It's the most-asked question in personal finance and it has the least useful standard answer. Here's why the big round numbers mislead, and how to get to yours.
You'll see figures like £500,000, or £1m, or “25 times your spending”. They're not wrong exactly. They're just answering a different question from the one you asked.
The number you need depends on four things, and only one of them is the pot: what you spend, when you stop, what other income you have, and how long you live. Change any one and the answer moves by hundreds of thousands. Two people with identical £400,000 pots can be comfortable and broke respectively.
“How much do I need?” is really “can I stop when I want to, and not run out?” That has an actual answer, and it's specific to you.
Work backwards. What does your life cost now, and what will it cost when you stop?
Most people's spending falls somewhat at retirement — commuting goes, the mortgage is often gone, children have usually stopped being expensive. But it falls by much less than people expect, because time you used to spend at work is now time available to spend money.
Guessing. Ask someone what they spend and they'll name a figure confidently, and they'll typically be out by several hundred a month. That error compounds through every year of a thirty-year projection.
The free Healthcheck connects your bank read-only and sorts the last few months automatically. It's the difference between a plan built on a fact and one built on a hope.
Take someone who wants £3,000 a month after tax from 62, in today's money. That's £36,000 a year.
The full new State Pension is roughly £12,000 a year — but not until 67 or 68. So from 62 to 67 they need the whole £36,000 from savings. After that, about £24,000.
Five years at £36,000 is £180,000 before the State Pension arrives. That gap is the single most expensive part of retiring early, and it's the bit people forget entirely.
£24,000 a year from 67 to 100 is thirty-three years. At 3% real growth, funding that needs roughly £530,000 at 67.
Around £710,000 at 62 — but it's not really one number, because tax, the order of returns and how long they actually live all move it. That's why a projection beats a rule of thumb.
Retiring at 62 instead of 67 doesn't cost five years of income. It costs five years of income plus five years of growth you no longer get plus five more years of drawdown. That's why the stop date is usually the most powerful number in the whole plan.
Comes from the 4% withdrawal research. A reasonable starting point, but it ignores the State Pension entirely — which for most people is the single largest inflation-linked income they'll ever have. Include it and the number drops a lot.
A hangover from final salary pensions. It anchors to what you earned rather than what you spend, and those can be wildly different. Someone earning £80,000 and living on £40,000 needs far less than the rule suggests.
Figures like this come from national research and are genuinely useful for orientation. They are not useful as your target, because they describe a standardised basket rather than your life, your mortgage or your children.
Your real spending, your real other income, your real stop date, run forward properly. It's more work and it's the only version that answers the question you asked.
When people want to improve the answer, they almost always reach for the weakest lever first. Here they are ranked by how much they actually move the number.
Comfortably the most powerful, and the one people treat as fixed. Every extra year is a year of earning, a year of contributing, a year of growth and one fewer year of drawing — four effects pulling the same way. Two more years often does more than a decade of saving harder.
Second, and it works at both ends. £300 a month less in retirement is roughly £90,000 less capital needed over thirty years. £300 a month more saved between now and then compounds on top. Finding money you were spending without noticing does both jobs at once.
State Pension gaps you can still fill, a deferred scheme from an old job, a rental, a business. Each guaranteed pound reduces what the pot has to produce, and people routinely leave whole pensions out because they have forgotten the job.
Last, and it is the one everybody starts with. Moving an assumption from 3% to 4% changes a projection far less than moving the stop date by two years, and unlike the other three it is not something you control. Chasing it is the least reliable way to fix a plan.
The first three are decisions. The fourth is a hope. A plan that gets to a comfortable answer by assuming better returns has not been improved — it has been rewritten to be more optimistic, which is a different thing and rather easier to do.
The most useful floor is whatever covers your non-negotiable costs from guaranteed income. Housing, food, energy, insurance, council tax: the things that carry on regardless of what markets do. If the State Pension and any defined benefit scheme between them cover that list, everything else you own is buying choices rather than covering survival, and a bad year in the markets changes your holidays instead of your heating. That is a far steadier position to plan from than a large pot doing all of the work on its own. Work the floor out first, in pounds per month, before you think about the pot at all. For a lot of people it comes to less than they feared and the State Pension covers more of it than they expected.
More than almost anything else in the calculation. Roughly £12,000 a year, rising annually, guaranteed for life, with no investment risk attached and no possibility of running out. For a couple who both qualify in full that is around £24,000 between them. To generate the same income from a pot at a cautious sustainable withdrawal rate you would need something in the region of £300,000 each, and even then it would not carry the guarantee. People routinely describe it as not much and then build a retirement around a pot worth less than the thing they dismissed. Check your own forecast on gov.uk rather than assuming the full amount, because a great many people have gaps in their record and some can still be filled.
Then two separate problems arrive at once. The first is access: normal minimum pension age is 55 and rises to 57 in April 2028, so if you were born after 6 April 1973 there is a stretch where you have stopped working and cannot touch a pension at all, which has to be funded from ISAs, savings or other income. The second is the bridge. Retiring at 55 rather than 62 means seven extra years of drawing before the State Pension arrives, and those are the most expensive years in any plan because nothing else is contributing to them. Between them, going at 55 rather than 62 can easily need a third to a half more capital, most of it funding the gap rather than the retirement itself.
Only if you genuinely intend to sell or downsize, and then only for the amount you would actually free up. That means the sale price less the cost of wherever you move to, less stamp duty, agents, legal fees and the move itself, which together take a larger bite than most people allow for. A £400,000 house becoming a £300,000 house does not release £100,000. Plenty of plans quietly rely on equity the owner has no real intention of releasing, and a plan that only works if you leave a home you love is not one you will follow. If downsizing is genuinely the intention, model it at a realistic figure and a realistic date. If it is not, leave it out.
Start with the free projection on this site. It takes about two minutes, it works in today's money, and it gives you both the shape of the answer and the age your money runs out. For a lot of people that is enough on its own. Before you run it, get two facts straight: what you actually spend, which almost nobody guesses correctly, and what your State Pension forecast says, which is free on gov.uk and takes five minutes. If several parts of the picture are pulling against each other, a business nobody has valued, a stop date you are unsure of, a defined benefit scheme with a penalty for taking it early, that is what the Financial Plan is for.
The projection runs in your browser in about a minute. Nothing is sent anywhere and no email is needed to see the answer.
Thirty minutes with a planner to work out whether you need a Plan at all. Some people leave that call having been told they don't.
One written document — every year to 100 with the tax done properly, what breaks it priced in years, and the ninety days after. £1,500.