The Diamond Is 7,900 Missing Juniors: UK's 3-Year Senior Gap
Open University's Sept 23 data says 19% of UK firms cut entry-level hiring. Here is the mid-level poach window that closes in 2028.
Open University's 2026 Business Barometer, published by CNBC on September 23, put a number on something UK engineering leaders had been muttering about all summer: 19% of 1,500 surveyed business leaders have cut entry-level recruitment, and 42% of those blame AI adoption. SHL's Lucy Beaumont called the new org chart a "diamond" instead of a pyramid: fewer juniors, a fat mid-level, and the same thin apex on top. That shape is a sourcing opportunity today and a crisis in 36 months, and most recruiters are only looking at half of it.
What Open University actually said, and why it matters in three years
AI is now the dominant force reshaping the shape of UK hiring, not just the volume, and the 2026 Business Barometer is the first dataset that puts hard numbers on it. Opinium ran the fieldwork in April and May 2026 across 1,500 UK business leaders. The headline figures, from the CNBC piece and Personnel Today's follow-up:
- 51% of UK business leaders say AI is changing how they hire.
- 19% have reduced entry-level recruitment in the past year.
- 42% of that 19% cite AI adoption as the direct reason.
- The 51% economy-wide figure rises to 59% at organisations with 250+ employees.
- 76% say economic uncertainty made recruitment or training harder, and 43% hired fewer staff overall.
The mechanism Beaumont describes is not complicated: "Those entry-level roles, they're your training ground. If you take that away, it has serious implications for every rung of the ladder." A junior cohort cut in 2026 does not appear as a senior in 2029. The diamond bulge sitting in the middle right now is the last full cohort. After that, promotion supply collapses.
The UK software pyramid is already thin. The diamond makes it worse.
Before the cut, the UK software talent pipeline was already running lean, with roughly 1.72 juniors per senior IC. A 19% haircut to the entry tier does not just dent the funnel, it inverts the shape entirely.
Here is what the UK software pipeline looks like right now, pulled from Refolk's index of professional profiles filtered on current title and country.
| Tier | Titles | UK profiles | Ratio to tier above |
|---|---|---|---|
| IC / Entry | Software Engineer | 41,544 | - |
| Mid-senior IC | Senior Software Engineer | 24,152 | 1.72 juniors per senior |
| Staff+ / Manager | Staff SWE, Principal SWE, Engineering Manager | 15,050 | 1.60 seniors per staff+ |
| Full pipeline | All three combined | 80,746 | 51% / 30% / 19% |
Apply Open University's 19% to the top row and the arithmetic gets ugly fast. That is roughly 7,900 UK software juniors missing from a single cohort year, which is nearly one-third of the current senior IC pool and half the entire staff+/manager population. There is no natural way to grow those bodies back in three years.
Why the 1.72 ratio is the number to memorise
The pyramid was already inverted enough that senior IC roles were the bottleneck, not junior ones. Every enterprise recruiter has felt this: a Senior Software Engineer requisition in London stays open longer than a graduate one. The diamond does not create that problem. It permanentises it. Sourcing mid-level engineers in 2026 is a defensive move against 2028's senior scarcity, not an opportunistic one.
The mid-level bulge is a poach window, not a plateau
The fat middle of the diamond exists right now because AI ate the juniors below without changing anything about the seniors above, but promotion pressure and retention risk will drain it within 18 months. Enterprises that lost their junior intake are staring at 24,152 UK Senior Software Engineers as their only remaining growth lever, and they will start bidding for them.
The top employers of UK Senior Software Engineers in Refolk's index tell you where the bulge actually sits:
- Monzo
- RVU
- Raytheon UK
- Spire
Notice who is not on that list. JPMorgan Chase, Lloyds Banking Group, and Deloitte Digital, which top the junior SWE employer list, do not show up at the senior tier. That is not because their seniors quit. It is because their seniors get poached out to scaleups like Monzo within a few years. The enterprises hired the training pipeline. The scaleups harvest it. Now the training pipeline is broken and the harvest has an expiry date.
The enterprises hired the training pipeline. The scaleups harvest it. Now the training pipeline is broken and the harvest has an expiry date.
Edinburgh and Cambridge over-index. London does not.
Refolk's index shows Edinburgh over-indexed at the senior IC tier relative to its share of junior roles, and Cambridge over-indexed at staff+. London holds 55%+ of the junior pool but a smaller share of the senior one. The practical read: the diamond bulge outside London is deeper and cheaper, because recruiter attention still defaults to the M25. Sourcing mid-level engineers in Edinburgh right now is the highest-leverage move on the board.
This is the exact gap Refolk closes for teams whose sourcers have been trained to type "London" into every filter. You describe the person in plain English ("senior backend engineer in Edinburgh, 4+ years, ex-scaleup, Go or Rust"), and get a ranked shortlist across LinkedIn, GitHub, and the open web without touching a Boolean string.
AI-native juniors are a different SKU. Source them on GitHub.
The juniors who did get hired in 2026 are not the same juniors the class of 2024 was. NACE reports 35% of entry-level jobs now require AI skills, and Handshake's Class of 2026 outlook shows the share of full-time postings mentioning AI has nearly doubled year over year to 4.2%. At London ad agency Catalyst, founder Tobias Green told CNBC that one AI-fluent junior now covers five client accounts that used to need several employees.
The 19% who got cut were largely non-AI-fluent grads. The minority who kept their offers are already being promoted into what used to be 2-year mid-level scope. That has two sourcing consequences:
- LinkedIn is a lagging indicator for this cohort. AI-native juniors have three months of "Software Engineer" on their profile and a two-year GitHub history that actually tells you what they can build.
- The old "2 to 3 years experience" filter is now a proxy for "hired before AI mattered." Some of your best mid-level candidates are 14 months into their first job.
If you are sourcing this slice, the query is "who has shipped agentic tooling on GitHub in the last 12 months, based in the UK, currently at a scaleup," not "Senior Software Engineer, 3+ years."
Segment by employer size or you will misread the whole market
The diamond is an enterprise problem, not an economy-wide one, and treating it as universal will send you sourcing the wrong pools. Open University's 51% AI-hiring-change figure rises to 59% at 250+ employee firms, and a separate Klarus survey of 500 UK and Irish mid-market leaders found 45% said AI was helping junior employees work better and faster, with almost a quarter saying it was creating new roles.
Two different markets, two different sourcing plays:
- Enterprise (250+ employees). The hollowing pool. JPMorgan Chase, Lloyds, Deloitte Digital, Google UK. These teams need to poach mids and lock down seniors before their internal promotion pipeline runs dry in 2028. Their own alumni are the target list.
- Mid-market and scaleup (under 250). The amplifying pool. Klarus's 45% who say AI is making juniors better. These teams are quietly training the next generation of mids on AI-native workflows, and enterprise recruiters will be trying to buy them wholesale in 2027.
The arbitrage is obvious once you see the two markets separately: scaleup juniors hired in 2026 will be the exact profile enterprise mid-level reqs demand in 2027 and 2028. Build that bench now, at scaleup salaries, before the enterprise bidding war starts.
A 3-year sourcing plan against the diamond
The right response to a shape change is a phased plan, not a panic. Here is the sequence that falls out of the numbers above.
- Now through Q2 2027: poach seniors. Focus on the 24,152 UK Senior Software Engineers, prioritising Edinburgh and Cambridge where competition is thinner. Target ex-Monzo, ex-Deliveroo, ex-Dojo profiles who have hit the 4-year mark.
- Q3 2026 through 2027: build the scaleup bench. Track AI-fluent juniors at sub-200-person companies via GitHub. Warm them for 18 months before enterprises notice.
- 2027 through 2028: lock retention on your own mids. Every mid-level engineer you hold onto is one you do not have to buy back at scarcity prices. Comp reviews now, not later.
- 2028 onward: plan a budget premium on staff+ hires. The 15,050 staff+/manager pool is the one that cannot be rebuilt from within.
The teams that will win this cycle are the ones treating September 23 not as a news item but as the start of a stopwatch. Sourcing tools that require a Boolean string per query will not keep up with a plan that has to run across three tiers, two employer segments, and four cities simultaneously.
FAQ
Is the "diamond" shape real across the whole economy or just at big firms?
It is concentrated at scale. Open University's 51% AI-hiring-change figure rises to 59% at 250+ employee firms, and Klarus's mid-market survey found 45% of leaders say AI is actually helping juniors work better, with almost a quarter saying it is creating new roles. Treat the diamond as an enterprise phenomenon and the mid-market as an amplification story. Sourcing strategy has to segment by employer size or you will misread which pool is shrinking and which is growing.
Which UK cities offer the best mid-level poach opportunity right now?
Edinburgh and Cambridge, based on Refolk's index. Edinburgh over-indexes at the Senior Software Engineer tier relative to its share of junior roles, and Cambridge over-indexes at staff+ and principal. London still holds the largest raw pool but also the highest recruiter density, so competition for each candidate is intense. The junior developer market in 2026 is a London story. The senior market is not.
How do I source AI-native juniors if the old "2 to 3 years experience" filter is now misleading?
Move the query off LinkedIn tenure and onto GitHub activity. NACE data shows 35% of entry-level jobs now require AI skills, and the survivors of the 19% cut are the ones with real shipped work on LLM tooling, agents, or evals in the last 12 months. Ask for "UK engineers with active GitHub contributions to AI tooling in the last year, currently at a startup," rather than filtering on years of experience.
What is the single number I should remember from this data?
1.72 juniors per senior IC in the UK software pipeline today, before the 19% cut lands in the promotion funnel. That ratio was already historically thin, which is why cutting entry-level recruitment by nearly a fifth removes roughly 7,900 juniors per cohort and breaks the natural supply into staff+ roles by 2028. Every sourcing decision you make for the next 18 months should be evaluated against that number.
Try it on the search you came here for
Stop building boolean strings. Just describe the person.
Type one sentence. I plan the search, read GitHub, public LinkedIn and Crunchbase records, and the open web as it is right now, and hand back a ranked list with the reason next to every name.
01Describe them
One plain sentence. Role, city, stack, stage, whatever matters to you.
02I read the web live
GitHub, public LinkedIn and Crunchbase records, the open web. Not a database that went stale last quarter.
03You read the shortlist
Ranked, with the reasoning under every name. Open a profile, ask a follow-up, narrow it down.
- Staff backend engineers in NYC who shipped Rust in production
- Series A fintechs in SF under 50 people, growing headcount this year
- Maintainers of fast-growing Rust web frameworks on GitHub
- No boolean, no filters, no seat to buy. One box.
- Read at search time, so a profile updated yesterday counts today.
- Every step visible as it runs, every name with its reason.
500 free credits on sign-up. No card, no demo call. See real searches.