Google’s John Mueller responded to questions round picture web optimization, particularly explaining that whereas the picture itself and the alt textual content is beneficial, so is the web page that picture is embedded on. That is all helpful for Google to know the relevance of a picture primarily based on a particular question.
John mentioned relating to understanding the picture, having a photograph of a seaside on a website is nice and all, however it would not inform you a lot. If that photograph is embedded on a lodge web site, it means, that’s the seaside by that lodge. Whether it is embedded on a photographer’s web site, it tells you this photographer can take images by the seaside, and so forth – you get the purpose.
A Reddit thread requested:
Are pictures alt texts nonetheless related for web optimization with all the pc imaginative and prescient and pictures recognition development? Is there any data of Google or different engines like google utilizing machine studying fashions to crawl pictures reasonably than counting on the consumer offered alt texts?
Right here is how John replied:
For picture search, there’s the context that comes from the web page + picture mixture that issues. A photograph of a seaside is likely to be a calming poster, it is likely to be the seaside from a lodge, it could possibly be the positioning of a chemical spill. Simply figuring out that the picture is of a seaside would not actually give ample background data to have the ability to present it in picture search appropriately. Plenty of it does come from the web page, and the alt-text is exclusive in that it is what immediately connects the picture to the web page with context.
Except your website is a photograph company, visitors for “photograph of a seaside” is not going to be that helpful – however for a lodge, having “lodge with seaside in X” may be related. Once more, a number of that may come from the remainder of the web page, however the alt attribute worth is a novel alternative to offer context. (And with that … should you use AI to create alt texts primarily based on the picture file, and get “photograph of a seaside” because the alt textual content for that picture, you are not getting essentially the most out of the alt textual content, each for customers & engines like google.)
So sure, the web page the picture is embedded on can also be necessary for a lot of causes.
Discussion board dialogue at Reddit.
