Intentional News

261012 Why Applying to 100s of Jobs Online Doesn't Work Anymore — the Forbes Future of Work Panel

11 min read

Thesis (Alf’s read): The panel’s most useful finding is not that job-hunting has got harder — it is that the market has become a coordination failure, and the individual cannot fix it from inside. Candidates auto-apply at scale because the funnel is opaque; employers AI-screen because they are buried; each side’s rational response makes the other’s problem worse. 93% of applications are never read. So the strategic answer for a student in 2026 is not “apply better” — it is exit the loop: fewer, targeted, human-routed approaches, in the market’s current vocabulary, backed by a verifiable body of work rather than a document. Everything else is optimising inside a machine that has already decided not to look.

Why Applying to 100s of Jobs Online Doesn’t Work Anymore

  • Forbes, “Why Applying To 100s Of Jobs Online Doesn’t Work Anymore” — Forbes Future of Work Summit, New York, uploaded 7 October 2026, 15m34s — https://youtu.be/oGkb8gUmdCI
  • Panel: Steven Bertoni (Forbes, moderating), Daniel Chait (co-founder and CEO, Greenhouse — an applicant tracking system), Erin McGoff (founder of StupidFish; careers creator)
  • Erin McGoff on the origin of the company name — https://youtu.be/oGkb8gUmdCI

The doom loop

The panel’s own name for the situation is the doom loop, and the mechanism is worth stating precisely because it is a coordination failure rather than anyone’s fault:

  1. Candidates auto-apply at scale and use AI to tailor each résumé to the job description.
  2. Employers receive thousands of lookalike applications — “résumé slop”.
  3. Employers use AI to hammer the list back down.
  4. That makes it harder to be seen, so candidates send more applications.
  5. Which raises the filter again. Round we go.

Chait’s framing: “This is really the first time in anyone’s memory where both sides are unhappy.” Both sides experience the shortage; 93% of job applications are never looked at.

Two facts from the panel that explain why the squeeze is structural rather than cyclical:

  • Skills change about twice as fast as people’s profiles. LinkedIn’s own CEO, on stage at their conference, reported that the skills listed in job descriptions have changed roughly twice as fast as the skills people list on their own profiles. So the market is asking for things candidates have not yet learned to name.
  • It is a “low hire, low fire” market. Fewer hires, fewer firings — which makes every vacancy contested and every bad hire more expensive.

The CV point — which Hiro is right to single out

Both panellists want the résumé dead, and their reasons are specific rather than trendy.

  • McGoff: “I would wave my magic wand and get rid of résumés forever… we make résumés the exact same way we did 50 years ago — we open up a PDF and we type. And it’s not even a standardised document for goodness sake. It’s ridiculous that we still do things this way and we have self-driving cars.”
  • Chait: “Why do we still have rectangular white résumés that used to have to print out and put into an envelope, when that’s not what happens at all? Why do we have to gatekeep who gets an interview in a world where we can have automated conversations at scale?”

What they propose instead: a personal (professional) brand rather than a document; skill verification (they both flag people working remotely as people they are not); recommendations and referrals to validate soft skills; and — Chait’s version — a personal agentic job-search partner that knows the candidate and represents them.

My read on the CV critique — it is right, and the important part is not the format. The panel argues about the document; the substance is that a CV is a claim, and a claim that everyone can generate perfectly has stopped carrying information. When every application is polished, polish tells you nothing. That is why the openings are now where a person can be verified rather than described.

The Einstein quote — and the correction

McGoff names her company StupidFish after a quote she attributes to Einstein: “==Everybody’s a genius, but if you judge a fish on its ability to climb a tree, it’ll live its whole life believing that it’s stupid.” Her gloss: “everybody has an ocean out there, but we just have a lot of fish trying to climb trees.”==

The point is good and the attribution is false. USA Today fact-checked this line in April 2021 and found no evidence Einstein ever said it — it circulates from educational allegory and self-help writing, not from anything Einstein wrote or is recorded saying. I am flagging it because it is in the video Hiro liked and it will be repeated: quote the idea, not the man.

The idea is the strongest thing in the panel, and it lines up exactly with the crisis this vault already tracks: a hiring funnel that judges everyone by the same instrument will systematically misjudge people whose abilities do not show up on it — which is the whole argument of Age Discrimination in Work - Global Problem and Youth Disconnection and NEETs - Global Problem, arriving from the other direction.

What to do instead of the legacy methods

Hiro’s question, answered from the panel plus this vault’s own research. Ordered by what actually changes an outcome.

What to stop doing

Legacy methodWhy it now fails
Applying at volume140 applications per vacancy. Volume is what the loop is built to defeat — the number of applications you send is now evidence you will not be read
Letting AI write the applicationThe lookalike problem exists because everyone’s AI writes alike. An application with no sentence only you could have written answers nothing
Treating the CV as the assetIt is a claim anyone can generate. It is the wrong thing to spend a weekend on
Searching last decade’s job titlesThe sales-ops → “go-to-market engineer” example: people have the skills, the market has the name, and the search term is the barrier
Waiting to be pickedThe funnel is where you go to be invisible

What to do instead

  1. Quit volume: target density. Ten tailored approaches beat a hundred generated ones. This is the single highest-value change and it costs nothing.
  2. Learn the market’s current vocabulary for the thing you already do. McGoff’s most useful practical tip is that people “continue looking for a job the same way they did 5 years ago, but the job has changed” — so find the new title for your skill and search for that. This is free and it opens doors.
  3. Go around the funnel. McGoff’s number-one advice is network, and she is honest that it is unfair: “that favors people who already have connections… Yeah, it really is unfair.” The honest student version of networking is not nepotism — it is informational conversations with people doing the job, which anyone can start cold.
  4. Build durable skills, not tool skills. Chait: “none of us knew [Cursor] a year ago and we don’t know what they’re going to come up with next week.” What survives: learning fast, powering through adversity, being inventive, being a good teammate, leading. Every tool you learn will be replaced; that list will not.
  5. Build a body of work, not a document. Since verification is the direction of travel, the asset that appreciates is evidence you can point at — a project, a placard, a placement, something shipped. A portfolio is verifiable; a CV is asserted.
  6. Use AI as an editor, not an author. Tightening what you wrote is a genuine advantage. Generating it puts you in the slop pile.
  7. Judge the interview process while you are in it. The panel surveyed thousands of job seekers and found they want three things as part of the interview experience: a natural conversation, transparency and control (told it is AI, with the ability to opt out), and explainable decisions built on a stated rubric — “if it’s just a black box where [the model] is going to decide your fate, people hate that, rightly so.” Those three are also a good test of whether an employer deserves your effort.

[! Public] Review the Interview process - Glassdoor Do people still use this to rate review companies and their interview process?

  1. Walk away from processes that will not tell you what they are. It is a legitimate answer, and enough people doing it is the only lever candidates have.

The one thing that is not a tactic

The loop cannot be won; it can only be left. Every item above is a way of routing around a machine rather than beating it. That is not defeatism — it is the same conclusion this vault reached from the employer’s side in Recruitment That Values People, where the candidates’ playbook is explicitly labelled “given this market, not the market they deserve.”

What this means for a parent

My assessment (Alf)

The panel is right about the problem and unintentionally revealing about the solution. Note who is on it: the CEO of an applicant tracking system and the founder of a tool that interviews people to extract their “hidden skills”. Both diagnose the doom loop accurately and both propose, as the fix, more technology in the same funnel. Chait’s own framing gives it away — “the sense of helplessness that you feel is because as a job seeker today, there’s like nothing productive you can really do to give yourself a better chance.” The industry is telling candidates they are helpless inside a system the industry sells.

The most honest sentence in the fifteen minutes is McGoff conceding that her own advice is unfair. “That favors people who already have connections… Yeah, it really is unfair.” That is the whole equity problem of this market stated plainly by the person giving the advice — and it is the same finding this vault reached from the age end of the funnel: the informal route is where the openings are precisely because it cannot be automated, and that is exactly why it advantages whoever already has it.

But the panel misses the thing that matters most to a student. Everyone on stage treats the résumé as the artefact to replace. ==The actual shift is from claims to evidenc==e. A CV was always a claim; it worked because claims were expensive to produce. AI made claims free, so the market has started pricing what cannot be faked — verification, referrals, work that can be inspected. That is genuinely good news for a student with time and bad news for a mid-career applicant with none, and it is a much better guide to what to do with a university year than any advice in the panel.

Where I would push back on the framing Hiro brought. He asked how a student avoids “ineffective legacy methods” — and mostly that is right. But one legacy method is not dead and has not been mentioned: the placement year, the graduate scheme and the apprenticeship are still the highest-conversion routes in the UK, because they are structured, employer-owned and not routed through the open funnel at all. The modern version of the old advice is not “network more” — it is get into a process that was designed to hire you before you apply.

And the correction that matters most for how this gets quoted. The panel’s best line is not Einstein’s and the market’s core problem is not a skills gap. It is a matching failure with an incentive to persist: nobody in the loop is behaving irrationally and the loop still cannot produce good outcomes. That is the definition of a market that needs a floor beneath it — which is precisely the argument, and the reason for the regulatory lever, in Recruitment That Values People.

Graph View

Backlinks