32 long titles, and what shortening them did

A title that is too long does not get a smaller font. Google cuts it, usually at a word boundary near 600 pixels on desktop, and shows an ellipsis. On one of our sites 32 of 63 pages were in that state. We shortened them on 4 September 2026. This is what the check measures, what changed, and the part that did not change.

What does this check detect?

Moonleap renders each title's width with the font metrics Google uses on desktop results rather than counting characters, because "Productfotografie voor e-commerce: de complete gids voor 2026 | OmniRoom" and a title of the same character count in narrow letters truncate at different points. The finding lists each page, the width, and where the cut falls. Descriptions are measured in characters because Google rewrites them freely anyway.

Why does truncation matter, and how much?

Google's documentation is direct: title links are generated from several sources, the <title> is the strongest, and titles that are too long or stuffed are more likely to be rewritten. A truncated "…complete gids voor 2026 | Omni" loses the brand and the year, the two things a returning reader scans for.

The size of the lever is the part people overestimate. A title change can move click-through on a query by a few points at the same position. It cannot create impressions, and it cannot win a click that an AI Overview or a shopping carousel above you is absorbing.

What happened after the rewrite?

At the 27-day reading the pages with shortened titles ranked higher and were shown more. Their clicks did not follow, and the per-query view showed why: the queries they gained were informational ones where an AI Overview sits above the first result. The rewrite did what a title can do. It could not do the rest.

Two things we would do differently. Shorten titles in one batch and leave them alone for 28 days; we changed a page inside the window for another reason and had to drop it from the measurement. And read the outcome per query, not as a site average; the average CTR fell for reasons that had nothing to do with titles.

What is the fix?

For a guide page: <what the page answers> | <Brand>, with the answer phrase under about 50 characters so the brand survives. For a product: <Product name> <key attribute> | <Brand>. Generate titles from data in the layout rather than writing them by hand, and add the width check to your build. Moonleap's finding includes the shortened suggestion for each page; the agent applies the pattern across the template, not 32 individual edits.

For descriptions, 120 to 155 characters, one sentence that says what the reader gets, no keyword lists. Google rewrites descriptions more often than titles, so spend the effort on titles first.

How do you verify it?

The visual check takes a minute per page and catches what the pixel rule cannot: Google deciding to use your H1 instead. For the measurement, filter Search Console to the pages you changed, compare the 28 days after against the 28 before, and read CTR only where position is stable; a page that also moved from 14 to 9 has two changes in one number.

FAQ

Is there a character limit for titles?

No fixed one. Google cuts by pixel width, roughly 600 pixels on desktop, which is 50 to 60 characters in a typical Latin font and fewer in wide fonts or capitals.

Does Google use my title at all?

Usually. It rewrites when the title is too long, boilerplate, missing, or does not match the page. A concise, accurate title is used most of the time.

Should the brand go first or last?

Last, unless people search for the brand plus the topic. The first words carry the query match; the brand is recognised even when cut.