How PickyPeacock measures specs and prices

Every spec on the site is read from the manufacturer's own datasheet and entered by hand. We do not pull specifications from merchant listings, where a figure is often rounded, marketing-led, or simply wrong. A datasheet is the closest thing to a primary source a component has.

A spec is only as current as the datasheet it came from. When a manufacturer revises a product without renaming it, the page reflects what the datasheet said when it was read, not a silent later change.

Prices are captured repeatedly over time, not read once. So a current price is measured against what the product has actually been selling for, rather than against a list price a merchant chose to print beside it.

This is why the site publishes no discount percentage of its own. A percentage is worth only what the number on the other side of it is worth: measured against a seller's list price it describes the seller, and measured against a product's own recorded history it describes the market. We show the record that comparison needs — what a product has sold for, when it was captured, and across how many merchants — and leave the percentage to you.

Where the site surfaces a price drop — inside a guide, beside the products that guide is about — that is an output of the tracking, not the reason the site exists. Such a drop is a product now priced low against its own recorded history, ordered by price and nothing else. It is never a curated list of promotions, and nothing is surfaced because a merchant paid for placement.

When two products are compared, the engine reports dominance, not a verdict. One product is shown beating another only when it is no worse on every axis being compared and better on at least one. Where that is not true — where each wins on something the other loses — the comparison says so and stops there.

The engine computes no overall score and names no single winner, because the trade-off between, say, capacity and speed is yours to weigh, not ours. A number that collapsed those axes into one ranking would be inventing a judgement the measurements do not contain.

Some product pages carry a short summary of what owners and reviewers commonly say. It is drawn from a named set of sources — retailer reviews, forum threads and professional reviews — that we read by hand. Each point is our own paraphrase of a theme those sources state, never a quotation, and a source is counted only where it names the exact model. The sources sit behind the summary on the page, so the claim can be checked rather than taken on trust.

This is not a rating the site computes. We publish no score of our own and no number the sources do not carry; where a source states a rating, we show that source's figure with its count. Where a product has no summary, it is because we have not yet read enough sources that name it exactly — a gap, reported as a gap, not filled with a guess.

We do not rank products by a figure we invented, and we do not call anything a best buy. We do not take payment to place a product higher or to surface it, and nothing on the site is ordered by what pays us. Where the site cannot measure something, it says nothing rather than guessing — an unknown is reported as an unknown, not dressed up as a recommendation.