Metadata entry

What metadata entry is

Metadata entry is the job of typing the hidden fields into your website, page by page, until none of them are empty. The page title that shows in the browser tab and in a results list. The description that sits under it. The address slug. The alt text on every image. The category and the excerpt where the system has them. It is data entry, and I am not going to dress it up as anything else.

It is not the writing of those fields. Deciding what a title should say, for which page, in how many characters, is title tags and meta descriptions. Deciding what a photograph should be described as is image alt text. Both of those produce a spreadsheet. This service is the person who takes that spreadsheet and puts every line of it where it belongs.

When you need one

The usual case: somebody has already done the thinking. You have a spreadsheet of titles and descriptions from a writer or a consultant, delivered three months ago, and it is still a spreadsheet. Sixty rows. Nobody has ten hours to type them in, and the person who does have the time is not confident about which field is which.

The second case: nothing has been written, but the site is running on whatever the system generated, which on most sites means the same eight words repeated across forty pages and no image description anywhere. Here the honest answer is that you need the writing first, then this.

You do not need this for a handful of pages. If there are five, open the editor and do it; it is twenty minutes. And if what you actually want is the whole page entered rather than its fields topped up, CMS content entry covers metadata as part of the price.

What gets filled in

• The page title. Typed into the title field rather than left to inherit the H1, checked for length so it is not cut off, and unique on every page.
• The description. The paragraph shown under the title in a listing, in the field the system actually reads, which on many sites is not the obvious one.
• The address slug. Short, readable, lowercase, hyphenated, with the date and the trailing numbers taken out where the system allows it without breaking the link.
• Alt text on every image. One line per image describing what it shows. Decorative images get an empty attribute on purpose, which is different from a missing one.
• The rest of the fields. Excerpt, category, author, and the social preview title and image, which are separate fields on most systems and are almost always empty.

Two things get flagged rather than typed. Anything longer than the field allows, so you can shorten it rather than have the system cut it mid-word. And any page where the supplied text plainly belongs to a different page, which happens in every batch of sixty.

What you get

The fields, filled in. Every page in the agreed list, done in the live system or on a staging copy, with the change saved and the page re-checked afterwards to confirm nothing else moved.

A completed sheet. Your original spreadsheet, returned with a column added: done, flagged, or not applicable, with a note against every flag. It is the record that says the job finished.

A blanks list. Every field that is still empty when I stop, and why. Usually images nobody could identify, or pages where the description was never written. It is a short, specific list of what to commission next.

A worked example

An illustration, not a client. A specialist tool supplier has a shop of four hundred product pages, sixty article pages and no descriptions anywhere. A consultant has written titles and descriptions for the sixty articles and the twenty product categories. The product pages themselves are left alone for now, which is the right call.

The eighty pages are done in four batches of twenty over two weeks. Nine titles come back too long for the field and are shortened with the owner on a fifteen-minute call. Alt text is added to two hundred and ten images across those eighty pages, about forty of which turn out to be the same stock photograph of a workbench and get the same line. Eleven images are flagged: nobody in the business can say what tool is in the picture. The returned sheet shows every row closed, and the blanks list names the eleven photographs and the four articles whose descriptions were never written.

How it runs

1. A half-hour call. What system you run, how many pages there are, and whether the wording already exists. No charge.
2. A fixed price in writing. A price per page, in batches, agreed before anything starts. The per-page price falls as the batch gets larger.
3. The inputs. Your spreadsheet of titles, descriptions and alt text, and an editor account on the site with no more permission than the job needs.
4. The first ten. Ten pages done and sent to you, so you can confirm the fields are the ones you meant before the other fifty are touched.
5. The batches. Twenty to thirty pages a day once the first ten are approved.
6. The handover. The completed sheet and the blanks list, sent when the last batch closes.

What it costs

Priced per page in CAD, in batches, and fixed in writing after the half-hour call. What moves the number is the number of images on each page, whether the system keeps its fields in sensible places or hides them behind a plugin, and whether the wording is supplied or has to be shortened with you. Small batches cost more per page than large ones because the setting up is the same either way. The pricing page shows the four published ways of working, and ongoing entry is usually folded into a monthly arrangement rather than bought as a batch every quarter.

What happens next

Once the fields are in, the blanks list is the next piece of work, and it is usually writing rather than entry. If new pages keep appearing with empty fields, the fix is not another batch in six months but page creation or CMS content entry that fills them in as it goes. And if you want somebody to confirm the fields survived the next site update, content QA catches it. This piece of work is complete on its own and nothing further is assumed.