SERP Simulator: Preview Your Google Snippet

Google truncates titles by the width they occupy in pixels, not by how many characters they contain. The search result has a fixed slot on the page; when the text does not fit, the engine trims it and adds an ellipsis.

The gap between the two ways of measuring is not a rounding error. In Arial, the letter "i" occupies 4.443px and "W" occupies 18.877px — more than four times as much. A sixty-letter title made of narrow letters measures 266.6px; the same sixty letters in wide characters measure 1,132.6px. That is 4.25× the width for an identical character count, which is why "keep it under 60 characters" is right some of the time and wrong the rest — and why a SERP simulator that measures pixels tells you something a character count cannot.

English is the most compact of the three languages this site publishes in, which means a title that fits comfortably in English can overflow once translated. We measured it rather than repeating the figure that circulates without a source.

The commonly repeated 20–25% expansion figure does not hold up here — we measured roughly 6%. And the honest caveat runs in one direction: these titles were hand-written and already optimised not to truncate, so they likely understate natural expansion. A title translated without regard for width would stretch further.

What matters is not the average but the truncation rate. Six per cent sounds negligible until you notice it triples the number of cut-off titles: most titles live close to the limit, so a small nudge pushes many over it. That is the real reason advice calibrated on English hurts everyone else, and why it pays to run each localised title through a SERP simulator rather than trusting a round number.

One more thing worth knowing if you localise: accents cost nothing. In Arial, "á" is exactly as wide as "a", and the same goes for ñ, ç, ü and ã — accented glyphs keep the base letter's advance width so the rhythm of the text is preserved. The single exception is "í", which grows 25% because "i" is so narrow to begin with. What does cost width in Spanish is the inverted opening marks ¿ and ¡, which English never pays for.

Once the SERP simulator shows you the cut-off point, the craft becomes concrete: decide what lands before it. A few things reliably buy back width.

Front-load the distinctive words. Search results are scanned, not read, and the first two or three words carry most of the click decision. If the page is about pixel width, "pixel" belongs in position one or two, not after a category label the reader already knows from the URL.

Prefer a pipe to a dash — or drop the separator. Separators are pure overhead: they carry no meaning and still consume width. Measured in Arial at 20px, spaces included: " | " costs 16.3px, " - " 17.8px, " – " 22.2px and " — " 31.1px, while a plain ", " costs just 11.1px. Swapping an em dash for a pipe hands you back 15px — on a title sitting at 98% of budget, that is exactly the nudge that saves it.

Watch what numbers do. Digits in Arial are tabular: every digit from 0 to 9 has an identical advance width of 11.123px, which happens to be exactly the width of a lowercase "o". So "2026" costs 44.5px — precisely the same as "oooo", and more than you would guess from four narrow-looking glyphs. Years and list counts earn their keep, but they are not free.

Decide the brand suffix once, globally. Appending the site name to every title is a site-wide budget decision, not a per-page one. If your brand costs 120px, you are spending roughly a fifth of the desktop title budget on every single page. Many sites are better off adding it only where the brand is the differentiator — the homepage, product and pricing pages — and reclaiming that width everywhere else.

Check the description separately. Descriptions have far more room (about 920px) but are also rewritten by Google more aggressively than titles. Treat the first sentence as the part that must survive: put the promise there, and let the supporting detail take its chances in the tail.