Last updated: · Dataset: 5,681 UK job adverts, 12 daily snapshots
Search for pharmacy assistant work and you will get hundreds of results. It looks like a healthy market. We tracked every live advert in that role for twelve days and counted who was actually behind them. 345 adverts. 14 employers. Asda posted 57% of them.
The advert count is the number everybody sees. The employer count is the one that decides how many genuinely different chances you have, and nobody shows you that.
You are not applying to 345 jobs. You are applying to Asda, repeatedly, in different towns.
Ten roles, and how few companies are behind them
We tracked ten occupations. For each one, here is the advert count, the number of distinct employers, and how much of the list the single biggest employer owns.
- Pharmacy assistant: 345 adverts, 14 employers. Asda is 57%.
- Optical assistant: 555 adverts, 36 employers. Specsavers is 50%.
- Paralegal: 1,110 adverts, 152 employers.
- Staff nurse: 714 adverts, 108 employers.
- Dental nurse: 72 adverts, 11 employers. Bupa is 49%.
- Occupational therapist: 377 adverts, 67 employers.
- Barista: 128 adverts, 32 employers. Costa is 60%.
- Receptionist: 843 adverts, 215 employers.
- Payroll administrator: 1,402 adverts, 492 employers.
- Library assistant: 135 adverts, 52 employers.
Payroll administrator is the healthy one. 1,402 adverts spread across 492 employers, and the largest of them posts 9%. If that search disappoints you, there are hundreds of other companies to try.
Pharmacy assistant is the opposite. Fourteen employers, and more than half the list belongs to one supermarket. Send forty applications and you are mostly sending them to the same recruitment team.
Most repetition is real, and that matters
53.5% of the adverts shared an employer and a job title with at least one other advert. That sounds like a job board full of duplicates. It mostly is not.
We counted how many copies of each repeated advert sat in each distinct location. The median was 1.14. Of the 625 repeated sets, 389 ran at 1.5 copies per location or fewer, which is what genuine multi-site hiring looks like: Asda posted 122 adverts across 96 locations, Vision Express 119 across 101. Those are real shops with real vacancies.
Only 18 sets ran above three copies per location. So the honest read is that the repetition is not a scam, it is geography. Which is still a problem for you, because a job at a Specsavers 200 miles away is not a job you can take.
The thing we could not prove
Pooled across everything, adverts inside a repeated set carried a median of 5 applications against 12 for adverts that stood alone. That looked like a finding: the listing you keep scrolling past is the one nobody is applying to.
Then we split it by occupation, which is where this sort of thing usually dies. It survived in six of eight roles and reversed in two. Receptionist showed a gap of 29 applications; occupational therapist showed one. Barista and paralegal went the other way entirely.
A pattern that flips in a quarter of the occupations we tested is not a rule, so we are not going to sell it as one. We have killed a better-looking finding than this before for the same reason.
What to do with this
Before you send anything, count the employers in your search rather than the adverts. Sort by company. If three names cover most of the page, you have three applications to get right, not thirty.
Then tailor for the company, not the job title. When one employer owns half the market in your role, their advert wording is effectively the marking scheme. Generic beats nothing, and specific beats generic.
And if your role is a concentrated one, widen the search before you widen the volume. Fourteen employers is a small number of doors to knock on, however many adverts they have between them.
Methodology and limitations
5,681 adverts across ten UK occupations, captured through the Reed API once a day from 1 to 12 August 2026, twelve snapshots. Occupations were pharmacy assistant, optical assistant, paralegal, staff nurse, dental nurse, occupational therapist, barista, receptionist, payroll administrator and library assistant.
Employers are counted by exact name match on the advert's employer field. That understates concentration, because one company filing under two spellings counts twice. Repeated sets are adverts sharing both an employer and an identical job title. Some of the employer names are recruitment agencies rather than the hiring company, so the true number of end employers is lower again.
Limits. One job board, one twelve-day window, ten occupations chosen for coverage rather than at random. This describes what a candidate sees when they search, not the whole UK labour market. We deliberately make no claim about how long adverts stay live: departures need two consecutive missed snapshots to confirm, so at daily cadence any lifespan figure would measure our polling interval rather than employer behaviour. The application comparison in this piece is reported as inconsistent because it is, and no causal claim is made about why repeated adverts draw different numbers.