Where our data comes from

Everything on this site is built from two things: publicly published retailer product data, and the measurements we derive from it over time. There is no third source.

Retailer product data

Most online stores publish their catalogue in a machine-readable form so their own product pages can display it. We read exactly what a visitor’s browser receives, and keep only factual fields: product name, link, item number and identifiers where published, price, sale and regular price, stock state, category, brand and attributes.

We do not republish retailer product descriptions. Descriptions are hashed so we can tell when copy changed; the text itself is discarded and never appears on this site.

Product images

Product photographs come from the retailer whose offer they sit beside. They are shown to help you identify the product, and every one of them appears on a page that sends traffic to that retailer.

• We do not download or re-host them — they load directly from the retailer’s own image host.
• We request a reduced size rather than the full-resolution original.
• We claim no rights in them and never present them as ours.
• Where a retailer publishes no image, we generate our own data visual instead.

If you are a retailer and would rather we did not use your images, tell us and we will stop using yours the same day — one message is enough, and no explanation is required. Your products stay listed; we simply show our own visual instead.

Data we create ourselves

• Canonical product identity — the decision that two listings are the same item.
• Price history: every observation with its source and timestamp.
• Daily market low, median, interquartile range and high.
• Real saving versus the price a product normally sells for.
• Deal Score, price percentile, price stability and store spread.
• Freshness state and source health.

This derived layer is the actual product. It is why a page here answers questions a retailer’s own product page cannot.

How we read a retailer’s catalogue

• We check and follow each retailer’s stated access rules before we read anything, and re-check them regularly.
• We read one page at a time from any one store, with a deliberate pause between requests and a firm hourly ceiling.
• If a store asks us to slow down, we slow down; if it keeps failing, we stop and come back later.
• We identify ourselves clearly on every request, with a link explaining who we are.

What we will not do

• Work around any access control a retailer has put in place.
• Read a catalogue from a store that has told us not to.
• Buy or use collected data of uncertain provenance.
• Copy retailer marketing copy, or re-host their images on our own servers.
• Fabricate a price, a trend, a review or a rating.
• Keep using a retailer’s images after they have asked us not to.

Retailers currently active

If you are a retailer

If you would prefer we read your catalogue less often, or not at all, tell us and we will comply. You can also refuse us in your site’s access rules, and we will stop at our next check without needing to be asked twice.