The Interview Invites Already Went Out. Now What?

Picture a recruiter who just realized the job description behind a whole batch of interview invites was the wrong version, minutes after the emails went out. The candidates are real, the invites are already sitting in their inboxes, and the only question worth asking now is whether there is any way to cancel interview invites already sent to candidates before more of them click through. That is a different, harder problem than deciding to halt a campaign. Halting is a decision made once. Pulling back invites that are already out is what happens next, to a room full of people who already have a link.

Why “Stop” and “Already Sent” Are Two Different Problems

We’ve written elsewhere about the decision itself, so this piece won’t retread that ground. What that earlier piece doesn’t answer is the question sitting in front of HR at 6 PM on a Tuesday: the invites are already out, so what happens to the people who haven’t opened theirs yet, and to the ones who open it five minutes from now?

That gap is where this piece lives. Pulling the plug on a run and clawing back the links it already fired off happen to be triggered by the same click, but they are answering two entirely different worries for whoever just clicked it: don’t let anyone new in, and make sure the ones already holding a link can’t use it either.

What a Candidate Sees When They Click a Cancelled Invite

Before this shipped, a join link had no way of knowing the batch behind it had been pulled. It just sat there, joinable, because nothing told it otherwise. Now, the instant HR pulls the batch, the link returns a 410 response, the web’s standard way of saying a page is gone on purpose rather than merely missing, with a plain note that the recruiter called off the screening.

That wording choice matters more on a college campus than almost anywhere else. A candidate who hits a broken-looking page screenshots it and drops it in the class WhatsApp group within minutes, and now thirty batchmates think the product is unreliable before a single one of them has spoken to you. A candidate who reads a note explaining the recruiter called this off tells the same group the company changed its plans instead, which is a very different rumor to have circulating before your next drive lands at that same college.

What Cancelling Leaves Alone

Cancel does not mean everyone loses their interview. It reaches exactly two states: a meeting marked scheduled and one marked waiting, meaning nobody has joined it yet. Two categories are left completely alone.

An interview already in progress keeps running until the candidate finishes. A candidate mid-conversation with the voice agent does not get cut off because someone in HR clicked cancel on the campaign. And an interview that already completed keeps its transcript, its score, and its verdict exactly as recorded; cancelling a campaign afterward does not retroactively erase work HR already has.

This is a narrower blast radius than “cancel campaign” sounds like on its own, and that’s deliberate. The whole point of a cancel control is to stop the parts of a campaign that are still avoidable, not to punish the candidates who already gave you fifteen to twenty minutes of a real conversation.

Picture a drive that invited 200 candidates. By the time HR notices the JD mistake, say 40 have already finished their interview, 10 are mid-conversation with the voice agent right now, and the remaining 150 have not clicked yet. Cancelling the campaign leaves the 40 finished interviews exactly as scored, lets the 10 in-progress calls run to their natural end, and turns the other 150 links into the HTTP 410 message. Three different outcomes from one click, sorted correctly without HR having to triage each candidate by hand.

The Credit Refund Nobody Has to File a Ticket For

Every unsent, unopened, or never-started interview represents a credit the company already paid for. When a batch is pulled, HireQwik sends the unused credit for every candidate whose interview never started straight back to whoever paid for it. Nobody on the HR side has to email support or open a ticket to get it back.

What actually protects that number is a single database update per candidate, which means a nervous double-click on the cancel button cannot pay the same credit out twice. In practice, the person who notices this isn’t usually the one who clicked cancel. It’s whoever reconciles the account later and finds the balance already matches what was genuinely spent, with no manual adjustment line to write for “batch pulled on the 4th, refund pending.”

The Eight Hours You Used to Be Stuck With

Before the link revocation piece existed, an emailed invite stayed joinable for up to eight hours no matter what else happened on the admin side. Eight hours is most of a working day. For a campus drive of the kind that screened 3,000 candidates in a single evening, that is easily long enough for a meaningful share of an already-invited batch to click through on a drive HR believed was long since over.

That number is the reason a dead link matters more than a database flip most days: a database can say “cancelled” the instant HR clicks the button, but for eight hours the actual experience a candidate got told a different story. Closing that particular gap, not the button press itself, is what turned “we marked it cancelled” into “the candidate cannot reach it.”

Where This Lives, and Where the Line Is Different for a Live Booking

The full walkthrough of triggering this from the dashboard is here; this post is only about the aftermath, so it won’t repeat that ground. What’s worth adding is the boundary case a recruiter runs into next: someone who already has a confirmed slot on a campaign that stays healthy just wants to move it, not kill it. HireQwik keeps that as a separate action, capped at three reschedules within 72 hours, because rescheduling one person shouldn’t touch the other hundred and ninety-nine invites sitting in the same batch.

Timing counts for whatever HR launches next, too. A wrong-JD batch that stays live keeps eating the same booking slots a corrected relaunch would want on that same morning. Clearing the bad batch first means the relaunch isn’t competing with its own mistake for candidate attention on the calendar.

One honest limit worth naming: none of this proactively emails anyone. The dead-link message only shows up for a candidate who actually clicks; someone who never opens the invite again never sees it. If HR wants those 150 still-unopened invites to know the drive is off before they even try, that’s a manual note the team sends, same as on any other platform. What changed is only what greets the person who does click.

What Most ATS Cancel Buttons Actually Cover

Almost every mainstream applicant tracking tool ships some version of a cancel button, but nearly all of them were built with a single applicant in mind, not a spreadsheet of two hundred names uploaded together for one drive. Greenhouse’s own self-scheduling cancellation guide is explicit that the system does not automatically notify the candidate; the recruiter has to message them separately, one at a time, if they want the candidate to actually know. Karat and Oracle Taleo document the same single-candidate pattern in their own scheduling tools. None of them describe what happens when the thing being cancelled is the whole invite list of a campus drive.

That gap is the reason the dead-link message earns its keep as much as the credit refund does. When you are cancelling one candidate’s interview, a recruiter sending a manual follow-up email is a reasonable ask. When you are cancelling two hundred, expecting a recruiter to message every candidate by hand before someone clicks a dead link is not a real plan, it’s a hope. Making the link itself explain what happened is what makes batch cancellation survivable instead of just possible.

Most of these systems were designed assuming a recruiter manages a handful of individual records at a time, not the couple hundred names pulled into one campus spreadsheet and screened together. A workflow built for onesies breaks down the moment the unit of work becomes a whole batch.

The Placement Office Hears About This Before You Do

On most Indian campus drives, the first complaint about a dead interview link doesn’t reach HR at all. It goes to the college’s Training and Placement Officer, the person who vouched for your company to get the drive scheduled on campus in the first place. If a few dozen students independently report a broken link within the same hour, that’s now the TPO’s problem to explain to their own batch, whether or not anyone at HireQwik or the hiring company ever hears about it directly.

A dead link that explains itself changes what the placement office has to say. “The recruiter called it off, they’ll follow up” is a sentence a TPO can repeat with confidence to a room full of anxious final-years juggling three other offers and a looming graduation deadline. A dead link with no explanation forces a guess instead, and a wrong guess from the placement office spreads through a student WhatsApp group faster than any correction HR sends afterward, and outlasts your next visit to that campus.

Reputation on a campus compounds across recruitment seasons. Seniors talk to juniors, juniors become next year’s applicant pool, and a placement office remembers which companies handled a mess gracefully versus which ones went silent. None of that shows up in any dashboard metric, which is exactly why it is easy to underrate until a TPO quietly deprioritizes your slot the following spring.

What to Check Before You Send the Next Batch

A wrong JD version is one reason a recruiter reaches for cancel. A wrong candidate list, a rubric that turns out to be broken (the kind that scores everyone at exactly 100% instead of spreading candidates out), or a WhatsApp template that went to the wrong role are others. HireQwik’s WhatsApp invites and reminders go out per JD for exactly this reason: a mistake stays contained to the one role it was configured for, not every open drive at once.

None of that makes an earlier mistake impossible to catch before launch; a separate review step exists for exactly that. What changes here is the cost of catching one late, after HR has already hit send. Cancel is the pullback: a dead link for anyone who hasn’t joined yet, untouched results for anyone who already finished, and money that comes back without a ticket. That is the actual safety net once “before” has already passed.

The follow-up message most teams skip

Cancel handles the technical side cleanly: dead links, untouched completions, automatic refunds. It does not handle the goodwill side, and that part is still worth doing by hand.

A short note to the placement office, even one line, saying the drive is cancelled and a corrected version is coming, costs a recruiter two minutes and buys a lot of patience. The alternative, silence until the corrected batch shows up unannounced, reads as if nothing went wrong the first time, which is rarely how a TPO fielding confused calls from students will remember it.

The same goes for the small number of candidates who already reached out asking why their link stopped working. A short reply pointing back to the message they already saw on the dead link does more good than it sounds like it should, mostly because it tells them a person actually read what they wrote, instead of leaving them to assume the silence means the company forgot about them entirely.

If a campaign you already launched needs to come back, or you want to see the cancel and refund flow before you’re mid-drive with two hundred emails already gone, talk to us and we will walk you through it on your own JD.