Welcome to the Q&A Forum
Browse the forum for helpful insights and fresh discussions about all things SEO.
Yep when/if Google shut down search operators there are a lot of processes that will be FUBAR!
100% agree a full ecommerce related Technical Audit is a must as a starting point. WQe recently saw a spider trap issue that was preventing a high percentage of a clients product pages being crawled and indexed. Resolving was a major 'quick win'.
Thanks for weighing in Everett. Appreciate it!
Calling all those with ecommerce SEO chops!
What are the high impact tasks that you would always start out your new campaigns with?
I've got this far in my thinking:
Hoping this can be one of those discussion threads that is so useful people bookmark it for future reference!
Many thanks!
Hi Dana,
Thanks for helping out! It's an ecommerce site.
You mention that "content not on Page 1 is not being crawled or indexed by search engines". Wouldn't the pages still be crawled and the products that they list still be crawled and indexed? It's just that Page 2/3/4/5 etc wont be indexed because of the canonical? If this is the case the only negative effect I can see is that if the product combination on a page 2/3/4/5 happened to convey relevance for a particular search query then the page wouldn't be returned in the listings?
Hi,
I have a new client that has pagination handled incorrectly on their website....
They have it setup as follows:
rel=canonical from page 2 to page 1
rel=canonical from page 3 to page 1
etc etc
i.e. they aren't using rel=prev, rel=next
To get them to invest in the development time need to change this I need to explain to the client how what they have is negatively affecting things? Anyone?
Thanks in advance!
Thanks for joining the discussion Everett!
The ecommerce site in question currently has most of its product pages noindexed. This was implemented back in 2013 because the pages either have no product description or just a single sentence. The noindexing did bring about a Panda recovery at the time.
The site is currentky being redesigned and relaunched and a decision has to be made re. whether to leave the noindex on product pages in place. Three years down the line from when I recommended the noindex and three years of listening to Google's rhetoric re. providing content that gives quality user experience and I'm now looking at these pages and thinking.... "You know what....these pages would actually be a quality search result for people. Anyone searching for what they're optimised for (the specific product name) is likely to have the intent to buy. These pages allow you to buy the product that you want!. A product description is pointless. The searcher knows the product and wants to buy it and this page allows you to buy it from a site that is well known and trusted in it's niche!"
Because this is obvious to me I was hoping that it is now obvious to Google's quality filters and I can remove the noindex and get more traffic. Perhaps not....
I have indeed, although it was more because all they carried on the page was literally a few lines per page and no value-add. Most of what was there was copied from the manufacturers site. Interesting, would you say that the pages were still a good quality search result for those using a product specific search query? i.e the searcher was able to fulfil their search intent and buy the product?
Perhaps I give Google too much credit...
I'm hoping that Mozzers can weigh in with any recent experiences with eCommerce SEO.....
I like to assume (perhaps incorrectly) that Google's 'Quality' filters (formerly known as Panda) have evolved with some intelligence since Panda first launched and started penalising eCommerce sites for having thin product pages. On this basis i'd expect that the filters are now less heavy handed and know that product pages with no or little product description on them are still a quality user experience for people who want to buy that product.
Therefore my question is this...
Do thin product pages still need noindex given that more often that not they are a quality search result for those using a product specific search query?
Has anyone experienced penalty recently (last 12 months) on an ecommerce site because of a high number of thin product pages?
Hi Dan,
Thanks for weighing in on this! Appreciated.
Judging by Dejan's study I was actually using suppression to mean a penalty of some sort as that's what it looks like on the pages that had the redirects pointed at them.
Why would the pages with the redirects pointed at them drop? Even if Google chose not to pass the signals through because of a lack of topical relavancy they have not lost any signals, just not gained any....?
In an interesting study by DeganSEO titled 'Negative Impact of 301 Redirects - A Case Study' a drop of rankings was observed when popular blog posts were redirected to product pages. One hypothesis is that the suppression is due to topical difference between the redirected pages (blog posts) and the target page.
The topical difference issue is an interesting one when you consider it in the context of website migrations. We always recommend that 301 redirects are done at a page level and that if an equivalent page doesn't exist to just 301 anyway but to the most logical page. If you think about it Google are likely to frown on this because
a) it's not a good experience for the user - 404 would be more accurate for them
b) it's lazy - if you have good content that has gained authority/trust then create the same content on the new site don't trytp pass that to an entirely different page.
Thoughts? Experiences?
I hear you loud and clear re. studying different query spaces.
Do we know that in those query spaces where social shares correlate it is not actually down to links i.e can we remove the influence of links from the studies?
Thanks Josh. Great video. This would match up with what SearchMetrics concluded in their ranking factors study.
I'm astonished by the number of SEOs that assume social shares are a ranking factor. You've only got to look at Moz's survey to see that people still think so despite no-one (that i've seen) having conclusive evidence.
It kind of feels like a share should help rankings so that might explain Moz's study of SEO's. BUT we all know that a share is a hell of a lot easier to get and much easier to manipulate than even links. If Google arent looking at them I guess this explain why.
Thanks Zoe!
Moz are careful to point out that their social results are correlation and likely to be because of a link between content that is shared heavily also achieving links (the real cause of good rankings).
i know that the recent Buzzsumo report highlights that this is not necessarilly the case but still...
Thanks John, I'll take a look at the video. When you say "social media does have an SEO value" do you mean that you believe that shares directly affect rankings? If so what evidence have you used to come to this conclusion?
I'm currently researching this area in order to show to a client that social shares aren't as valuable for SEO as they might think. Can anyone point me in the direction of the best studies done on this topic?
Thanks in advance!
Good point David. I will be digging into things further rather than taking things on face value!
Hey Scott,
Thanks for taking the time out for a comprehensive response.
Because my clients press releases have contained interesting and insightful data the volume of the links have come from lazy publishers that have taken/scraped the press releases from newswires and then re-published them verbatim.... including the brand links that have been included in the bio sections. This has resulted in lots of brand links.
I think i've been looking at it from the point of view that a press release with a branded link rather than an exact match link is OK but as you say "press releases are meant to get coverage and not _to get links directly". _Maybe Google's problem is with the fact that my client included a brand link in their press releases where it wasn't necessary or providing any value. If the links had been to 'find out more' type info perhaps this is a different matter.
I've been reading a couple of posts by the guys at Branded3 (here and here. Thanks for sharing Stephen Kenwright) that touch on how brand links in press releases were flagged as problematic by Google. Which helps to confirm suspicions that it is the press releases that have brought the penalty.
I'm wrestling with something that I'm hoping members of the community can provide input on....
I've working with an enterprise level client that is in the business of data capture and distribution.
I've diagnosed a clear drop of traffic on May 22nd, i.e a loss of search visibility post Penguin 2.0.
Their link profile is big! Discussions with internal stakeholders who have been with the company 10's of years confirm that no "link building" service providers have ever been hired and no over-zealous employee is ever likely to have tried to "do" link building internally. They are just one of those lucky companies that by their nature publish information that people want to link to and share.
As a first port of call I've grouped links by anchor text and can see groups of hundreds of matching anchors based on their brand URL and specific page titles. The matching anchors have resulted from big take up of interesting data that they have marketed via press releases. NOT for link purposes.
My question is this....
Does the community think or have evidence (or can point me toward any case studies) that show that Press release syndication alone could result in:
a) a penguin penalty or...
b) a devaluing of press release type links during Penguin 2.0 that could have resulted in a loss of search visibility and give the impression of a penalty
Your thoughts are much appreciated!
Hey Miki,
Great suggestion for product pages themselves but do you see it working for product list pages (product category pages) ?
Suppose we could have the shop owner running through and giving a brief intro to the different types/brands that have 17 inch bike frames and then produce a transcript of that.
When working on ecommerce websites we insist that product list pages need at least 250 words of copy that is optimised for our keyword phrase ... lets say "17 inch bike frames". So we have some crappy copy written that goes something like this....
"We have a great 17 inch bike frame for you whatever your requirement. Take a look at the frames below .... blah blah blah totally pointless text blah blah blah........."
This text is of no use to the user as the page is merely a means of them getting to a suitable product page. However, the copy is pretty essential if we want to rank well for "17 inch bike frames" and not having copy on product list pages could land us in hot water with Panda ...especially if we have lots of them on a site using the same page template and with no copy on them.
Does anyone else feel uneasy with adding this crappy text to pages? It's only there for search engines and that is something that Google say's we shouldn't do but I know for sure they're not going to rank me as well if I don't have it.
I'd be interested to hear other people's opinion on this. It's always annoyed me. Does anyone have any good tips for making this type of copy on product list pages less forced and crappy?
Hey thanks Irving,
The distinction for me here is that over-optimised links / on-page is being penalised as you say 'on the fly' rather than seeing penalties corresponding to Penguin Updates like in the past.
My question to people is ...do you think it's likely that separate Penguin Updates are a thing of the past (hasn't been one sine start of October) and 'Penguin like' ability has now been merged into the main ongoing algorithm?
Hey guys
We think we have been seeing some over-optimisation penalties outside of Penguin updates. One possible penalty seems for over-optimisation on page and one penalty for a page with an over-optimised exact match link profile.
Does anyone else suspect, or have seen word elsewhere, that Google's main ongoing algorithm now has Penguin like capabilities and is able to bring over-optimisation penalties without a separate refresh being run?
Hi everyone.
I have a client who has a lot of sponsorships etc and therefore a lot of inbound image links (many of them sitewide). Unfortunately most of these don't have alt text, and [img no alt-text] links now make up over 50% of their link profile.
Should I be trying to correct this and requesting updates from the people who are linking? Obviously I wouldn't want loads of keyword stuffed alt texts, but maybe I should request alt texts on the brand name or URL instead.
Do you think this would make a significant difference and be worth the time it will take to contact all these webmasters?
Thanks in advance.
Something I always do to assess how easy/difficult it will be for a clients site to rank against competitors is to compare domain and page authority scores with the scores of the competitors on page 1.
This is great when Moz have spidered my clients site and can provide scores but what about where a clients site is brand new with only a few natural links? Anyone suggest a ball park figures for Domain authority on brand new websites?
I guess as with most things SEO .... I'll test it.
Thanks both for joining the discussion.
I'll post results ... if indeed there are any.
Hi Matt,
Thanks for confirming my thoughts.
What I forgot to mention is that the 2 links to same category pages is being done from the sites main navigation (i.e not just an issue on home page). Therefore, i'm thinking that when all this lost PR is added up it could be incredibly damaging to the ranking potential of the category pages. What would you think?
I am working on a site that has a home page containing 2 links to each category page. One of the links is a text link and one link is an image link.
I think I'm right in thinking that Google will only pay attention to the anchor text/alt text of the first link that it spiders with the anchor text/alt text of the second being ignored.
This is not my question however. My question is about the page rank that is passed to each category page.....
Because of the double links on the home page, my reckoning is that PR is being divided up twice as many times as necessary. Am I also right in thinking that if Google ignore the 2nd identical link on a page only one lot of this divided up PR will be passed to each category page rather than 2 lots ..... hence horribly watering down the 'link juice' that is being passed to each category page??
Please help me win this argument with a developer and improve the ranking potential of the category pages on the site 
Thanks once again Brian.
I have double checked the header response and the redirect from "index.aspx" to the international pages is definitely a 302.
I'm hoping that Google is not assuming that the 302 is intended to be a 301 as I want the link juice to stay at the root domain and I assume this is why 302s were used by whoever setup the IP redirects to international pages.
Thanks Dr. Pete,
I am comforted slightly by what you say about Google having a strong preference for the root and that you've seen it before. With the volume of links coming in to "index.aspx" you can see why seeing the root in Google listings made me nervous that all of those links were being ignored.
This comment also comforts me..
301/302 chain could exacerbate the problem, although it would typically leave link-juice after the 301
as that is what I would expect too.
I am hesitant to make changes because of the risks it involves and might just have to accept that I'll never know for sure whether those links to "index.aspx" are definitely being counted.
Thanks again for giving this some thought and trying to help!
OK, I left out some info that I thought would complicate my description of setup but I wonder whether it is pertinent to the problem....
The site has international content areas and is geo-targeted using the directory method i.e /en-gb, /en-au,, /en-us etc
I have mentioned that www.domain.net does a 301 redirect to www.domain.net/index.aspx but I did not mention that www.domain.net/index.aspx then 302 redirects to the relevant international landing page based on IP address.
I'm pretty sure that the Google spider crawls from a US IP address so I believe that what it will see is:
www.domain.net 301 redirects to www.domain.net/index.aspx
www.domain.net/index.aspx 302 redirects to www.domain.net/en-us/index.aspx
Could the 302 redirect from www.domain.net/index.aspx be responsible for Google keeping www.domain.net in their index despite the 301??
Also, and concerning me more, do you think the considerable link juice to www.domain.net/index.aspx is being lost in all of this??
Thanks for your response Brian,
it could be that most of the links are initially directed at www.domain.net.
According to Opensiteexplorer www.domain.net/index.aspx has more links than www.domain.net though?
Where is your non-www domain directed? Is it directed to www.domain.net or or www.domain.net/index.aspx ? If it's directed to the former then that could explain some or all of your issue.
domain.net redirects to www.domain.net/index.aspx
I have taken over the SEO for www.domain.net.
The way i've inherited the setup is that www.domain.net is 301 redirected to www.domain.net/index.aspx Looking at top pages and linking root domains in Opensiteexplorer I can see that
www.domain.net/index.aspx has 1,006 linking root domains www.domain.net has 806 linking root domains.
I assume that www.domain.net is passing the value of it's 806 domain links to www.domain.net/index.aspx via the 301 redirect and because of this would expect www.domain.net/index.aspx to be the strongest page on the site and be the url that ranks in the listings for many relevant searches.
It appears however that www.domain.net is what is shown in listings and not www.domain.net/index.aspx ??
Can anyone explain why this might be??
If I do a site: search in Google then www.domain.net is indexed and not www.domain.net/index.aspx ??
OK. I'm working on an international site.
The site is setup with folders for UK, US, AU e.g www.site.com/UK/index.aspx
The root (non folder based) is the international version of the site e.g www.site.com/index.aspx
www.site.com/index.aspx has the lions share of links.
Therefore, the pages immediately linked from www.site.com/index.aspx have page rank distributed between them.
My UK, US and AU home pages are linked via a country selector from the www.site.com/index.aspx page via an aspx redirect page that 301's to the appropriate country home page. Therefore the home pages of UK, US, AU are recieving some of the 'juice' that is coming in to www.site.com/index.aspx (but only a fraction via the redirect links)
Am I right in thinking that pages on the international version of the site will have much more potential to rank (because of their 'juice') than the pages on UK, US and AU versions of the site?
If so, am I right in thinking that these will tend to rank over the equivalent UK, US and AU versions of the pages in each country version of Google despite having set directory level Geo-targetting in GWT?