The £9,940 gap and the roles closing it

Where the median comes from

We build the distribution from real, posted salaries on sanctioned job boards — only roles that actually disclose a band, deduped across sources. For machine learning engineers the disclosed bands run from roughly £57,806 at the 25th percentile to £87,328 at the 75th, with a median near £67,746. The people furthest below it are almost always those who’ve stayed put: pay rises in place rarely track the external market.

Why the gap hides

Two reasons. First, internal raises anchor to last year’s you, not this year’s market. Second, you only see the roles you happen to scroll past — not the full set paying above your current number. The gap isn’t that you’re underqualified; it’s that you’re comparing against the wrong reference point.

The roles that close it

In this snapshot, 106 of the 200 postings for machine learning engineers paid at or above the median — each one, on its own, enough to erase a bottom-quartile gap. That's the whole game: don't apply everywhere, apply to the roles that already pay median or above, with a CV tailored to each one. This is also one of the more competitive titles for role-name drift — "ML Engineer," "AI Engineer," and "Applied Scientist" often describe near-identical day-to-day work at very different bands, so widen your search terms before you widen your salary target.

What to do this week

1. Anchor on the median for your exact role and level — not your current salary.
2. Filter to roles whose disclosed band closes the gap.
3. Tailor each application to the posting; send a few excellent ones, not fifty generic.
4. Track responses so you learn what actually converts.