Optimizing for Product Colors: Long Tail Gold or Duplicate Content?
Colors are search modifiers that can bring a lot of long-tail traffic. When someone searches for a particular product and color, it often indicates someone is close to a purchase, or at least further along the sale-trail than one who goes broad.
But you can’t create a separate product page and URL for each color because that’s duplicate content, and duplicate content is the worst of sins, right? That’s what I thought until I started testing it - and it turned everything my momma ever told me about duplicate content on its head.
(If your momma never had “the talk” with you - you know, *content reproduction,* we recently did a duplicate content post that included a PG13 explanation. I made sure this post was completely different so nobody mistakes it for duplicate content).
Yes, Virginia There Is A Santa Clause…And You Can Optimize Product Pages for Color
Here’s an example:
Jessica Bennett Shoes sells its product through its own e-store and various retailers like Amazon, Zappos and ShoeBuy. One of its styles is called “Harli.” It’s made from burlap and comes in navy, beige and brown.
Shoebuy.com has 3 indexed product pages for Harli – one for each color.
Each page has an identical meta description, and according to Webconfs’ Similar Page Checker, these pages are 100% identical.
But Shoebuy not only owns top spot for each color, Google’s also throwing in some indented result love. When you search for “jessica bennett harli navy” (at time of writing and from my data center):
Top ranking… and for “jessica bennett harli brown”:
“jessica bennet harli beige”:
The only differentiators between the 3 color pages are the URLs (just numbers, no keywords) and the title tag. I’ve scoped out other sites that use different pages for different colors and they all seem to rank fine when color is included in the search query. The technique seems to be create color-specific pages in addition to one main product page (hence, indented results). Since all pages are indexed, the color pages are selected to appear when someone searches for the color, with the non-color, main product page potentially appearing as an indented, second result.
This leads me to believe that as long as your color pages are getting indexed, you don’t need to worry about duplicate content smackdown.
Ever try shopping for wedding stuff? Color modifiers are a must! But, no one optimizes for color attributes well at all. I see big upside for sites able to do this. One of our customers has done a decent job with color/shade attributes: http://www.makeup.com
Great advice, I agree that this is something that most ecommerce sites should be taking advantage of.
The only thing I would be concerned about is making sure that if the colors you optimize for and get listed under are always in-stock or that you’re suggesting an alternative on the product page. We don’t want to see 404 errors because the color the customer clicked on in Google isn’t available any longer.
The great thing about the tactic above is, you have SERP control from using various properties to augment the brand in addition to your search modifiers.
Since your pages are linked (the color specific page and the standard brand) the double listing is evoked from the similarity in the Title.
The quandary about duplicate content is only to prevent the same result across multiple domains, so in essence, even if you toggled a filter, it would just remove one of the URL’s from the list on that page. You still have great positioning regardless as a result of diversification, so technically you can rank all the way to the bank with that strategy (as long as the domains are not inter-linked).
Great post, Linda.
Just goes to show you that you can’t take anything for granted that you’ve read on SEO until you run a few tests. I long thought that using a tactic like this would simply be fruitless because Google would filter out the additional, or duplicate, pages. Here we see evidence that that is most definitely not the case.
Definitely something to keep in mind when putting together an SEO strategy for any website.
I wonder what other modifiers are not in Google’s filter. I would assume there would be others like size of clothes, large, small and such.
I’ve been playing around with testing duplicate content aimed at different countries and it’s definitely possible that search engines can recognize www.site.com/product/en/us and www.site.com/product/en/uk for example, even when the product descriptions are identical. However this depends on whether the searcher has selected “search only pages from ‘country’” vs. “search the web.” This testing will probably turn into a blog post soon enough ;)
This interesting..^^ too bad my products don’t have any color variations..^^
This site has a facebook banner running that shows their colors (I uploaded the photo so you can see what I mean)
http://www.getelastic.com/wp-content/uploads/dessy-shot.jpg
Great for usability - you can look at dresses and click the color swatches to change - but they’re all in Flash. Example: http://www.dessy.com/dresses/bridesmaid/2721/
Throw this into Google to see if you can find any pages optimized for pink:
pink site:http://www.dessy.com/
Nada.
There should be special landing pages for each of these colors at the very least. “Pink bridesmaid dresses” bla bla bla - link to Flash app or have the Flash app embedded.
Like they do with branded terms:
http://www.dessy.com/dresses/bridesmaid/d400/
@Commercestyle
Perhaps a nocache attribute (that keeps page out of Google cache) will prevent that. However, if you show stock levels on site, and you’re sold out, you could direct them to similar colors or similar styles. Yes a lot of custom programming…
Let me guess … your client is Zappos? ;) Just kidding hehe.
Anyways, Jeff makes a good point - duplicate content on the same site = canonicalization issue. Duplicate content on two different sites = duplicate content issue.
As to colours, if someone wants shoes in a specific colour, it seems that indexing the various product pages featuring the specific colours would be a good idea.
From a merchant’s perspective, you could get around the canonicalization issue by listing matching outfits that most people would have in their wardrobe. Obviously it helps to know who your clients are in this situation, and it gets expensive to scale this. But you can templatize it to an extent (e.g. “white shoes” should often be highlighted with black skirts and white tops, or perhaps with beige/summery pants/shorts…)