“I’d like to manage the move before the sun runs out of fuel.” Read on its own, that sounds like a joke about deep time. It wasn’t written as one — it’s a single line from an April 2024 letter to Alison Green at Ask a Manager, buried partway through a much longer complaint: a new resume for every posting, a “vision for the company” nobody asked for, cover letters that had turned into short stories, prep for what the writer described as a thousand possible interview questions.
Alison’s answer amounted to: you’re doing several times more work than the job requires. One master resume, trimmed per posting. One or two cover letters, lightly reworded. Interview answers built from five or six real stories you already have, reused across whatever gets asked. The letter and Alison’s reply ran under the headline “job searching is so much work,” and it drew 254 comments, a good number of them from people describing exactly the system the letter writer didn’t have yet.
That’s the version of the cost anyone can see: hours, spent inventing something from scratch that should have been a five-minute edit. It’s real, and it’s the one people put into their complaints. It isn’t the expensive part.
What actually wears a search down isn’t the hours
“For the stats,” a commenter going by “yet again” wrote further down that same thread, after describing a job search that had taken about a year: applied to around 50 to 60 jobs, got first-round interviews at roughly 20 companies, made it through a full interview cycle at four or five, and came out with two offers. Read backward, once it’s over, that’s a funnel that worked precisely the way funnels are supposed to — wide at the top, narrowing in stages. Read forward, in month six, it’s twenty simultaneous relationships with companies who each expect you to remember which resume version they got, which interviewer said what, and whose reply is overdue.
One commenter, “bamcheeks,” put the wrong intuition about all this more plainly than I could: “jobhunting isn’t a lottery where the more tickets you buy the more chances you have.” Buying twenty tickets and forgetting which drawings you’re even in doesn’t raise your odds. It just adds twenty things you can get wrong in a way that has nothing to do with your qualifications.
Alison Green, Ask a Managerjob searching is so much workThe letter and its 254 comments — read past Alison’s answer for the commenters describing their own master-resume-and-two-cover-letters setups.The waiter forgets the check the moment it’s paid
In the 1920s, the Gestalt psychologist Kurt Lewin noticed something at a Berlin café: a waiter could rattle off a long, unpaid table’s order from memory without looking, but couldn’t recall a thing about it once the bill was settled. He mentioned it to a student, Bluma Zeigarnik, who turned it into a real experiment — people asked to do a set of small tasks, some interrupted before they finished, remembered the unfinished ones noticeably better. The finding got her name attached to it: the Zeigarnik effect, published in 1927.
A job search generates the same kind of open check, dozens of times over, except none of them ever visibly close. A rejection is closure, in its flat way. Silence isn’t. An application nobody has responded to in six weeks is a table that never got its bill, sitting open in whatever part of your attention keeps track of things you haven’t finished, whether you’re consciously thinking about it or not.
I want to be straight about how solid the ground under that idea actually is, because it’s shakier than the version that gets repeated in productivity blogs. Several later attempts to reproduce Zeigarnik’s original result failed outright, and a 2025 systematic review of everything published on it since found no reliable memory advantage for unfinished tasks at all. What did hold up across that review was a related, less famous finding from her colleague Maria Ovsiankina: not better recall, but a measurable pull back toward an interrupted task — an urge to finish it, whether or not you remember the details clearly. So the honest claim isn’t that a stale application is quietly eating your memory. It’s narrower and, if anything, more relevant to a job search: an open thread doesn’t sit still. Some part of you keeps trying to go back and finish it, at moments you didn’t pick.
WikipediaZeigarnik effectThe waiter story, and — further down the page — the 2025 meta-analysis that complicates it. Read past the headline finding for the honest version.The other side of the exchange has stopped closing loops
None of this would matter much if employers reliably told you when a thread was over. Increasingly, they don’t. Greenhouse’s 2024 State of Job Hunting report, built from its own platform data plus a survey of 2,500 workers, found that 42 percent of candidates want stronger recruiter communication above anything else they’d change about the process — ahead of wanting to stand out more, ahead of wanting clarity on how AI is used in screening. Three in five candidates said they suspected they’d applied to a “ghost job,” a posting live with no real hiring intent behind it, and Greenhouse’s own internal numbers back that suspicion: 18 to 22 percent of jobs posted to the platform in any given quarter get classified as exactly that.
Greenhouse2024 State of Job Hunting ReportThe recruiter-communication and ghost-job-suspicion numbers here are different figures from the ghosting and mass-application stats this blog has cited elsewhere from the same report.A posting that was never going to result in a hire still generates an open thread on your end. Nothing on the company’s side is ever going to mark it closed, because as far as their system is concerned, there was nothing to close.
The offer that never happened
This is the version of the cost that’s hardest to see, because it never announces itself as a loss. Two processes are moving at once — one slower and further along, one that started later but is suddenly asking for a decision — and attention, being finite, goes to whichever one is currently loud. The quieter thread doesn’t get a dramatic ending. It just stops being followed up on, until the recruiter on the other end reasonably assumes you’ve gone with someone else and moves down their own list to a candidate who replied.
Nobody sends a rejection for that.
There’s no version of the story you ever get to hear the ending of, which is exactly why almost nobody counts it among the costs of a bad search — you can’t total up an offer that quietly stopped being possible. What you can count, in the smaller and less dramatic version of this, are the ordinary mistakes a lot of open threads produce on their own: the follow-up note that goes to the wrong recruiter with details from a different company still in it. The application to a role you already applied for in March, because the posting resurfaced and nothing told you that. The deadline that passed because the email with the date on it got archived along with forty others the same week.
- Which resume version went where, and what it emphasized.
- Who said what, and what you told them you’d follow up on.
- Whose move it currently is — waiting on them, or waiting on you.
- The actual date something is due, not a rough sense of “soon.”
Four fields, kept current, is a smaller ask than it sounds, and it’s the specific, unglamorous thing Vigil exists to do off the inbox that’s already collecting all of it, whether or not anyone is watching closely enough to notice a thread went quiet.
A spreadsheet does the same job for free, if you’ll actually open it.
What you don’t get, with any system, is proof of which door quietly closed while you were looking at a louder one. That absence of proof is most of why the cost stays hidden. It’s also not a reason to stop looking for the door.