Title tags and meta descriptions

What title tags and meta descriptions are

A title tag is the page title stored in your site's code. It is the clickable blue headline in a list of search results and the text on the browser tab. A meta description is the short summary printed underneath it. Together they are the advert for a page you have already written, shown to someone who has not yet decided whether to open it.

They are not the headline on the page. The visible heading a reader sees after clicking is a different piece of text, dealt with under heading structure. They are also not a ranking device on their own: a good title will not rescue a page that does not deserve to be in the list, and this service does not pretend otherwise.

The limits are practical rather than absolute. A title tag beyond roughly sixty characters is cut off in the list, and a meta description beyond about a hundred and fifty-five characters gets the same treatment. The important words go first.

When you need this

You need it when your site has grown to a few dozen pages and nobody has ever written these deliberately. The usual state of affairs is a site where most titles were generated from the page name, several pages share the same title word for word, and half the meta descriptions are empty or repeat the company slogan.

The second case is a site where the pages are sound and the listings undersell them. A page can sit in the results for a search it deserves and still be passed over, because the result beside it says what a buyer gets and yours says the name of a service category.

You do not need this if the pages underneath are weak. Rewriting the advert for a page that does not answer the question is effort spent in the wrong place, and an audit of what earns and what drags is the cheaper first purchase. You also do not need it for pages you would rather nobody found at all.

What gets written

• A unique title tag per page. Inside the length limit, carrying the words a buyer actually types, and different enough from every other title on the site that two of your own pages stop competing for the same result.
• A meta description per page. One or two sentences saying what the page contains and what the reader will be able to do afterwards, rather than a sentence about the company.
• A priority order. Money pages first: the service and product pages that produce inquiries, then the category pages above them, then the articles. Pages that produce nothing go last or not at all.
• The duplicates, listed. Every place where two or more pages currently share a title, with a note on which page should own the search and what should happen to the others.
• The blanks, filled. Pages with no description at all, which on most sites is a surprising number of them.

One honest note. Search engines rewrite these. A title you wrote may be replaced by text taken from the page itself, and a meta description is replaced more often than not, usually with whichever sentence matches the search. That is not a reason to skip the work: a well-written title is what the engine has to beat. It is a reason not to agonise over one page for an afternoon.

What you get

The sheet. One spreadsheet, one row per page: the address, the current title and description, the new ones, the character count of each, and the search the page is aimed at. Sorted in the order the work should be applied.

The duplicate list. A short separate tab naming the pages that currently collide, so the person applying the changes does not reintroduce a clash.

The application. You can paste the sheet into your site yourself; most people do. If you would rather not, I can do it as metadata entry and hand back the finished pages.

A worked example

An illustration. A Halifax firm selling and servicing commercial kitchen equipment has a hundred and forty pages. Twenty-two are service and product pages that produce inquiries, eleven are category pages, and the rest are articles going back seven years. Nineteen pages carry the same title as at least one other page, and sixty-one have no meta description.

The batch is written in tranches. The first covers the twenty-two earning pages and the eleven categories, with titles built around what installers and restaurant owners actually type: the equipment, the make, and the words “repair”, “parts” and “Halifax” where they belong. Four of the nineteen collisions turn out to be genuine duplicate pages rather than a titling problem, and those are flagged for merging. The second tranche covers the eleven articles that still bring visits. The remaining fifty-odd are left alone, because rewriting their adverts would change nothing, and the sheet says so plainly.

How it runs

1. A half-hour call. How many pages you have, which of them produce inquiries, and how you will apply the changes once they are written. No charge.
2. A fixed price in writing. Priced per page, with the count agreed before anything starts.
3. The export. A list of your pages with their current titles and descriptions, from your site or from read-only access to your search data. The collisions and blanks come out of this in an afternoon.
4. The writing. A batch of forty pages takes about a week; a hundred and fifty runs over two to three weeks, delivered in tranches so the earning pages can go live first.
5. Delivery. The sheet and the duplicate list, with a short call if the batch is large.

What it costs

Quoted as a fixed price in CAD after the half-hour call, priced per page, because that is how the work behaves. Larger batches come down per page, since the research into what your buyers type is shared across them. This is the cheapest item in this pillar and the one that shows a return soonest, which is why it is usually bought as a whole-site batch rather than a page at a time. The pricing page shows the four published ways of working, and the number here is fixed in writing before anything starts.

What happens next

Two of the pages in every batch turn out to deserve more than a new advert, and the usual follow-up is a refresh of the page itself. The batch also exposes pages competing for one search, which is handled by pruning and de-indexing or by merging them. Where the sheet shows you never knew what each page was for, keyword and topic research settles it once so the next batch writes itself. This piece of work is complete on its own, and nothing further is assumed.