This page is for the person the benchmark tables quietly shame. It works through what starting from zero at 50 actually builds by 67, in today's money, and why the answer is better than the panic suggests, without pretending it matches starting at 30.
A full new state pension pays £12,548 a year from 67, inflation-linked for life. Buying that much guaranteed, escalating income privately would cost several hundred thousand pounds. Someone at 50 "with nothing" very often has most of it already banked.
So the first move costs nothing: check the state pension forecast on gov.uk. It needs 35 qualifying years for the full amount, and at 50 there are 17 more working years to fill gaps. Where a gap year exists, a voluntary Class 3 top-up costs about £907 and typically buys around £358 a year of pension for life, arithmetic our top-ups page works in full. Everything in the next section stacks on top of this floor, which is precisely why a late start is recoverable at all: the plan does not have to build a whole retirement income, only the gap between £12,548 and the life you want.
The example: age 50, salary £40,000, nothing saved, retiring at the state pension age of 67. Growth of 2.5% a year above inflation, so every figure below is in today's spending power. The three tiers are total contributions including any employer money.
| Contribution tier | Per month | Pot at 67 (today's money) | Income at a 4% draw | Total with state pension |
|---|---|---|---|---|
| Auto-enrolment minimum (8% of salary) | £267 | £67,700 | £2,700 | £15,300 |
| Serious catch-up (15%) | £500 | £126,900 | £5,100 | £17,600 |
| Maximum effort (25%) | £833 | £211,600 | £8,500 | £21,000 |
Two honest readings. The auto-enrolment minimum alone does not rescue a late start: £15,300 a year is close to a floor-only retirement. But the 15% tier changes the story, and its real cost is smaller than it looks: £6,000 a year by salary sacrifice costs a basic-rate earner about £360 a month of take-home, because £140 of every £500 is tax and National Insurance that was never going to be kept anyway, and the employer's 3% arrives on top. The gap between the £267 row and the £500 row is the most valuable £233 a month in this table.
Information and guidance only. Not personal advice. The projections are arithmetic on stated assumptions, not forecasts: real returns above inflation are not guaranteed and a bad early decade changes outcomes. A 4% draw is a starting convention, not a promise the pot lasts, as our will-it-last page works through. Investment values can fall as well as rise and you may get back less than you put in. Verify decisions with a qualified FCA-authorised adviser before acting.
Three levers do the work in every late-start plan, and none of them is picking better funds.
Tax relief means the state funds a fifth to nearly half of every contribution, at your highest marginal rate, which makes the pension the default catch-up vehicle over an ISA for most late starters, worked £1 for £1 here. The employer match is the only guaranteed doubling available anywhere: a workplace scheme matching beyond the minimum outranks every other move, and unused annual allowance from the previous three tax years can carry forward for anyone able to contribute lump sums later, within this year's earnings. And retirement age is itself a lever: the table stops at 67, but each further year worked adds contributions, adds growth, removes a drawdown year, and can defer the state pension upward, a compounding worked on its own page. A late starter planning to 69 or 70 is not failing; they are using the strongest lever on the board.
Compounding needs time, and 17 years is enough for contributions to matter but not enough for growth to dominate: in the £500-a-month row, roughly £102,000 of the £127,000 pot is money paid in, not returns earned.
That has a liberating implication: the late starter's outcome depends far more on the savings rate, which they control, than on market returns, which they do not. It also has a sober one: the table's incomes fund a modest retirement, not an early or lavish one, and pretending otherwise is how bad plans get made. What the arithmetic genuinely supports: a floor-plus-comfort retirement from a standing start at 50 is achievable at savings rates ordinary households have managed when the mortgage ends or the children leave, which for many people is exactly what turns 50 into the first year saving becomes possible at all.
Methodology. Future values: monthly contributions compounded at 2.5% a year real (bundle defaults: 5% nominal growth, 2.5% inflation) for 204 months, expressed in today's money; contributions assumed level in real terms. Salary-sacrifice cost: 20% income tax plus 8% employee National Insurance on a £40,000 salary. Auto-enrolment minimum 8% of salary used as the entry tier (statutory minimum applies to a qualifying-earnings band, so the true statutory minimum is lower; 8% of full salary is shown as the round-number entry tier). State pension £12,548 (full new state pension, 35 qualifying years) and Class 3 rate £17.45 a week from rules bundle UK-2026.1.1. Drawdown income shown at a 4% initial rate. All arithmetic machine-checked before publication. Published 8 August 2026 · figures last verified 8 August 2026.
Sonuswealth is pre-launch. The app builds this plan from your actual salary, gaps and employer scheme, and keeps it current as rules change.