A Technical SEO’s Look at Endless.com
April 21st, 2007 by Michael Gray in SEO, conferenceIf you're new here, you may want to subscribe to my RSS feed. Read my top posts or learn more about Michael Gray. Want more frequent updates follow me on Twitter. Thanks for visiting!
Over the past few months I’ve had more than one client ask me my opinion about Endless.com Amazon’s new shoe shopping site. Since Rae talked about Endless and Danielle gave an Endless review I thought this would be a perfect opportunity to share some of what I saw, what I liked and what I didn’t.
I’m not exactly sure what it is that draws me to women’s fashion websites but sites like Styledash, StyleCritics, StyleCynics, Stylehive and the Shoe Blog are fascinating to me. I guess I’d have to say while I’m not a hot babe I do play one on the internet with some regularity, and that’s what keeps me curious, but back to SEO.
Endless.com launched right around the beginning of the year if we look in Google we see Google says they know about 900-ish pages
However when you try and see all the pages you see they go supplemental at around 100 results, and even worse they really only have only about 300 pages in the index
So why does Google tell you 900 when they only say 300, well that’s the magic of Google trying to obfuscate you from seeing as much real data as possible. They are totally into creating as much FUD and “pay no attention to the man behind the curtain” stuff as they can. Also the “In order to show you the most relevant results, we have omitted some entries very similar to the 299 already displayed.” message tells you they are “seeing” lots of internal duplication.
For most brand new websites getting 100 solid pages into the index in less than 120 days is OK (not great but OK), however for Endless.com who is a sub division of Amazon and has a real marketing budget it’s really pretty crappy. However where it really gets bad is when we look at the pages that are actually indexed:
There are some shopping/product pages but the overwhelming majority of the pages are “help” pages. Help pages aren’t bringing you any shopping traffic. So why is Endless having such a problem getting into the index? Is it a trust issue and are they victims of Superficial Crawling? While I’m sure that plays a role I think the real issue is that Endless is built with AJAX.
For example go to women’s shoes and click the ‘pumps’ category from the left menu. I see this page
However I can’t provide you with a direct link because the page is built with AJAX. Using the side controls I can very easily narrow my selection down to Taupe Wedges priced between $30 and $100 and get the page below, but I can’t send you the link or bookmark it myself for later use because of AJAX
My inability to directly link to pages also means search engine spiders are going to have an incredibly hard time deep crawling sites like this.
Now Endless could help this a bit by actually adding a sitemap (an HTML sitemap we’ll get to XML in a minute). However they have thousands of shoes so they are going to need multiple sitemaps and that could end up coming off a bit like a doorway page. I’m a big fan XML sitemaps but if you’re depending on am XML submitted sitemap to get your URL’s indexed your going about it the wrong way. Spiders should be able navigate all of your pages without a site map, it’s just good information architecture. From an SEO perspective you want to be able to leverage your own internal anchor text as much as possible.
There are workarounds where you can get direct links to products, by saving them emailing them to yourself, like these red Steve Madden Platform Pumps, but it’s way too difficult, and not intuitive. Although you should really try the multiple views and zoom features on that product page they totally kick ass.
So what’s Endless to do, do they totally give up the long tail because they have a really slick but search engine incompatible implementation? I would suggest they go with an IP delivery solution. Sure IP delivery is nothing more that a politically correct way of saying cloaking, but Google needs to come down of the “cloaking is evil” high horse, the New York Times gets away with it on a flagrant level anyway. Endless should serve a straight version of the site to people/bots who browse without Javascript instead of giving me this message
By serving it to people they will avoid the whole “evil croakers” moniker and be able to use the “better user experience” angle.
So what’s the takeaway here. With the current state of spider development using an AJAX or other JS based navigation is an extremely risky tactic. You can overcome it with traditional marketing and brand awareness, but you would be taking a huge financial risk cutting yourself off from deep crawling and long tail searches. If you insist on using AJAX or JS use the proper detection tools and serve up a machine readable version of the same content, and don’t depend on XML sitemaps to overcome usuability, design, or information architecture issues.
Sphere It
April 21st, 2007 at 10:53 am
I got side tracked by all the pretty shoes. Am I the only woman who says “Yummy” when I see pretty shoes like this one: http://www.endless.com/dp/B000LUWIOK/ ? LOl
I digress… here is somethng that I would add to your solution:
Each item does have an individual url(such as the one above), so why not add both a Breadcrumb Trail and a “Link to this page” to each page… of course we would also follow this with Keyword Rick Anchor Text in the link. This could easily be done on the back end. And that combined with the xml sitemap (and ip delivery) should help this site.
Now I must go back to spending more than I should on shoes I will only be able to stand in for a few hours.
April 21st, 2007 at 11:37 am
Can’t handle the bow… but these are indeed yummy - thanks for sending me back Natasha - just purchased those. :/
>>>Each item does have an individual url
But figuring that out is near impossible. However, as a consumer with JS enabled, they’re perfect, so they need to find SEO balance without killing the current functionality.
April 21st, 2007 at 11:38 am
The link also didn’t show in my comment above that I purchased the *black* ones, not the white ones. But, anyone who knows me likely would have guessed that.
April 21st, 2007 at 1:37 pm
@RAE: You started it - lol. I love the bow. Maybe they could figure out the urls from their database and dynamically insert some portion of the product description into the anchor text.
April 21st, 2007 at 7:44 pm
Endless.com has thousands of noisy urls (12K+ on Xenu and counting) and a weak IBL profile. Reduce the noise and double up on link pop and things will improve.
Cloaking isn’t gonna help with an AJAX site with all the content on one page. To get a page crawled you need links pointing at it and in this case there’s nothing pointing at the “pumps” page, for example, is there?
I still don’t get why the “men’s shoes” page is represented by a url like
http://www.endless.com/b/241993011/ref=topnav_sd_mn_b/?deptLanding=1
instead of endless.com/mens-shoes/
I’ll have to email my sister the url to this site so she’d stop blowing $500 on shoes.
April 22nd, 2007 at 10:37 pm
Here is your answer:
http://www.ragepank.com/articles/45/ajax-is-here/
I would quote the above article, but it’s too technical for a beginner like me.
April 22nd, 2007 at 11:09 pm
Justin, that’s an interesting solution, but it doesn’t apply to endless.com because the “pumps” page doesn’t exist. In other words, if there’s no link, you can’t convert it.
April 23rd, 2007 at 12:04 am
Dynamic sites are a bummer. You sure you just dont like to look at the shoes? I have to admitt for once my girlfriend was interested in SEO, more shoes please
April 23rd, 2007 at 11:28 am
I happen to hate women’s shoes, my wife’s are stuffed under the bed and spilling out all over the room. The cat hair globs in between them. She probably only wears 3 pair but just can’t get rid of the rest. Friggin’ women and their shoes! ;-(