This is for recruiters, talent acquisition leads and hiring managers at PR and communications agencies who keep watching good account directors and senior account managers resign for in-house jobs. By the end you'll be able to find out why your people leave, rebuild the pay and progression offer, and put a stay conversation in front of the people most likely to go. One caution first. You'll see statistics about how many UK PR professionals are leaving agencies for in-house roles, but I can't source them, so I won't lean on them. What you can see in your own leaver data is more useful anyway.
1. Count who left, and where they went
Pull every resignation from the last two years. Record seniority, tenure, salary at exit, and the destination: in-house, another agency, freelance, out of comms altogether.
You're checking whether the leak is real and where it sits. Most agencies I speak to find it's concentrated at senior account manager and associate director level, three to six years in, which is exactly when someone has learned enough to be valuable and not yet been promoted.
How you'll know it worked: you can say in one sentence which grade is leaving and for whom. "Our SAMs go to tech scale-ups" is a finding. "People are leaving" isn't.
Beyond Tick Boxes: Diversity Recruitment Strategies That Actually Transform UK Workplaces
Master the Virtual Hot Seat: 7 Video Interview Techniques Recruiters Don't Tell You
AI Governance Careers Surge in 2026: Why This Role Will Define the Next Decade
2. Ask the leavers what they were really offered
Email people who left in the last year. Keep it short: "I'm not looking to win you back. I'd like to know what your new employer offered that we didn't, and what would have kept you." Say the first part and mean it.
Expect the answers to split. Some left for money. Plenty left for a four-day week, a single client to care about, or a boss who wasn't tied to billable targets. Money gets named first because it's the polite answer.
You've done it properly when you've got at least five candid replies and can sort them into pay, hours, progression and the work itself.
3. Benchmark pay against in-house, not other agencies
This is where agencies trip. You've been comparing yourselves with the agency down the road, but your leavers are comparing you with a corporate comms team that has a bonus scheme, a pension contribution and perhaps share options. Check the ONS earnings data for the broad picture, then get real offers from candidates who've turned you down.
As a rough market observation, in-house offers for a senior account manager equivalent often land above agency base. You won't always match that. You may not need to if the rest of the package is strong.
How to tell it's working: your salary bands are built from live offers, and someone senior signs them off at least twice a year.
4. Restructure what you pay for
A flat pay rise is the blunt tool. It costs the most and persuades the fewest. Try this instead:
- A retention bonus paid at 12 and 24 months, so staying has a visible price tag
- Account-linked variable pay, so winning and growing a client shows up in someone's pay packet
- A defined, dated route from SAM to associate director with the salary attached
The third one gets overlooked constantly. People will wait eighteen months for a promotion they can see. They won't wait for one nobody will put in writing.
Success looks like a SAM being able to tell you, without prompting, what they need to do to reach the next grade and what it pays.
5. Sell what in-house can't match
Agencies shouldn't try to out-benefit a big corporate. The real advantage is range: a PR professional here sees consumer, B2B, crisis and public affairs in the same year, while in-house they might spend three years on one product line. If you want to keep your senior staff, say so out loud, and back it with moves between accounts, secondments to clients, and proper training budgets.
The old pull of the move is also worth understanding from the other side. This piece on why the switch can feel like a demotion covers what leavers often discover after they land, which is useful ammunition for an honest stay conversation.
How you'll know: when you ask staff what they'd lose by leaving, they name something specific.
6. Hold stay conversations before the resignation
Book 30 minutes with each person in the at-risk grade. Not a review. Try wording like: "If a good in-house offer landed tomorrow, what would make it hard to say yes?" Then stop talking.
If someone has already been approached, you'll often sense it from the pause. Don't panic, and don't counter-offer on the spot. A rushed counter buys a few months.
The test: you come away with one concrete thing to fix for each person, and a date by which you'll have done it.
7. Build a bench so exits stop hurting
Retention won't hit 100%, and it shouldn't. Some people are right to leave. What hurts is losing someone with no replacement warm. Keep a live pipeline of agency and in-house candidates, and make your hiring process quick and transparent. Hiring PR professionals across the agency to in-house line is worth a read for the other direction of travel, because former agency staff who've done a stint in-house often return with sharper skills. Platforms like OHUB's recruitment tools for employers can help you keep that bench visible.
If a step goes wrong
Leavers won't reply in step 2. Ask someone they trusted, not HR, to send the message. Or use a short anonymous survey and accept you'll get less detail.
Finance blocks the pay changes in step 3 and 4. Cost one resignation instead: recruiter fees, the months of reduced billing, the handover, the client wobble. Put that figure next to the retention bonus. It usually shifts the conversation. For wider context on why the squeeze exists, see the 2026 PR talent crisis.
People still leave after you've fixed things. Check whether the cause is the line manager. Pay and promotion paths rarely rescue someone from a boss they've stopped trusting.
You don't know where to start. Do step 1 today.
This week: export your leaver list for the last two years and mark the destination against each name. Fifteen minutes, and you'll know which grade to protect first.
