Product descriptions

The gap between a parameter and a decision is small but it has to be crossed by someone. Either the customer does it, using knowledge they may not have, or the description does it for them. A rating that means the product survives rain but not a swimming pool is more useful to more people than the code that expresses the same fact. That is the whole job.

The first is conversion. A visitor who can tell whether a product suits them buys or leaves quickly. A visitor who cannot tell opens a competitor's listing to check, and often finishes the purchase there.

The second is pre-sale questions. Every message asking whether something fits, includes a part or works with a given system is a question the page failed to answer. Those messages cost support time and delay purchases past the moment of intent.

The third is returns. Faults and changed minds are outside anyone's control, but the share caused by mismatched expectations is genuinely addressable. Stating clearly who a product is not for costs a small number of sales and saves more in return shipping, restocking and support.

The practical approach is tiering. Bestsellers, new lines and high-margin products get individual descriptions with real attention to the buying decision. Everything else gets a consistent template: the same section order, the same level of detail, differences between models made visible. Comparable products should be comparable, and a catalogue where each item was written from scratch by a different person is harder to shop, not easier.

We settle that template on a small sample before writing at volume. It is much cheaper to argue about structure across ten descriptions than to discover the problem across four hundred.

Original copy is one of the few genuinely controllable differentiators at product level. We write around the phrasing people actually use when searching for the item, including the informal and misspelled variants, and structure the page so headings, attributes and meta fields carry that language properly.

We do not stuff keywords. A description that reads like it was written for an algorithm is a description a customer will not finish, and search engines have been penalising that pattern for a long time.

Customers are used to being sold to. A description that tells them one honest inconvenient thing is unusual enough to be trusted on everything else.

Writing that way depends on having something solid to write from, which is most of what we need from you: accurate specification data, product photos, and access to whoever fields customer questions. That last one is the most valuable input we get and the one companies most often forget to offer, because the questions people ask before buying are exactly the sections a good description needs.