← All writing

Field Notes / Resume

Resume Optimization in 2026: Write for the Human, Not the Robot

I read stacks of resumes from the hiring chair for a decade. The ATS is not the gatekeeper you fear. A tired human deciding in seconds is. Build for that person.

By Nicolai Imset8 min read
resumeatscareers
The fear vs the fact

What the fear sells

  • A robot rejects you on missing keywords
  • You win by gaming the ATS
  • Stuff in every keyword you find
  • A bold creative template stands out

What happens

  • A human decides in about seven seconds
  • Stay parseable, then write for the reader
  • Stuffing reads as spam, not signal
  • Tables and graphics break the parser

I read resumes from the other chair

I spent thirty years in enterprise tech. For the last ten I hired solution engineers. I screened the stacks. I built the shortlists. I rejected strong people for resumes that buried their value. Now I write my own again. Oracle made my role redundant. The view from both chairs changed how I read the advice. Most of it sells fear.

The robot is not your gatekeeper

The industry runs on one story. A machine reads your resume. It counts keywords. It rejects you before a human looks. So you must beat the algorithm. One marketing page I read this week printed a figure, sixty-seven percent of agencies screen with AI, with no source attached. The fear sells software. The story is wrong in the part that matters.

Modern parsers rank and route. They rarely auto-reject a qualified candidate over a missing phrase. The decision that ends most applications is human. A recruiter opens your resume. They spend about seven seconds on it. They read your current title. They read your last title. They check the dates for a clean line. Then they move on. Or they read more. The 7.4 second figure comes from Ladders eye-tracking research.

Optimize for that human. Stay readable for the machine. The order matters.

What the seven seconds land on

The recruiter eye moves in an F. Across the top. Down the left. Ladders mapped it with heat maps. Build for that path.

  • The top third does the work. Your title, your current company, one line of proof. Put your strongest result where the eye lands first.
  • Titles carry the story. A clear title beats a clever one. The reader scans titles to map your trajectory.
  • Numbers earn the pause. "Cut onboarding time forty percent" stops the eye. "Responsible for onboarding" does not.
  • White space reads as confidence. A cramped page signals a cramped thinker. Give the eye room.

Stay parseable, then stop

The machine work is real. It is also small. Clear it once. Move on.

  • Use a clean single column. Tables and text boxes break parsers. Your skills vanish into a void.
  • Keep contact details in the body. Headers and footers drop out of many systems.
  • Send the format they ask for. DOCX parses cleanly. PDF holds your layout. Match the application.
  • Name sections plainly. "Work Experience." "Skills." The parser knows those words. It guesses at the clever ones.

That is the whole machine checklist. It takes one pass. Spend the rest of your time on the human.

Keyword stuffing is a tell

The fear story ends with stuffing. Paste the job description in white text. Pack a skills wall. List forty tools. Recruiters read this in a glance. So do modern filters. A wall of keywords with no result attached signals a weak candidate, not a strong one. Match the role's real language where it is true. Tie each skill to a result. Drop the rest.

Tailor in twenty minutes, not two hours

Build one master resume. Keep every real achievement in it. For each role, adapt the top. Rewrite the summary to name the company's problem. Reorder bullets so the relevant wins sit first. Cut the lines the role does not need. Twenty minutes per application. No blank page. No burnout.

One more 2026 note. Illinois HB 3773 took effect on January 1. Employers there now notify applicants when AI screens them. More states will follow. The tools that read you are becoming visible. Build a resume that survives both readers. The machine, and the person behind it.

Where PrepEdge fits

I built PrepEdge to do the resume work honestly. It compares your resume to the target role. It flags the gaps a recruiter would question. It rewrites bullets toward impact. It keeps the format parseable.

It does not promise to trick a robot. There is no trick. A clean resume that names real results clears the parser and holds the human. That is the whole game.

Write for the person reading in seven seconds. Stay readable for the machine doing the sorting. Tie every claim to a result. Do that. You stop fighting a robot that was never the real judge.