Let's say your resume is three pages of everything you were ever near, and it has won two interviews in four months. Cut it to one page, about the job you want.

The life story gets skimmed and dropped. In Ladders' eye-tracking study the first screen ran about seven seconds, and page two only gets read if page one grabbed somebody.

Laszlo Bock, who ran hiring at Google, is blunt: the resume's sole job is to get you past the first screen and into an interview. One page keeps only what does that.

The resume's job is the interview, not the job.

Let's say every line starts with "Responsible for" and no line has a number. Rewrite each bullet as an action, a number, a result: Grew, Trained, Cut, then how much.

Passive language and missing results sit in Harvard career office's top five resume mistakes. "Duties included" is both at once.

You have numbers: customers a shift, money in the drawer, hours saved. Estimate honestly, then back the number up.

A duty says you were there. A number says what changed because you were.

Let's say you pictured a person opening your application, and it went into a database nobody mentioned. That's the ATS: recruiters search and rank it by keywords, titles, years.

The HBS and Accenture Hidden Workers study found over 90% of employers filter or rank this way before a human looks. Qualified people get vetted out routinely.

Knowing the mechanism changes the question: not "am I good enough" but "does my page carry this posting's words".

A database reads it first. Write for it, then for the human after it.

Let's say you ran stocktakes every Friday, but your resume says "inventory duties", and the posting never says inventory. Borrow the posting's exact words for skills you have, each in a sentence that shows it happening.

Stuffing is the failure mode: hidden word lists and invented skills die with the first human. Ladders found it on the worst resumes.

The honest version is translation, not decoration: you did the thing, so say it in the employer's words.

Match the words, not the vibe.

Let's say the posting went up three weeks ago, and you spent tonight polishing for a queue that already has a shortlist. Keep a ready base resume, apply inside two or three days, tailor in minutes.

Perfect-and-late loses twice: bigger pile, and some employers read as applications arrive. Ask A Manager: some review fast, some batch weeks later. Early helps either way.

The base resume makes early possible: structure, verbs, numbers, ready. Each posting gets fifteen focused minutes.

Done and early beats perfect and late.

Let's say you've sent 40 applications, heard nothing, and read the silence as feedback about you. Read it as the machine: in CareerPlug's 10-million-application dataset, about 3 in 100 got an interview, and 27 in 100 interviews became hires.

Treating every silence as a verdict is the fastest way to burn out, and it's based on a false model of hiring.

Roughly 19 of every 20 applications get no interview, for everyone. Offers land for people who keep working through the silence.

Silence is the funnel's default, not your grade.

Let's say Saturday disappeared into 40 quick-fire applications and not one used a word from the posting. Spend the same hours on ten roles you fit: posting words in, applied early, one human notified.

Spraying feels productive because the counter climbs, but each generic application gives the filter nothing to match. The HBS study found qualified people vetted out for exactly this.

Tailored wins twice: the filter finds the words, and the human finds someone who clearly read the posting.

Apply to ten like you mean it, not a hundred like you don't.

Let's say the applications live in your memory, which rehearses them at 1am like a roll of shame. Move them to five columns on one sheet: company, role, sent, reply, stage. Update when something happens, not daily.

A tracker that turns into a scoreboard, checked hourly, feeds the anxiety it was meant to calm.

The row also carries the follow-up date, so page fifteen's email goes out by calendar instead of by worry. What's next is written down, so you can stop holding it.

The tracker remembers so you don't have to rehearse.

Let's say your answers run four minutes, three of them backstory, and end with "so, yeah". Use the STAR shape. MIT weights it 20% Situation, 10% Task, 60% Action, 10% Result.

The classic fail is all Situation and no Action, told in "we" until nobody can tell what you did. MIT is explicit: "we" erases you.

Write five to seven stories in four labelled beats: a conflict, a failure, a deadline, a thing you improved, a thing you led.

Short runway, long flight, land the plane.

Let's say the first call asks your salary expectations and you freeze. Ask their range back: 'What range do you have in mind?' Some will just tell you.

A number named blind is a coin flip: high screens you out, low anchors you cheap. Vague ducking just gets you asked again.

If pressed: 'based on what I know so far, I'm looking at X to Y, same ballpark?' The homework comes from advertised ranges.

The first number anchors. Make it theirs, and make yours researched.

Let's say the interview went well and you're deciding whether a thank-you email is cringe or expected. Send it the same day: short, one specific detail, interest in next steps. USC's career center says 24 hours.

A mass-produced template, or a second and third chase, undoes it. One note, then quiet until their timeline passes.

Surveys of hiring managers keep finding most say a short thank-you helps, and few send one. The detail proves you were in the room.

Thanks, one true detail, done. Before the day is out.

Let's say the rejections feel like a crowd voting on your worth. Look who else the funnel rejects: HBS found 88% of employers agree qualified candidates get vetted out on exact criteria.

Treating each no as a personal finding makes every application cost too much, and shortens the search you still need to run.

Study patterns, not single results: zero interviews from twenty is a materials signal, and fixable. One rejection is one door.

The funnel counts. It doesn't judge.