Your busiest vacancy and your emptiest one look identical on the dashboard
Across 233 staff nurse adverts tracked for a month, not one reached ten applications. The busiest tenth of them carried two. Over the same period the busiest tenth of receptionist adverts carried 53. The roles with the loudest shortage headlines are the ones nobody is applying to, and the quiet office roles are where the queue actually forms.
Most of us would have guessed that the wrong way round, and the instinct is reasonable. Hard-to-fill clinical roles are assumed to be hard because the pool is small but motivated, and administrative roles are assumed to be easy because the pool is enormous. If that held, nursing adverts would show a short queue of serious candidates and receptionist adverts would show a long queue of weak ones. What the numbers show is closer to the opposite, and the difference is not marginal.
Since 1 August we have been recording UK job adverts from the hour they appear and going back every day to count the applications on each one. The panel now holds 2,482 adverts across 16 occupations. Every advert is enrolled within 24 hours of going live, so its age is measured rather than guessed, which matters more than anything else in the method. An application count means nothing until you know whether the advert went up this morning or three weeks ago.
Counting applications on adverts more than seven days old, the busiest tenth in each occupation runs like this. Receptionist 53, from 187 adverts. Paralegal 52, from 155. Payroll administrator 48, from 171. Administrator 37, from 150. Optical assistant 32, from 65. Customer service adviser 19, from 72. Then the drop: marketing executive 6, from 584. Care assistant 5, from 249. Occupational therapist 3, from 104. Staff nurse 2, from 233. Teaching assistant 2, from 48.
Marketing executive is the one we did not expect. It is the largest set here by a distance, it is the role everyone assumes is drowning in graduates, and its busiest tenth carries six applications. The number of vacancies and the number of applicants are different things, and we had been quietly treating them as one.
There is a second finding buried in the timings. The median UK advert collects one application in its first 24 hours. Across all sixteen occupations the first-day median is zero, one or two, and that is the entire range. Reaching ten applications takes a median of 24 hours for an administrator advert, 71 hours for a receptionist and 116 for a paralegal. If your process assumes a flood on day one, it is calibrated for something that does not happen.
We should be straight about the limit before drawing anything from this. These are Reed application counts, so they capture applications made through Reed. NHS nursing largely recruits through NHS Jobs and TRAC, and a nurse who sees a Reed listing may well apply somewhere else entirely. So we are not saying 233 nursing adverts received almost no applications. We are saying they received almost none through this channel while receptionist adverts on the same site in the same month received plenty. We lean toward the gap being mostly real, because care assistant sits at five and occupational therapist at three and neither routes through NHS Jobs the way band 5 nursing does. If the whole effect were an NHS Jobs artefact, those two should look like receptionist, and they do not. That is suggestive rather than settled.
For HR teams the consequence is the same either way, and it is a measurement problem before it is a recruitment one. If you recruit across a mixed workforce you are almost certainly measuring all of it the same way. One time-to-fill target, one escalation trigger, one weekly report where every open vacancy is a row. On this evidence you are running two unrelated problems through one process, and the process cannot tell them apart.
For the contested roles the problem is triage. A receptionist advert whose busiest version pulls 53 applications does not need more reach. It needs a faster and fairer route through the pile, because every day it stays open adds candidates nobody will read, and the strongest ones are the first to take something else. Reach spending on those roles is money going nowhere.
For the shortage roles reach is the entire problem, and the open advert is doing nothing at all. A nurse vacancy sitting at zero applications after ten days has not been read and turned down by anybody. Nobody has seen it. If your process escalates at day 28 because that is when it escalates things, you have spent a month waiting for a queue that was never going to form, and that month is the whole cost. It is invisible on a report that counts days elapsed rather than applications received.
Three things worth doing with this.
First, set your escalation trigger on applications received rather than days elapsed. An advert with zero applications on day three is telling you something urgent. An advert with 40 on day three is telling you something completely different, and a rule that fires on day 28 serves neither.
Second, segment your time-to-fill target by application volume rather than by department. A role that attracts one applicant a fortnight will never hit the same target as one that attracts eight in a week, and holding both to the same number makes the report meaningless rather than making the recruiters faster.
Third, check where your shortage roles are actually being advertised and where the applications are arriving. If the traffic is going to a channel nobody is watching, the dashboard is measuring the wrong thing and every decision downstream of it inherits that error.
None of this is expensive. It is mostly a matter of noticing that a number already sitting in your own system tells you which of two opposite problems you have, and that treating them as one guarantees you handle both badly.
The figures come from 2,482 UK job adverts across 16 occupations, collected daily from the Reed API between 1 August and 2 September 2026 across 33 snapshots. Only adverts first seen within 24 hours of going live are included, so advert age is measured rather than inferred, and occupations with fewer than 30 tracked adverts are excluded. Application counts are Reed’s own figures as displayed on each advert, which means applications through Reed rather than total applications, and this matters most for NHS roles that also recruit through NHS Jobs and TRAC. One job board, one month, and sixteen occupations weighted toward office, care, health and education work. Roughly 84% of adverts in a comparable sample came from recruitment agencies rather than direct employers, so an agency reposting one vacancy across several towns appears more than once, which would push the busy end of the range down rather than up. No causal claim is made: this describes how many people applied, not why.
https://sausagedog.io
https://sausagedog.io/blog/receptionist-vs-nurse-applications