Prateek Nair
06

Job Hunting in 2026

We’ll Keep Your Resume on File — A Data Report

2026
Role
Concept, data design, copy direction and motion direction.

I turned 287 logged applications into a report, film and deck: every next step came from a role I’d marked as a strong fit.

The report’s cover: a clock face beside a large 75, “minutes,” the fastest rejection; below, a red tab reading “Job Hunting in 2026” above the title “We’ll Keep Your Resume on File” and the line “287 applications. All sent by a person. 1 human wrote back.”
01 · Cover

A data report on my own job search. I logged 287 applications from July to October 2026 and turned the tracker into an 11-page report on how hiring actually responds: how fast the no’s arrive, who never answers, and what moved forward. The clearest finding: every application that reached a next step came from a role I’d marked as a strong fit.

The motion version: 100 seconds, cut to the beat.
Slide 2 of the readout deck, “A person sent every one of these”: 287 in red beside a grid of 287 marks, the last one red, and a “Sent by a human” stamp.
Slide 4, “1 in 287 reached a human”: five gray bars on one scale, the single human reply in red, marked “the 1”.
Slide 9, “Template bingo”: the rejection bingo card, each phrase under a red ink daub, “Hi Nair,” as the free space, and a “Bingo!” stamp.
The readout deck: the same findings as 12 slides, built to be presented.

Why I made it

Page 2, “A person sent every one of these”: a rubber stamp reading “Sent by a human” above the report’s data notes.
Page 3, “Two-thirds never answered”: bars for 190 with no reply, 85 rejected, 8 closed or withdrawn, 3 automated screens and 1 human reply.
02 · How this was made  /  03 · The response

I was applying to design jobs full time and logging every one: role, date, pay, which resume I sent and what happened. By October the tracker had 287 rows and almost no answers. I wanted to know if the silence had a shape.

Rules before charts

I set the definitions before drawing anything. Ghosted means no reply after 30 days. A human reply is a screening, an interview or a real conversation with a person; AI interviews, online tests and virtual-assistant screens don’t count. Undated entries count in the totals, never in the timing charts. No company or person is named.

The rules changed the story. My first count of human replies was four. Under my own definition, it’s one.

What the data says

Page 4, “Then I started applying like it was my job”: weekly applications rising from July to October, with no rise in replies.
Page 5, “After 30 days, it’s cooked”: one mark per silent application by days since applying, day 30 marked, and a “Ghosted” stamp.
04 · The volume  /  05 · The silence

Two-thirds never answered: 190 of 287 got no reply at all. The no arrives fast, with a median wait of five days and a fastest of 75 minutes. Rejections cluster on Mondays and Thursdays, 21% each. And every next step (any reply that moved an application forward, AI interviews and tests included) came from a role I’d marked as a strong fit: 4 of 75, against 0 of the 212 others.

Page 6, “The no arrives fast”: a five-day median wait, the 75-minute fastest no, and a calendar with one mark per rejection.
Page 7, “Mondays are rough for both of us”: applications and rejections by day of the week side by side, rejections tied at 21% on Monday and Thursday.
06 · The speed  /  07 · The timing

Designing it

Page 8, “Template bingo”: a bingo card of nine real rejection phrases, with “Hi Nair,” as the free space.
Page 9, “The bots are applying too”: a grid of bot icons with one person among them, and a job posting with a hidden code word.
08 · The letters  /  09 · The bots

I set it as a print-style research report, not a dashboard: one finding per page, a headline that states it, one chart that proves it. Kumkumam, the red of the powder used in temple rituals, is the only accent, saved for the number that matters, and on dark pages it marks things but never carries text. The humor lives in the objects, like rubber stamps, a bingo card of real rejection phrases and a clock for the 75-minute no, so the charts can stay straight.

How it was made

I art-directed it, set every rule and signed off on every line and number. The charts and first drafts were built with AI from those decisions. I also used AI to help draft and tailor my written answers, trained on how I actually write, then read every posting and sent every application myself. The raw tracker stays private; the report uses totals only.

What it can’t tell you

Page 10, “What almost worked”: 4 of 75 strong-fit roles moved to a next step, against 0 of 212 others, and three resume variants compared.
The closing page, on red: “We’ll keep your report on file,” an “On file” folder tab, three closing numbers and a “Still hiring?” tape.
10 · What almost worked  /  11 · The end, for now

It’s one person’s search. Fit was my own read of each role, mostly made before applying, and 81 entries were logged after the fact with no date. The pattern in this data is clear. It isn’t a law of hiring.

Download the full report (PDF) Download the deck (PDF)

This started as a spreadsheet I kept to stay sane and turned into the most fun project I’ve worked on this year.

Every number is real; the jokes are the only thing made up.