You're about five years out

Everything else is measured against this. It runs a survival curve alongside a portfolio simulation and tells you the odds of one outlasting the other.

Run it once with your real figures. Resist the urge to enter what you hope to spend rather than what you spend.

Step 1 assumed you spend the same amount, adjusted for inflation, every year for thirty years. Nobody does that.

Real retirement spending has a shape: higher in the first decade while you are well enough to travel and have suddenly acquired the time to spend money in, lower through the middle, and then potentially much higher at the very end if care is needed. Planning against a flat line gets the total roughly right and the timing badly wrong — and the timing is what determines whether you can afford the good decade.

Take the ramped first-year figure back to step 1 and re-run it. This is the single most common reason a plan looks worse on paper than it is.

This step produces no action. It changes how you think, which is why it comes before the steps that do.

Take one set of yearly returns and reorder it. The average is identical. If you are drawing an income, the order alone can decide whether the money lasts. Tick the "take no withdrawals" box and all three lines finish on exactly the same number, to the dollar — untick it and they diverge wildly. That contrast is the whole idea, and it is worth ten minutes because it explains why the next two steps matter.

This is the practical answer to step 3, and it is the highest-value action in these five years.

You cannot control whether a downturn arrives. You can control whether it forces you to sell. Holding two or three years of spending in cash and short bonds is not a good investment — it is deliberately a bad one — but it means a terrible first year is something you read about rather than something you pay for permanently.

How to build it without bad timing. Do not sell a third of your portfolio to cash tomorrow. Redirect new contributions, dividends and interest into the buffer over the remaining years, and let it fill up gradually. You have time; use it rather than making one large bet on today's price.

Two calculators, in this order, because they answer different halves of one question.

Calculator 09 asks what your plan can survive — the largest fall it can absorb while still funding your essentials — and compares that with what you are actually exposed to. Calculator 04 then builds a mix from your age, longevity, income needs and temperament, and stress-tests it.

The order matters. Capacity is arithmetic; temperament is you. Where the two disagree, the smaller one governs — a plan you abandon in the third bad month is worse than a more cautious plan you actually keep.

Claim your state pension as soon as you can, or wait for a larger one? It is one of the few genuinely irreversible decisions in retirement, and it is much better made now, on a quiet afternoon, than in the month you stop working.

How this page applies depends heavily on where you live — the UK cannot claim early at all, Australia has no deferral bonus, Canada needs a second run for OAS. Your country's page explains which version of the question you're actually facing.

Neither is exciting. Both are easier now than they will ever be again.

Calculator 08 scores the whole estate checklist rather than just asking whether you have a will. Most people have a will and assume the job is done; it usually isn't. Look at where the score loses points rather than at the score.

Calculator 16 asks whether you could explain what you own in two sentences. If you couldn't, someone else may have to one day, in a hurry, without your help. It also totals what the extra layers cost you each year — which is frequently the largest number on the page and the only certain one in your whole plan.

Consolidating accounts, closing dormant pots and writing down where everything is takes a weekend and removes a burden from people who will be in no state to work it out.