From "where do I even start?" to a clear plan — in three steps

Every option is scored out of 100. Nothing is weighted by who the provider is, and no provider can pay to appear or to rank higher — the site takes no commission from any institution it lists. The score is built from five components, and each product card shows you its own breakdown.

The five maxima below add to 100, but the figure you see is held between 10 and 99. The ceiling is there because nothing on a list of financial products deserves a perfect mark, and the floor because a score of zero reads as a verdict on the product rather than on its fit with you. So the arithmetic will not always match what the card shows at the extremes — which is worth knowing if you are checking our working, and you should be.

Affordability — up to 25 points

Jointly the heaviest component, alongside risk match, because an option you cannot buy is not an option. A product priced comfortably within your capital scores full marks. One whose minimum would consume most of what you have scores partially, because committing everything to a single holding is a concentration risk rather than a plan. One you cannot meet at all scores zero and drops down the list rather than being hidden — you are shown what you would unlock at a higher amount.

Risk match — up to 25 points

Scored against the risk tolerance you selected, not against an abstract notion of safety. If you said you cannot tolerate losses, low-risk products score full marks and high-risk ones score zero. If you said you are willing to take more risk for higher returns, the ordering inverts — a savings account is a poor fit for a goal it cannot reach, and the model treats that as a mismatch rather than as prudence.

Horizon match — up to 20 points

Scored against how long you said you can leave the money, using the product's risk level as the measure — because what a short horizon cannot absorb is volatility. Over a few months, a high-risk holding scores close to nothing: you may have to sell at whatever price the market happens to be offering on the day you need the cash. Over five years or more the ordering reverses, and an instant-access savings account scores below a growth holding, because you are paying for liquidity you said you do not need. For horizons under three months, liquidity is read directly as well, since there a product that cannot be exited quickly fails on availability rather than on price.

Worth knowing what this does not use: the lock-in period shown on each card is not an input to the score. It is displayed because it decides whether you can live with the product, and you should read it — but the score reaches that judgement through risk and liquidity rather than through the stated term. We would rather say so than let the field's presence on the card imply it was counted.

Goal match — up to 20 points

Weighs what the product does against what you said you want. Capital preservation favours sovereign-backed and low-risk instruments. Income favours products that actually distribute cash rather than accumulate it. Growth is scored on the return available relative to the rest of your market, so the standard adjusts to where you live.

Relative yield — up to 10 points

Deliberately the smallest component. Return is scored relative to the best net yield available in your own market rather than against a fixed global threshold, because 13% is a poor return in a high-inflation economy and an exceptional one in a low-inflation economy, and a fixed divisor cannot tell those apart.

Fees are subtracted from a product's yield before this comparison, so a high headline rate carrying high charges is scored on what it actually pays. Fees are not then penalised a second time, which would charge the same product twice for one cost.

Why yield is weighted lowest

Ten points of a hundred looks dismissive of the thing most people think they are shopping for. It is deliberate. A high return you cannot access when you need the money, or cannot stomach the volatility of, or cannot meet the minimum for, is not a good outcome — it is a mis-sold one. The model is built to put suitability first and let return break ties among options that already fit.

Product details — yields, minimums, fees, lock-in periods and regulatory status — are compiled from published central bank figures and the providers' own published terms. This is a periodically updated reference set, not a live market feed, and each card carries the date its data was last checked.

That distinction is stated plainly rather than buried, because it changes how you should use the output: treat a quoted rate as a starting point for your own check with the provider, not as today's price. Rates move, and a comparison engine that implied otherwise would be misleading you about its own precision.

Economic context — policy rates, inflation — is applied per country, which is why the same capital produces different recommendations in different markets. Where we hold verified local products for your market they are compared directly; elsewhere you are shown the general categories available in most countries with your local inflation applied, labelled as such.

It is not advice. The model knows four things about you, all of which you typed in a minute. It does not know your debts, your tax position, your dependants, your job security, your existing holdings or your plans — all of which a licensed adviser would ask about before saying anything.

It does not execute anything. We never hold your money, open accounts or place orders. You act through your own bank or broker.

And it will tell you not to invest when that is the right answer. If your capital is below the useful threshold for your market, or local fees would consume the return, the result says so instead of ranking something unsuitable at the top to look useful.