Hotlinking to merchant feed images

[speedshopping]

Hi,

We pull in product feeds from various 3rd party sources which all have a standard image. We display these images within our search results, but hotlink to the image via the merchant URL.

We notice a lot of our bigger competitors actually grab the image and store it locally on their servers to avoid hotlinking - does anyone have any experience in this area, and if it could be damaging to our rankings. We are currently in a 950 penalty (we think) so are looking into legitimate reasons why.

Cheers

[tedster]

I've never seen hotlinking to manufacturer images as a cause for a minus 950 penalty, or any severe penalty for that matter. So I doubt that serving your own copy will fix your situation, but I do think it makes sense as one step to take.

That said, I've never been willing to hotlink to a manufacturer's image or even to serve an exact duplicate. For one thing, most manufacturer's images are too limited in gamut because they've been created for CMYK, and need some restoring to the more vivid gamut that makes an image truly appealing online.

For another, it's rare the the image is properly cropped for the page I'm building and i want the entire effect of the page to be "just so". That's just two of the reasons for me.

[speedshopping]

Hi Tedster,

Thanks for your reply - since recently looking, our site has over half a million images indexed - all of which are merchant image hotlink URLs from the feeds they supply us with.

We are gaining quite a lot of traffic from these images, so could Google be perceiving us as some kind of scraper and is punishing us in the Google search index?

[tedster]

It sounds possible - who knows for sure except Google engineers?

... from the feeds they supply us with

Do you republish text content from those feeds, without any enhancements? If so, that is a major cause of Panda devaluation. It's called "thin content" and Google was already cutting back rankings for that kind of page even before Panda.

Panda is targeting "shallow" content, but from what people have been bringing to me, thin content is one kind of shallow content they did take a big swipe at.

[speedshopping]

yeh we are trying to seperate the Panda from the 950 right now, so we can tackle the "thin" based content once we can establish our panda rankings for real.

[Makis77]

I have the same question.
One of my WP websites that fetched amazon product details like descr, tags, images etc went from 1st page to 50+ in one week.
After posting some orginal content for all products and changing meta tags I saw a small improvement but still too far from my old rankings.
I m still using amazon product images as hotlinks for many of my products so I wonder if this is the reason that panda is still punishing my website rankings.

[lucy24]

Take off the webmaster hat and put on the user hat for a moment. Have you ever tried to load up a page that's composed entirely of hotlinked images? Especially the ones that take the original image and resize it dynamically to a thumbnail. Those things take hours. (This is a hypothetical figure. I'm on slowish DSL and have never had the patience to load up the whole thing.) Now put the webmaster hat back on and calculate the extra bandwidth, which is probably a bigger concern than storage space.

Somewhere hereabouts there's a current thread about users' preferred maximum load time for a page. The people who hotlink to me probably don't have users--or even look at their own pages. But you do. Why risk it?

[scooterdude]

I 've actually had this at the back of my mind for a while, but held of because

a, i am uncertain about the impact it would have

b, would i have to pay for these images if i downloaded them on to my server as opposed to using the merchant supplied url

c, it would ad another massive task to back end work when i need to be promoting low traffic sites like mine

Sometimes improvements like this can lead to giant or extraordinary leaps in performance, an webmasterworld has been a place for this :)

thanks guys

[speedshopping]

Hi,

I just wanted to pick this up again and ask for some advice.

We have pages that sit in directories such as:

/xyz/apple/a/
/xyz/apple1/a/
/xyz/apple2/a/
/xyz/banana/a/
/xyz/pear/a/

Within our 3 apple pages, we have product feeds where we hotlink to our merchants images URLs.

My question is:

If I added the following line to our robots.txt file, would this prevent Googlebot-Image from crawling the apple pages (or any beginning with A) and remove the images from Google images?

User-agent: Googlebot-Image
Disallow: /xyz/a*

Cheers in advance.

[aakk9999]

Remove the asterisk after a
It should be like this:

User-agent: Googlebot-Image
Disallow: /xyz/a

This will disallow anything starting with /xyz/a

[robotstxt.org...]

Specifically, you cannot have lines like "User-agent: *bot*", "Disallow: /tmp/*" or "Disallow: *.gif".

If you want to be more specific, then you can use

User-agent: Googlebot-Image
Disallow: /xyz/apple

Which will disallow anything starting with /xyz/apple therfore covering all three of your apple folders, but will allow other folders/files starting with /xyz/a

Not sure if already crawled images would be removed from Google images though.

[indyank]

I think it is fine to have asterisk - [google.com...]

But make sure that you don't block any other directory or pages starting with a.

[edited by: indyank at 3:31 am (utc) on Jun 3, 2011]

[indyank]

Moreover googlebot supports the allow directive but other bots may or may not support it.