HOW · Method
How we score
"Lowest price" means the lowest price you will actually pay
When we say Lowest price or Cheapest, we mean the lowest total price: the item price plus delivery, after any free-delivery offer the retailer runs (for example "free delivery over £30"). Only retailers that show the item in stock count. If two retailers charge exactly the same total, we show the one with the stronger trust rating first. How much we might earn never affects which offer gets this badge.
When a retailer doesn't tell us the delivery cost
Some retailers (Amazon's product data, for example) don't tell us what delivery costs. We never treat an unknown delivery cost as free. Those prices are shown as "£249.00 + delivery", and they can't take the Lowest price badge from a retailer whose full price including delivery we do know. Only when no retailer states its delivery cost do we compare item prices, and then the page says delivery is extra.
"Our pick" — and why it may not be the cheapest
We earn a commission from some retailers when you buy through our links. Sometimes the offer we earn more from is a few pence or pounds dearer than the cheapest. We will only mark such an offer as Our pick when it is within £2.00 or within 2.5 % of the lowest total price (whichever allows more), and only when it is in stock. When this happens:
- the badge says so ("within £X of the lowest price; we may earn more from this retailer"), and
- the cheapest offer is always shown right next to it, so you can choose.
If no offer meets that test, Our pick is simply the cheapest offer.
Some links reach the retailer through a sub-affiliate service (Skimlinks or Sovrn Commerce) that keeps part of the commission. We count only the part we would actually receive, so such a link is less likely to become Our pick — and it never affects which offer is the lowest price.
The HOT score (0–100): how good is this deal right now?
A HOT score rewards real, recent price drops — not inflated "was" prices.
| What we look at | Up to |
|---|---|
| How far below its own 90-day average price it is | 45 points |
| How far below the manufacturer's RRP it is | 20 points |
| How recently the price dropped (while still low) | 20 points |
| How many retailers have it in stock | 15 points |
A price 30 % or more below the 90-day average earns the full 45 points; 50 % or more off RRP earns the full 20. A drop earns recency points for 14 days and only while the price is still below its 90-day average. Eight or more retailers in stock earn the full 15. A price that is above its usual level never earns drop points.
The WHAT score: how good a buy is this among comparable products?
The WHAT score compares a product with others in the same category and the same price tier, using published specifications and retailer data only. We do not run hands-on tests, and we never invent ratings. It is shown on a scale from 4.5 to 9.8 (one decimal place).
First we sort each category into three price tiers by its current lowest price: the cheapest third is Budget, the middle third Mid-range and the dearest third Flagship. A product is compared with the products in its own tier — so a flagship is judged against other flagships, not marked down for costing more than a budget model. When a tier has fewer than 3 products, we compare with the whole category instead. With fewer than 2 products in a category there is nothing to compare, so we show no WHAT score.
Scores are worked out separately for each storefront (UK, US): a product is compared only with the products on sale in that region, using that region's prices, retailers and price history (the specifications are the same everywhere). So the same headphones can score slightly differently in the UK and the US.
For each of four things we work out the product's percentile among those products (how it ranks, from 0 = bottom to 1 = top; equal values share a rank):
| What we look at | Weight |
|---|---|
| Specifications — how each of its key specs (e.g. battery life, weight) ranks, averaged | 45 % |
| Price vs its own history — how far today's price is below its own 90-day average (no history = middle) | 25 % |
| Retailer trust — the trust rating of the retailer selling it at the lowest price | 15 % |
| Choice of retailers — how many retailers have it in stock | 15 % |
The weighted result (between 0 and 1) becomes the score: WHAT score = 6.2 + 3.6 × result, kept between 4.5 and 9.8. Because every part is a ranking among similar products, a product in the middle of its group scores about 8.0 — the score says how a product compares with its peers, not that it is good or bad in absolute terms.
For each specification we know whether higher is better (battery life) or lower is better (weight). Specifications without a clear "better" direction (colour, port type) are not scored.
| Badge | Score | Colour |
|---|---|---|
| Exceptional | 9.0 or more | green |
| Great buy | 8.0 – 8.9 | green |
| Good | 6.5 – 7.9 | amber |
| Fair | below 6.5 | neutral |
A score below 8.0 is never shown in green.
The deal meter
| Verdict | When |
|---|---|
| Great time to buy | at least 10 % below the 90-day average, or within 2 % of the lowest price in 90 days |
| Wait for a drop | at least 5 % above the 90-day average |
| Fair price | anything in between |
We need at least 14 days of price history before we show a verdict. Until then the meter is hidden — we would rather say nothing than guess. Some retailers (for example Amazon) do not allow their prices to be stored as history, so their prices are not part of the 90-day figures.
Your privacy when you click "View deal"
When you click through to a retailer we record that a click happened, with a random reference so the network can tell us about a sale. We do not store your IP address or your browser's user-agent string. To spot automated click loops we keep a one-way scrambled fingerprint of the browser type that changes every day and cannot be turned back into the original. Search-engine crawlers and automated tools still reach the retailer, but are not counted.