Every few weeks a list goes round of the highest-paying remote jobs you should be aiming for. We took the ten job titles that keep appearing on those lists, pulled 1,208 live UK adverts for them, and counted how many people had already applied to each one.
The role at the top of the competition table is the one everybody describes as a skills shortage.
AI engineer is the most contested title on the list
AI engineer adverts carried a median of 61 applications each. Across all ten titles the median was 18. Senior sales manager, advertised in a similar pay bracket, carried 5.
So an application to an AI engineering role lands in a pile roughly twelve times deeper than an application to a sales management role that pays about the same. That is the opposite of how the market gets described.
Median applications per advert
| Job title | median apps |
|---|---|
| AI engineer | 61 |
| Software engineer | 44.5 |
| Marketing manager | 23 |
| Product manager | 22 |
| Senior project manager | 18 |
| Customer success manager | 17 |
| Data scientist | 16 |
| Senior account manager | 12 |
| Operations manager | 7 |
| Senior sales manager | 5 |
There is no contradiction in an employer struggling to fill a role that hundreds of people applied to. A shortage of people who can do the work and a shortage of people applying are separate problems. Only the second one is visible from outside, and it is the one that decides whether your application gets read.
What actually predicts the queue
We expected competition to follow pay, or seniority, or how hard the work is. It follows none of those closely. What it follows is whether you can do the job from your kitchen.
Adverts describing fully remote work took a median of 34 applications. Adverts for the same ten job titles that mentioned neither remote nor hybrid working took 13. Hybrid landed in between at 27.
Remote working is not a quiet perk that employers hand out. It is the feature that turns one vacancy into a crowd, which means every list of high-paying remote jobs is also a list of the most oversubscribed openings on the board. Those are not the same as the easiest ones to get.
Only one advert in seventeen is genuinely remote
Of the 277 adverts, 5.8% described fully remote work. Another 40.8% described hybrid working, which in practice means two or three days in an office. Everything else was on site or said nothing at all.
The word remote is doing a great deal of work in these round-ups. Sixteen adverts out of 277 were the thing people picture when they read the headline. We found the same pattern in a wider sample last month, where almost every UK advert that states a working location states on site.
The pay does not transfer
These lists are usually built on United States salary data. Of our adverts that stated an annual figure, and 64.6% did, the median across all ten titles was £51,000. Only 3.9% reached £100,000.
AI engineer was the best paid of the ten, at a median £83,750. It is also the most contested, and we think the pay explains the crowd rather than sitting beside it as a separate fact. Good money, a title people want on their profile, and it gets advertised as remote more often than most.
What to do with this
The obvious move is to stop treating a shortlist of popular job titles as a shortlist of opportunities. They are the same thing everyone else read this week.
If two roles pay similarly and one takes 61 applications while the other takes 7, the second is worth several applications to the first. That holds across our wider work too. When we counted applications across 1,657 adverts in frontline occupations, the jobs that sound easiest to get were consistently the hardest, and the pattern here is the same one in a different pay bracket.
The quiet titles are quiet for reasons that have nothing to do with how good the job is. Sales roles carry targets. Operations work does not photograph well. Neither gets written up in careers advice, so the queue never builds.
None of that means avoid the popular roles. It means know the depth of the pile you are joining, and put your effort into the applications where being read is likely.
How we did this
We collected 1,208 live UK job adverts from the Reed jobs API on 28 July 2026, covering ten job titles that recur on published lists of high-paying remote work, then cleaned the set down to 277.
Three filters did that. We kept only adverts posted within the previous 30 days, because an advert accumulates applications for as long as it stays live and older adverts look more competitive through age alone. We removed near-duplicate reposts, 54 of them. And we rejected any advert whose title was not actually the role, which removed 558 on its own, because the Reed search endpoint matches the full advert text and will return a warehouse vacancy for a product manager query.
That last filter matters more than it sounds. Skipping it is how you end up publishing a market that does not exist, and it is the step that caught us out once before, when 49.3% of results for "nurse" turned out to be nursery nurses.
Sample sizes per title after cleaning: software engineer 44, product manager 41, senior project manager 40, senior account manager 31, senior sales manager 28, customer success manager 24, operations manager 20, AI engineer 19, data scientist 15, marketing manager 15. The four smallest are directional. The gap between the top and bottom of the table is far too large to be a sampling artefact, but we would not defend any single median to the decimal place. The overall figures, including the 5.8% and the 34 against 13, rest on all 277.
Working arrangements were read from the advert title and body. Fully remote counts explicit phrasing such as fully remote, 100% remote or home-based. Salary figures use the midpoint of the stated range and cover the 64.6% of adverts that gave one, so they describe what employers advertise and not what anyone was paid.
These figures describe one moment in the UK market. They are not a measurement of change over time, and the pay comparisons should not be read across to other countries. More of our work is at Sausage Dog research.