Image processing via API and search engines

Hi All,
we prepare an E-Commerce relaunch and would like to use the image API to process our assets. Most of our images are product images – and we want them to rank high in image search.

We know that Search Engines need to re-index source images, when anything about them gets updated: URL, Alt-Text, Title, dimensions, or other pixel data. Consequently, that image loses its Search Engine ranking.

I wonder how Search Engines treat the processed duplicates of images provided by an Image API and CDN.

Assume you have many product images with transparent backgrounds. You have uploaded and Gumlet-processed these images a year ago. Search Engines found them, they provide traffic and lead to sales.

One day have the fun idea to insert a pink background in all images, using the background parameter – or you use any of the colour-operations and enhance their look in some other way (without changing the source image).

Once the cache refreshes, all modified versions would appear online.
How would Search-Engines react?

Hey there, That’s a very interesting question.

Yes, changing URLs will lose SEO value. In that sense, using a third-party CDN or API is bad for image SEO.

But, Gumlet takes care of this to ensure your original images do not lose SEO ranking.

We achieve this by adding a “canonical link” to every image we deliver for you. The canonical link points to the original image on your server. That way, the image on your domain and server gets all the SEO benefits.

You do not lose any SEO advantages as long the original image is not changed.

This also means you can modify any parameters in the Gumlet API freely without worrying about SEO. All your image links, old and new, will point to the same original image on your domain :slightly_smiling_face:

Read more about the canonical links from Google.

URL Canonicalization and the Canonical Tag | Google Search Central  |  Documentation  |  Google Developers)&visit_id=637914013139269779-371702148&rd=1

You can check the canonical links in your Gumlet delivered images by going to the Network tab, clicking the image URL, and checking the header. (Attached image for reference).

I hope this helps. [d13c952-gumlet-canonical-link.png]

Thanks a lot for your explanation, Divyesh! It likely also would help others to publish this info in the documentation :grinning:. I even knew what canonical links are, and read about them in your docs. It sometimes still helps to speak out what this means in practice, like you just did …

“This means you can modify any parameters in the Gumlet API freely without worrying about SEO. All your image links, old and new, will point to the same original image on your domain :slightly_smiling_face:

I assume Search Engines will then only crawl the source image on our server. In our case, images are lossless compressed PNGs with alpha channel and large pixel dimensions – they may have 10 MB or more.

Bots will therefore need more time to crawl individual images. They might burn through the crawl budget for the site earlier and some assets could remain out of the index.

Is this a justified thought – or are such file-sizes unproblematic, from your experience?

Hi Again,

Thanks for the suggestion. Canonical links are mentioned in our WordPress guide. We would try to make a general guide for the topic. Your questions have already helped us prepare the base structure for it.

Search engines crawl both separately. You page’s SEO depends on its web vitals, which are only affected by the actual image file present on the page.

The canonical reference just acts as a reference for Google to attribute the ranking power to the original image. That way, if multiple duplicate URLs for the original image rank for different things (say on different blogs) they will both give ranking power to the same original image.

Now whether google will crawl the original image or not, is beyond my knowledge.

Let me know if you would want to know anything further on this.

Many thanks again, Divyesh!