It is a seemingly minor update, which mainly concerns those operating in foreign markets with alphabets other than Latin, but it is nonetheless interesting to report on it also because it confirms Mountain View’s focus on the issue. And so, while the SERPs are still at the mercy of the May 2022 Core Update, Google has also dedicated itself to an ‘algorithmic improvement’ for the way it selects titles for search result snippets of pages with multilingual or transliterated titles.
Google’s new update for title generation
Presenting the novelty is Koji Kojima’s article on the Search Central blog: thanks to the algorithmic improvement, Google is now able to identify documents in which the title element is written in a language or script system different from its content, and consequently chooses a title similar to the language and script of the document pursuing a criterion of uniformity and consistency.
More precisely, this update is based on the general principle that ‘the title of a document should be written in the language or script of its main content’, and Kojima points out that this mismatch is one of the reasons why Google can override the webmaster’s settings and change the title of a page shown in the web results.
Which sites and titles are affected by the update
There are two example cases mentioned in the article, which represent the focus of this Google update on title management.
Firstly, Google’s algorithm now intervenes to improve multilingual titles, in particular those that repeat the same content or the same sentence in two different languages or scripts (usually by adding an English version to the text of the original title).
The model shown by Google clarifies this particular case with a title composed of two parts, the first in Hindi characters and the second in English, divided by a hyphen, but expressing the same content in different languages (Hindi and English).
Even the title is in both languages, the document is only written in Hindi: Google’s system detects this inconsistency and decides to only use the Hindi title text for the SERP snippets, removing the English addition.
The other use case concerns transliteration, i.e. when content is transcribed from one language into a different language that uses a different script or alphabet. For example, Kojima explains, the title of a page for a song written in Hindi, but transliterated to use Latin characters instead of the native Hindi Devanagari script, would be as follows:
In this case, Google tries to find an alternative title using the predominant script on the page, which in this case could be:
Title and Google, what changes with the update
The new update to the algorithmic system underlying the generation of titles for Google’s results pages does not change things much, at least for those operating with only one language or in any case with a unique alphabetical system.
In general, the update does not upset the guidelines set by Google, and mostly affects sites with alphabetically mixed content, so to speak, affecting pages with multilingual titles or with different alphabetical systems (which may have seen changes in the SERPs in recent days, perhaps also related to changes in click-through rates).
Extending the discourse to titles in general, Google reiterates that as a rule the systems ‘tend to use the title element of the page‘, i.e. the title tag that we can set in HTML, and as a general advice it is preferable to provide a title that corresponds to the language or alphabet of the page’s main content, and in any case not to use different languages or scripts for the title that are not also present in the page’s content.
Indeed, in cases with multilingual or transliterated titles, Google’s systems may look for alternatives that match the predominant language of the page.