Google Not Indexing XML Sitemap Images
-
Hi Mozzers,
We are having an issue with our XML sitemap images not being indexed.
The site has over 39,000 pages and 17,500 images submitted in GWT. If you take a look at the attached screenshot, 'GWT Images - Not Indexed', you can see that the majority of the pages are being indexed - but none of the images are.
The first thing you should know about the images is that they are hosted on a content delivery network (CDN), rather than on the site itself. However, Google advice suggests hosting on a CDN is fine - see second screenshot, 'Google CDN Advice'. That advice says to either (i) ensure the hosting site is verified in GWT or (ii) submit in robots.txt. As we can't verify the hosting site in GWT, we had opted to submit via robots.txt.
There are 3 sitemap indexes: 1) http://www.greenplantswap.co.uk/sitemap_index.xml, 2) http://www.greenplantswap.co.uk/sitemap/plant_genera/listings.xml and 3) http://www.greenplantswap.co.uk/sitemap/plant_genera/plants.xml.
Each sitemap index is split up into often hundreds or thousands of smaller XML sitemaps. This is necessary due to the size of the site and how we have decided to pull URLs in. Essentially, if we did it another way, it may have involved some of the sitemaps being massive and thus taking upwards of a minute to load.
To give you an idea of what is being submitted to Google in one of the sitemaps, please see view-source:http://www.greenplantswap.co.uk/sitemap/plant_genera/4/listings.xml?page=1.
Originally, the images were SSL, so we decided to reverted to non-SSL URLs as that was an easy change. But over a week later, that seems to have had no impact. The image URLs are ugly... but should this prevent them from being indexed?
The strange thing is that a very small number of images have been indexed - see http://goo.gl/P8GMn. I don't know if this is an anomaly or whether it suggests no issue with how the images have been set up - thus, there may be another issue.
Sorry for the long message but I would be extremely grateful for any insight into this. I have tried to offer as much information as I can, however please do let me know if this is not enough.
Thank you for taking the time to read and help.
Regards,
Mark
-
Hi Mark,
I'm just following the thread as I have a similar problem. Would you mind sharing your results from the tests?
Thanks,
Bogdan -
Thanks Everett - that's exactly what I intend to do.
We will be testing two new sitemaps with 100 x URLs each. 1. With just the file extension removed and 2. With the entire cropping part of the URL removed, as suggested by Matt.
Will be interested to see whether just one or both of the sitemaps are successful. Will of course post the outcome here, for anyone who might have this problem in future.
-
It isn't always that simple. Maybe commas don't present a problem on their own. Maybe double file extensions don't present a problem on their own. Maybe a CDN doesn't present a problem on its own. Maybe very long, complicated URLs don't present a problem on their own.
You have all of these. Together, in any combination, they could make indexation of your images a problem for Google.
Just test it out on a few. Get rid of the file extension. If that doesn't work, get rid of the comma. That is all you can do. Start with whatever is easiest for the developer to implement, and test it out on a few before rolling it out across all of your images.
-
Cheers for that mate - especially the useful Excel formula.
I am going to try a few things in isolation so that we can accurately say which element/s caused the issue.
Thanks again, mate.
-
Ignore the developer - what worked for one doesn't mean it'll work for you
The easiest way to test this is to manually create a sitemap with 100 or so 'clean' image URLs. Just pull the messy ones into excel and use the formula below to create a clean version (Use A1 for messy, B1 for formula).
Good luck mate.
=CONCATENATE("image:imageimage:lochttp://res.cloudinary.com/greenplantswap/image/upload/",RIGHT(A1,LEN(A1)-(FIND("",(SUBSTITUTE(A1,"/","",(IF(LEN(TRIM(A1))=0,0,LEN(TRIM(A1))-LEN(SUBSTITUTE(A1,"/",""))))))))),"</image:loc></image:image>")
-
Thanks for the responses guys, much appreciated.
In terms of the commas, that was something that I put to the developer, however he was able to come back with examples where this has clearly not been an issue - e.g. apartable.com have commas in their URLs and use the same CDN (Coudinary).
However, I agree with you that double file extension could be the issue. I may have to wait until next week to find out as the developer is working on another project, but will post the outcome here once I know.
Thank you again for the help!
-
Hello Edlondon,
I think you're probably answering your own question here. Google typically doesn't have any problem indexing images served from a CDN. However, I've seen Google have problems with commas in the URL at times. Typically it happens when other elements in the URL are also troublesome, such as your double file extension.
Are you able to rename the files to get rid of the superfluous .jpg extension? If so, I'd recommend trying it out on a few dozen images. We could come up with a lot of hypothesis, but that would be the one I'd test first.
-
Hmmm I step off here, never used cloudinary.com or even heard of them. I personally use NetDNA, with pull zones (which means that they load the image/css/js from your origin and store a version on their servers) while handling cropping/resizing from my own end (via PHP and then loading that image, example: http://cdn.fulltraffic.net/blog/thumb/58x58/youtube-video-xQmQeKU25zg.jpg try changing the 58x58 to another size and my server will handle the crop/resize while NetDNA will serve it and store for future loads).
-
Found one of the sites with the same Cloudinary URLs with commas - apartable.com
See Google image results: https://www.google.co.uk/search?q=site:apartable.com&tbm=isch
Their images appear to be well indexed. One thing I have noticed, however, is that we often have .jpg twice in the image URL. E.g.:
- http://res.cloudinary.com/greenplantswap/image/upload/c_crop,g_north,h_0.9,w_1.0/c_fill,d_no_image_icon-720x720**.jpg**,g_center,h_900,q_80,w_900/v1352574983/oyfos82vwvmxdx91hxaw**.jpg**
- http://res.cloudinary.com/greenplantswap/image/upload/c_crop,g_north,h_0.9,w_1.0/c_fill,d_no_image_icon-720x720**.jpg**,g_center,h_900,q_80,w_900/v1352574989/s09cv3krfn7gbyvw3r2y**.jpg**
- http://res.cloudinary.com/greenplantswap/image/upload/c_crop,g_north,h_0.9,w_1.0/c_fill,d_no_image_icon-720x720**.jpg**,g_center,h_407,q_80,w_407/v1352575010/rl7cl4xi0timza1sgzxj**.jpg**
Wonder if that is confusing Google? If so, none of this is consistent, as they do have a few images indexed with exactly the same kind of URL as those listed above.
-
Thought I had them on email but must be within our fairly cumbersome Skype thread... let me have a dig through when I get chance and I'll post them up here.
-
Hmmmm, okay... Could you post the examples they gave, and an example page where the images are located on the site?
-
Hi Matt,
Thought I should let you know that (i) the X-Robots-Tag was not set, so that's not the issue and (ii) the URLs, although ugly, are not the issue either. We had a couple of examples of websites with the same thing (I'm told the commas facilitate on-the-fly sizing and cropping) and their images were indexed fine.
So, back to the drawing board for me! Thank you very much for the suggestions, really do appreciate it.
Mark
-
Hmm interesting - we hadn't thought of the X-Robots-Tag http header. I'm going to fire that over to the developer now.
As for the URLs, they are awful! But I am told that this is not a problem - but perhaps this is worth re-chasing up as other solutions have, so far, been unfruitful.
Thanks for taking the time to help, Matt - I'll let you know if that fixes it! Unfortunately it could be another week before I know, as the developer is currently working on another project so any changes may be early-mid next week.
Thanks again...
-
This is a bit of a long shot but if the files have been uploaded using their API it may have been that the 'X-Robots-Tag' http header is set to no-index...
Also, those URLs don't look great with the commas in them. Have you tried doing a small subset that just has the image id (e.g. http://res.cloudinary.com/greenplantswap/image/upload/nprvu0z6ri227cgnpmqc.jpg)?
Matt
-
Hi Federico,
Thanks very much for taking the time to respond.
To answer your question, we are using http://cloudinary.com/. So, taking one of the examples from the XML sitemap I posted above, an example of an image URL is http://res.cloudinary.com/greenplantswap/image/upload/c_crop,g_north,h_0.9,w_1.0/c_fill,d_no_image_icon-720x720.jpg,g_center,h_900,q_80,w_900/v1352575097/nprvu0z6ri227cgnpmqc.jpg (what a lovely URL!).
I had a look at http://res.cloudinary.com/robots.txt and it seems that they are not blocking anything - the disallow instruction is commented out. I assume that is indeed the robots.txt I should be looking at?
Assuming it is, this does not appear to get to the bottom of why the images are not being indexed.
Any further assistance would be greatly appreciated - we have 17k unique images that could be driving traffic and this is a key way that people find our kind of website.
Thanks,
Mark
-
Within that robot.txt file on the CDN (which one are you using?) have you set to allow Google to index them?
Most CDNs I know allows you to block engines via the robots.txt to avoid bandwidth consumption.
In the case you are using NetDNA (MaxCDN) or the like, make sure your robots file isn't disallowing robots to crawl.
We are using a CDN too to deliver images and static files and all of them are being indexed, we tested disallowing crawlers but it caused a lot of warnings, so instead we no allow all of them to read and index content (is a small price to pay to have your content indexed).
Hope that helps!
Got a burning SEO question?
Subscribe to Moz Pro to gain full access to Q&A, answer questions, and ask your own.
Browse Questions
Explore more categories
-
Moz Tools
Chat with the community about the Moz tools.
-
SEO Tactics
Discuss the SEO process with fellow marketers
-
Community
Discuss industry events, jobs, and news!
-
Digital Marketing
Chat about tactics outside of SEO
-
Research & Trends
Dive into research and trends in the search industry.
-
Support
Connect on product support and feature requests.
Related Questions
-
Google has discovered a URL but won't index it?
Hey all, have a really strange situation I've never encountered before. I launched a new website about 2 months ago. It took an awfully long time to get index, probably 3 weeks. When it did, only the homepage was indexed. I completed the site, all it's pages, made and submitted a sitemap...all about a month ago. The coverage report shows that Google has discovered the URL's but not indexed them. Weirdly, 3 of the pages ARE indexed, but the rest are not. So I have 42 URL's in the coverage report listed as "Excluded" and 39 say "Discovered- currently not indexed." When I inspect any of these URL's, it says "this page is not in the index, but not because of an error." They are listed as crawled - currently not indexed or discovered - currently not indexed. But 3 of them are, and I updated those pages, and now those changes are reflected in Google's index. I have no idea how those 3 made it in while others didn't, or why the crawler came back and indexed the changes but continues to leave the others out. Has anyone seen this before and know what to do?
Intermediate & Advanced SEO | | DanDeceuster0 -
Our client's web property recently switched over to secure pages (https) however there non secure pages (http) are still being indexed in Google. Should we request in GWMT to have the non secure pages deindexed?
Our client recently switched over to https via new SSL. They have also implemented rel canonicals for most of their internal webpages (that point to the https). However many of their non secure webpages are still being indexed by Google. We have access to their GWMT for both the secure and non secure pages.
Intermediate & Advanced SEO | | RosemaryB
Should we just let Google figure out what to do with the non secure pages? We would like to setup 301 redirects from the old non secure pages to the new secure pages, but were not sure if this is going to happen. We thought about requesting in GWMT for Google to remove the non secure pages. However we felt this was pretty drastic. Any recommendations would be much appreciated.0 -
XML Sitemap Questions For Big Site
Hey Guys, I have a few question about XML Sitemaps. For a social site that is going to have presonal accounts created, what is the best way to get them indexed? When it comes to profiles I found out that twitter (https://twitter.com/i/directory/profiles) and facebook (https://www.facebook.com/find-friends?ref=pf) have directory pages, but Google plus has xml index pages (http://www.gstatic.com/s2/sitemaps/profiles-sitemap.xml). If we go the XML route, how would we automatically add new profiles to the sitemap? Or is the only option to keep updating your xml profiles using a third party software (sitemapwriter)? If a user chooses to not have their profile indexed (by default it will be index-able), how do we go about deindexing that profile? Is their an automatic way of doing this? Lastly, has anyone dappled with google sitemap generator (https://code.google.com/p/googlesitemapgenerator/) if so do you recommend it? Thank you!
Intermediate & Advanced SEO | | keywordwizzard0 -
Our login pages are being indexed by Google - How do you remove them?
Each of our login pages show up under different subdomains of our website. Currently these are accessible by Google which is a huge competitive advantage for our competitors looking for our client list. We've done a few things to try to rectify the problem: - No index/archive to each login page Robot.txt to all subdomains to block search engines gone into webmaster tools and added the subdomain of one of our bigger clients then requested to remove it from Google (This would be great to do for every subdomain but we have a LOT of clients and it would require tons of backend work to make this happen.) Other than the last option, is there something we can do that will remove subdomains from being viewed from search engines? We know the robots.txt are working since the message on search results say: "A description for this result is not available because of this site's robots.txt – learn more." But we'd like the whole link to disappear.. Any suggestions?
Intermediate & Advanced SEO | | desmond.liang1 -
Google & Bing not indexing a Joomla Site properly....
Can someone explain the following to me please. The background: I launched a new website - new domain with no history. I added the domain to my Bing webmaster tools account, verified the domain and submitted the XML sitemap at the same time. I added the domain to my Google analytics account and link webmaster tools and verified the domain - I was NOT asked to submit the sitemap or anything. The site has only 10 pages. The situation: The site shows up in bing when I search using site:www.domain.com - Pages indexed:- 1 (the home page) The site shows up in google when I search using site:www.domain.com - Pages indexed:- 30 Please note Google found 30 pages - the sitemap and site only has 10 pages - I have found out due to the way the site has been built that there are "hidden" pages i.e. A page displaying half of a page as it is made up using element in Joomla. My questions:- 1. Why does Bing find 1 page and Google find 30 - surely Bing should at least find the 10 pages of the site as it has the sitemap? (I suspect I know the answer but I want other peoples input). 2. Why does Google find these hidden elements - Whats the best way to sort this - controllnig the htaccess or robots.txt OR have the programmer look into how Joomla works more to stop this happening. 3. Any Joomla experts out there had the same experience with "hidden" pages showing when you type site:www.domain.com into Google. I will look forward to your input! 🙂
Intermediate & Advanced SEO | | JohnW-UK0 -
Panda Recovery - What is the best way to shrink your index and make Google aware?
We have been hit significantly with Panda and assume that our large index with some pages holding thin/duplicate content being the reason. We have reduced our index size by 95% and have done significant content development on the remaining 5% pages. For the old, removed pages, we have installed 410 responses (Page does not exist any longer) and made sure that they are removed from the sitempa submitted to Google; however after over a month we still see Google spider returning to the same pages and the webmaster tools shows no indicator that Google is shrinking our index size. Are there more effective and automated ways to make Google aware of a smaller index size in hope of Panda recovery? Potentially using the robots.txt file, GWT URL removal tool etc? Thanks /sp80
Intermediate & Advanced SEO | | sp800 -
Google+ Pages on Google SERP
Do you think that a Google+ Page (not profile) could appear on the Google SERP as a Rich Snippet Author? Thanks
Intermediate & Advanced SEO | | overalia0 -
Do you bother cleaning duplicate content from Googles Index?
Hi, I'm in the process of instructing developers to stop producing duplicate content, however a lot of duplicate content is already in Google's Index and I'm wondering if I should bother getting it removed... I'd appreciate it if you could let me know what you'd do... For example one 'type' of page is being crawled thousands of times, but it only has 7 instances in the index which don't rank for anything. For this example I'm thinking of just stopping Google from accessing that page 'type'. Do you think this is right? Do you normally meta NoIndex,follow the page, wait for the pages to be removed from Google's Index, and then stop the duplicate content from being crawled? Or do you just stop the pages from being crawled and let Google sort out its own Index in its own time? Thanks FashionLux
Intermediate & Advanced SEO | | FashionLux0