Google Image Search Accuracy
- SEO |
Hello my SEO colleagues, I need help
I have an idea to check the*efficiency*of google sitemap for images, if it affects significantly the image's ranking on a specific keyword's SERP. It's something like a small SEO research.
The technical part of the project is almost done. I wrote a script that analyzes google images search results. And it's not a problem to export the*data*from google analytics and get a list of keywords that brings main traffic from the images search.
I have 2 projects to work with right now, with 10k and 200k of images. But I suppose that it's not enough. So I'm looking for projects with 1k+ images in the google index, but with no google sitemap for images.
I'm ready to help with creating google sitemap for images (especially if site is done with php). We export the data from google analytics with current stats of the google images search traffic, and then export it again when the google sitemap for images will be ready. I plan to do the export daily during 30-50 days. It should be enough time for google to update its index based on sitemap, if it will do any noticeable update at all. Anyway it will not harm the project.
It'll be something like report at the end.
Besides there is another idea about the google images search stats, but it's more complex -links 'accuracy' in the Google Image Search*(%). Every image of the site has to be linked with a certain page. For example an image of some item from the site's catalogue has to be linked with the page for this very item, but not with its category's page or with the page that presents a small announce of this item.
The script can get the most popular image/page pairs from the Google Index using search queries like "site:someSiteName someKeyword" and the Google API. It's possible to analyse 'accuracy' of the Google Index based on 'right' rules of relationship between different types of images and the structure of the site. It's possible to check whether traffic from mismatched image/pages is worse*performing*than traffic from correctly matched images. Afterwards, you could test whether using sitemaps fixes the mismatches, as well as what percentage of total images are missmatched. Maybe it sounds complex, but technically it's a rather simple task.
If you have any ideas about this*'accuracy'*stats you are welcome.
Special for moderators: it's not a span or something like this. I do not sell any goods or services. All results of this 'research' will be free. I'd like just to discuss such idea.
Best regards.
I have an idea to check the*efficiency*of google sitemap for images, if it affects significantly the image's ranking on a specific keyword's SERP. It's something like a small SEO research.
The technical part of the project is almost done. I wrote a script that analyzes google images search results. And it's not a problem to export the*data*from google analytics and get a list of keywords that brings main traffic from the images search.
I have 2 projects to work with right now, with 10k and 200k of images. But I suppose that it's not enough. So I'm looking for projects with 1k+ images in the google index, but with no google sitemap for images.
I'm ready to help with creating google sitemap for images (especially if site is done with php). We export the data from google analytics with current stats of the google images search traffic, and then export it again when the google sitemap for images will be ready. I plan to do the export daily during 30-50 days. It should be enough time for google to update its index based on sitemap, if it will do any noticeable update at all. Anyway it will not harm the project.
It'll be something like report at the end.
Besides there is another idea about the google images search stats, but it's more complex -links 'accuracy' in the Google Image Search*(%). Every image of the site has to be linked with a certain page. For example an image of some item from the site's catalogue has to be linked with the page for this very item, but not with its category's page or with the page that presents a small announce of this item.
The script can get the most popular image/page pairs from the Google Index using search queries like "site:someSiteName someKeyword" and the Google API. It's possible to analyse 'accuracy' of the Google Index based on 'right' rules of relationship between different types of images and the structure of the site. It's possible to check whether traffic from mismatched image/pages is worse*performing*than traffic from correctly matched images. Afterwards, you could test whether using sitemaps fixes the mismatches, as well as what percentage of total images are missmatched. Maybe it sounds complex, but technically it's a rather simple task.
If you have any ideas about this*'accuracy'*stats you are welcome.
Special for moderators: it's not a span or something like this. I do not sell any goods or services. All results of this 'research' will be free. I'd like just to discuss such idea.
Best regards.