On page optimization is the work of making a single page answer one question completely, in words a buyer wants to read and a machine can quote. On a store that means titles, headings, product descriptions and the order information appears in.
Most product pages are a photo, a price, a bullet list from the manufacturer and a buy button. They are transaction screens, not pages. There is nothing on them a search engine can use to decide they are better than the twelve other stores selling the same item with the same manufacturer text.
The stores that win are the ones that treat a product page as something worth writing.
Why Does Manufacturer Copy Hurt You?
Because every store selling that product has the same words. A search engine choosing between twelve identical pages will pick the biggest store, and that is rarely you.
Using the supplier’s description is the default because it is free and it is there. The cost is invisible until you understand what it does. Your page carries no information that any other page lacks, so there is no reason to prefer it.
Rewriting does not mean writing more. It means writing what the supplier did not. Who this is right for and who it is wrong for. What it is usually compared against and how it differs. What people get wrong when they buy it. What to check before ordering. How it behaves after six months.
That information exists inside your business already, in your support inbox, your returns data and the heads of the people who sell it. Getting it onto the page is most of the work.
What Do We Change on a Page and in What Order?
Title first, then the opening answer, then headings, then body, then the supporting detail. The order matters because the early elements do most of the work.
The opening answer is the highest value element on the page and the one almost every store leaves out. Two sentences in the right place regularly outperform a thousand words in the wrong one.
How Do You Write for Machines Without Writing Badly?
By writing more plainly, not less naturally. Machines extract text they can parse without guessing, which is the same text a hurried human finds easiest to read.
There is a persistent belief that optimising for search means degrading the writing. It came from an era when it was true, when pages were stuffed with repeated phrases that nobody would say aloud.
That approach now actively fails. What works is short sentences, a clear subject in each one, questions as headings, direct answers under them, and specific facts instead of adjectives. A page written that way reads better to a person and is far easier for a machine to quote.
The one place the two audiences differ is repetition. A person is happy for you to say the product name once and then say it. A machine benefits from the name appearing where it matters, in the title, the opening answer and the key headings. That is a small adjustment, not a compromise.
What About Pages With Thousands of Products?
You cannot hand write ten thousand product pages and you should not try. Above a few hundred products the work becomes templates, rules and data quality.
The approach splits in two. Your top pages, usually a few hundred products and every collection page, get written individually because they carry most of the revenue. Everything else gets a template built from the product data you already hold.
A good template is not a mail merge. It builds a genuinely different page for each product from its own attributes, so a page reads specifically about that item rather than filling blanks in a sentence. Doing that well depends entirely on the quality of your product data, which is why data work and writing work belong together.
The practical result is that improving one field across your catalogue can improve ten thousand pages at once. That multiplying effect is the whole reason large catalogues are worth working on.
How Long Until a Rewritten Page Moves?
Usually three to twelve weeks, depending on how established the page already is and how competitive the phrase is. Pages that already rank somewhere move faster than pages starting from nothing.
A page sitting at position eleven with weak copy often moves within a month of being rewritten, because search engines already consider it relevant and only need a reason to prefer it.
A page with no history takes longer, because it has to be found, evaluated and tested before it settles anywhere. Three months is a fair expectation and it can be longer in a crowded category.
This is why we rewrite in revenue order rather than alphabetically. The pages most likely to move quickly and matter most are the ones that pay for the rest of the work.