The short version: You negotiate a raise by proving your market value with numbers before you ask, timing the conversation around budget cycles rather than your own frustration, and naming a specific figure instead of hoping your manager offers a fair one. Most people lose this negotiation weeks before the meeting even happens, because they walk in with feelings instead of evidence.
Something nobody tells you before the meeting
Here's a fact that annoyed me for years before I understood it: your raise was probably decided before your appraisal, not during it. Most companies set their pay review budget months in advance, handed down from finance as a single percentage, often 3%, sometimes 5%, rarely more. Your manager isn't sitting there weighing your worth against thin air. They're working out how to split a pot that already has a lid on it. That doesn't mean asking is pointless. It means your actual job is to make yourself the exception to the pot, not to win a debate about fairness.
I learned this the expensive way in my twenties, working at a marketing agency in London. I'd had a strong year, brought in two new accounts, stayed late through a nightmare product launch, and I went in with a list of everything I'd done and asked my manager for a raise. He said "we'll see", which meant no, and I came out three months later with 2%. A colleague on my team, same level, asked for a raise the same round and walked out with 15%. The difference wasn't effort. He'd had a competing offer in hand from another agency, a real one, with a real number on it. He didn't even bring it up directly, he just mentioned he'd had "an interesting conversation" the week before. That was the whole negotiation. I've never forgotten how little it had to do with who worked harder.
Do the homework before you say a single word
The conversation is the easy part. The research is where most people quit early, and it's the part that decides the outcome.
- Pull three real market figures for your exact title, level and location, not a national average. Use Glassdoor, LinkedIn's salary tool, Reed or Totaljobs listings for equivalent roles, and if you want an official baseline, the ONS Annual Survey of Hours and Earnings gives median UK pay by occupation.
- Write down every concrete result from the last 12 months with a number attached. Not "improved processes", but "cut invoice turnaround from 9 days to 3" or "grew the email list from 4,000 to 11,000". If you don't have numbers, spend two weeks gathering them before you ask for anything.
- Decide your target figure and ask for slightly more than you'll settle for. If you want an 8% increase, open at 12 to 15%. Almost nobody gets offered their first number in full.
If you're unsure how to frame any of this into an actual sentence, I've written a longer piece on what to say when negotiating your salary with the exact phrasing that gets a better response than "I feel like I deserve more".
Pick your moment, not your mood
Do not ask for a raise the week after a bad quarter, straight after a round of layoffs, or during the actual appraisal cycle when the budget is already fixed. The better windows are right after you've closed a visible win, or a few months before the company sets its next budget, which for most UK businesses running an April fiscal year means asking in autumn, not spring. Ask for a dedicated slot in the diary rather than dropping it into a normal one-to-one. "Can we set aside 30 minutes next week to talk about my role and compensation" gives your manager time to prepare a real answer instead of a stalling one.
What to say once you're in the room
Lead with the market data, then your results, then the number. Something close to: "I've looked at market rate for this role in our industry, and based on what I've delivered this year, including X and Y, I'd like us to move my salary to £Z." Then stop talking. This is the hardest part of the whole thing and it's not the number, it's the silence after you say it. Most people fill the gap with an apology or a softener, and that's exactly where the negotiation gets weaker. Say the number and let it sit.
Follow up in writing whatever the answer is
Whether you get a yes, a no, or a "let me think about it", send a short email the same day summarising what was discussed and restating your figure and evidence. It protects you if the conversation gets forgotten in a busy week, and it puts the request on record for the next review. If you're not sure how to write that email without sounding pushy, there's a full breakdown on writing a salary negotiation email that gets a positive response that covers tone as well as content.
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
If the answer is no
A flat no is rarely the end of the useful part of the conversation. Ask directly: "What would need to be true in six months for this to be a yes?" Get the answer in writing and set a follow-up date, ideally 90 to 180 days out, not "sometime next year". If the salary itself can't move, other things often can: title, a training budget, extra remote days, a signing bonus disguised as a retention payment, or an earlier review date. Henry Ford figured out over a century ago that paying workers more than the going rate cut staff turnover so sharply it paid for itself within months, there's more on that in the business lessons from Henry Ford, and it's still true today: a company that won't invest in keeping good people is telling you exactly what it values, and it isn't you.
When staying isn't the right maths
If you've asked twice, brought clean evidence both times, and been told no with no real path forward, that's not a negotiation problem anymore, that's information. At that point it's worth doing the actual sums on alternatives rather than asking a third time and hoping for a different result. I've written before about using AI to price small business services without racing to the bottom, and the same logic applies to your own time: if you keep accepting below-market pay because asking feels risky, you're doing the race to the bottom to yourself. Some people cover the gap with a side income while they job hunt, and it's worth knowing what that's realistically worth: I've broken down what tutoring online part time in the UK pays, which for most people sits between £15 and £35 an hour depending on subject and level, a useful bridge rather than a permanent answer.
Frequently asked questions
How much of a raise should I ask for?
Open 12 to 15% above your minimum acceptable figure. Employers rarely offer their first number in full, so asking for exactly what you'd settle for usually means settling for less than that.
Is it a bad idea to mention another job offer?
It's the single strongest lever there is, but only if it's real. A vague reference to being "approached" can work if it's true, naming a company that never made you an offer is a risk that can badly damage trust if it ever comes out.
How often can I reasonably ask for a raise?
Once a year is standard unless your role or responsibilities have changed significantly, in which case asking sooner with clear evidence of the change is reasonable. Asking more than once a year without new evidence tends to read as impatience rather than confidence.
What if my manager says there's no budget at all?
Ask what would need to be true for that to change and get a follow-up date in writing. If the answer is never, that's useful information about the company, not a reason to stop asking anywhere.
