." Candidates who were a close second are frequently the first call when the next role appears, or when the person they hired does not stay.

And protect the process

Interviewing while working is exhausting, and fatigue shows up as flatness in interviews, which produces more rejection. Batch your applications, cap the number of live processes, take a whole day off from it every week, and do not schedule three final rounds in one week. This is a stamina problem as much as a skill problem.

11. The day-before checklist

  • Re-read the job description and your own CV, line by line — they will be looking at both.
  • Six stories, out loud, once each. Beats, not sentences.
  • Your one-line answer for anything awkward: a gap, a short role, a career change.
  • Five questions written down, with two you would be sorry not to have asked.
  • Your three salary numbers, written down, decided in advance.
  • Names and roles of everyone you are meeting.
  • Logistics confirmed: the address and how long it takes, or the link, the platform and the phone number if it fails.
  • Kit tested — camera, microphone, screen share — on the actual platform.
  • Clothes decided, one notch smarter than what you saw people wearing.
  • Notes reduced to one page of prompts you could glance at, not read.
  • An early night. It outperforms another hour of preparation, every time.

12. Questions people actually ask

How much should I rehearse?
Enough that you know the beats of your stories and can start them confidently; not so much that you deliver them word-perfect. Rehearsed-sounding answers cost you the thing the interviewer is looking for — a sense of what you are actually like. Practising out loud twice is usually the sweet spot; twenty times is not.
Is it acceptable to take notes, or to use them?
Yes to both, and it reads as prepared rather than rude. Ask at the start — "do you mind if I take a few notes?" — and keep your prompts to bullets. On video, put them beside the camera so your eyes do not travel.
What if I completely blank on a question?
Say so and buy time honestly: "Give me a second on that one." Silence for five seconds is fine. If nothing comes, ask to return to it at the end — and then actually return to it. Interviewers remember the recovery more than the blank.
Should I mention other interviews I have going on?
Only if it is true and relevant — most usefully when a timeline is involved: "I should mention I'm at final stage elsewhere, so I wanted to be transparent about timing." Do not invent a competing offer as leverage. It is discovered surprisingly often, and it ends things.
How do I answer a question about something I have not done?
Say you have not done it, then bridge to the closest true thing and how you would approach learning it. "I haven't run a migration at that scale — the largest I've done was about a tenth of it, and here's what I'd want to check first." Honest plus adjacent evidence beats a bluff every time.
Is it worth asking about a company's culture?
Asking "what's the culture like?" invites a marketing answer. Ask about behaviour instead: how disagreements get resolved, what happened the last time a deadline slipped, how often the team changed something about the way it works. You learn the culture from the answers to those.
Can Tessira practise interviews with me?
No — there is no interview simulator, and we would rather say so than imply otherwise. What Tessira does is upstream of the interview: it stores your CVs, renders them as real A4 PDFs, and can run an AI review of the writing or match a CV against a job description you have saved, when you ask it to. The stories are yours to prepare.
What does the job match actually tell me before an interview?
You save the job description, pick a CV, and Tessira compares them — returning what it found clear evidence for and what it could not find. The "missing" side is the useful part before an interview: it is a decent proxy for the questions you are most likely to be pressed on. Like the review, it only runs when you click it, and it is analysis by DeepSeek, an external AI provider, on an anonymised copy of your CV.

Before the interview comes the CV. If yours is doing the bare minimum, our CV writing guide covers the part that decides whether you get to have this conversation at all.

The week before

A preparation plan that fits around a job

Three evenings, in this order. It is more effective than one panicked night, and it is roughly the minimum that shows.

1

Understand the business

Use the product. Read what they write. Work out how they make money and what pain created this role, then write it in one sentence you could say out loud.

2

Build the six-story kit

Pull six real stories out of your own CV — a build, a conflict, a failure, a fast learn, a judgement call, an unasked-for improvement. Say each one out loud once.

3

Prepare your side of it

Five questions you want answered, your three salary numbers, your one-line answer for anything awkward in your history, and the kit tested on the actual platform.

Walk in with a CV you would be happy to be questioned on

Every line of your CV is a question waiting to be asked. Tessira helps you write ones you can answer — designed templates, real A4 PDFs, and an optional AI review. Free to use.