A Crazy Thought: Should I No Index My Products?

[Planet13]

Hi there, Everyone:

Maybe it is lack of sleep on my part, but I was thinking; Could it be possible that meta noindexing product-level pages would actually boost ranking for category pages - even for long tail terms?

Suppose you have a category called Iron Widgets. And in that category you had several individual products, like 15 Inch Tall Iron Widget, or Japanese Forged Iron Widget, and stuff like that.

And due to the fact that their isn't a HUGE difference between one widget and another, the description of each widget product page kind of overlaps with the description of all the other widget product pages.

In situations like that, would it be possible that all the different individual widget product pages would be "competing for google's attention" with the category level Iron Widgets page?

I ask this because I see that for MOST of my site, the category level pages get much better landing page traffic than individual product pages. There are a few exceptions (i.e., a couple product pages that are big landing pages), but for the most part, it is the category level pages that rank well in the serps and get the clicks.

I realize that page rank can't be "sculpted" by using noindex, and that by noindexing a bunch of product pages, the way that page rank flows through a site would be seriously changed.

But maybe by noindexing the product pages, it will allow google to "focus" all its attention on the category-level pages?

[onebuyone]

Problem is that you will have to wait until next panda refresh/update to see real effects of those changes. If you noindex those pages now Google will see it. Your most relevant category pages will takeover positions of products, but this won't change your traffic in any way.

[Planet13]

Your most relevant category pages will takeover positions of products, but this won't change your traffic in any way.

Hmm... I am interested in why you say this. Could you elaborate further?

I am thinking if a category page were to rank, say, 5th in the serps for the term Iron Widgets, maybe noindexing the various Iron Widgets Product-Level pages would help the Iron Widgets category page rank higher for Iron Widgets?

As for Panda, I am less concerned. I haven't been hit by panda, although since I described my site as having "overlapping" content, it may appear that I was worried about it.

But there must be tons of ecommerce sites where the product pages are nearly identical except for one attribute (say, a 10 inch widget instead of a 12 inch, 14 inch, 16 inch, etc.,)

And while they may not be susceptible to Panda, maybe they are somehow getting caught up in the rest of the algorithm?

[topr8]

>>maybe noindexing the various Iron Widgets Product-Level pages would help the Iron Widgets category page rank higher for Iron Widgets?

but what about all the (internal) links back to the category page from the product pages.

[driller41]

Won't you loose actual product searches.

It is something I thought about a while ago but never followed through the idea, I suppose the best option would be to try it on a single category branch and see where it leads you.

[Sgt_Kickaxe]

Look at how big shopping sites handle this, in particular ebay since they've gained 20% in traffic this year after recent concentrated SEO efforts.

Nothing is noindex - but the best pages are brought front and center including a category page for every TYPE of item. Gone are all of the fluff pages including forums and even review pages, ANYTHING that might compete with actual category/item pages is gone. Number of fluff links at the bottom of the page are also reduced and condensed.

Yet search for an exact user generated item title and they dominate anyway, even if the page is 50 clicks into a 20,000 results category. The key is getting rid of any fluff and redundant pages, imo.

When ebay has 20 of the same items for sale, with same wording, they now show one item with links to the other 20 from that item page, not from the category page. They've condensed pages but not noindex'd them, it's working.

[Robert Charlton]

My empasis added...

I realize that page rank can't be "sculpted" by using noindex, and that by noindexing a bunch of product pages, the way that page rank flows through a site would be seriously changed

No, PageRank flow is not something you want to challenge. If you're going to noindex, which I don't recommend, you should use meta noindex,follow... not meta noindex,nofollow.

Nofollow, on individual page metas would also be a serious mistake. I can't see where it is ever helpful except when you don't trust the page you're linking to.

...but what about all the (internal) links back to the category page from the product pages.

Similarly, the nofollow attribute on nav links, which is what I assume this comment is suggesting, is generally a disaster... nothing to be gained, but lots of link juice recirculation within the site to be lost. You definitely want that link juice to circulate up vertically to the next higher category... maybe not across categories... but definitely up and down.

Suppose you have a category called Iron Widgets. And in that category you had several individual products, like 15 Inch Tall Iron Widget, or Japanese Forged Iron Widget, and stuff like that.

And due to the fact that their isn't a HUGE difference between one widget and another, the description of each widget product page kind of overlaps with the description of all the other widget product pages.

It may be that the issue is that these pages are targeting nuances that aren't really distinguished in most searches, and that they are perhaps better left to an order form or a product sub-page.

It occurs that some thoughts I just posted in a different discussion about whether or not to combine product characteristics under one url, might have bearing on this discussion here....

Dynamic Single Product Page - Can this be successful for SEO?
http://www.webmasterworld.com/google/4424746.htm [webmasterworld.com]

In the above thread, I was basically arguing against combining too many diverse product categories into one dynamic page. In the course of explaining what I wouldn't combine, I also described what I would combine. Note that these considerations aren't restricted to dynamic pages. I'm suggesting the use of form option values to specify different shirt sizes, eg. Why not also for different heights of iron widgets?

...if you were targeting, eg, a particular piece of clothing, like a shirt, you might be able to manage to include characteristics like color, size, and pattern all on one page without skewing the search targeting. These are characteristics I wouldn't normally try to target on separate pages anyway. I'd normally leave these choices to the option values in the ordering interface. The lists of colors, sizes, and patterns that would appear in your source html would in fact be expected.

The dilemma comes if your "15 Inch Tall Iron Widget" and your "Japanese Forged Iron Widget" differ in much more than height... and I suspect they do. Then they are basically different products, and they do require different pages. I think you then really need to strain to uncover and emphasize differences between them that are likely to show up in a search query, and to reflect those differences in your product descriptions, which should describe the features of the product, why the customer might choose it over another, and where and how the customer might want to use it.. briefly, but beyond simply product specs.

I should add, btw, that I know of many site owners who are straining like crazy to rank for their category level pages rather than for their product level pages, because the number of searches is much greater for the category level. I mean, do people really search for [15 inch tall iron widgets]? Is that traffic you're really expecting?

At issue here, if you can settle for category level rankings, is to enhance those category level pages... making sure your navigation on the category level pages is laid out in such a way that the choices are more easily navigated and the product choices are prominent and readily seen. Many shopping carts, eg, lay out the products in checkerboard patterns, which I feel is a user-unfriendly display. There's no place for the eye to settle.

Perhaps you'd have better luck with a list format, resembling a serps page, eg, with product thumbnails at the left of each text listing that would run in one or two or three lines across the page. This compresses the listings vertically, and simply scrolling down the page exposes each product in sequence much more clearly. The thumbnails on the left, in a column, hold your eye. Scrolling down the list is almost like navigating a mobile app. There are ways to provide separation that are attractive. If you do this, list most popular products, or the ones you want to sell the most, at the top of the list.

[realmaverick]

I don't recommend you do this. I think it's unlikely that your product pages are causing a problem. Unless there is something specific about them, that Google doesn't like.

If your website is anything like mine, despite category pages individually receiving most traffic, the product pages traffic, when added up together, far exceeds the category pages.

I agree we should noindex junk, but your product pages are not junk, they're a big part of your website and they are ultimately, the thing your visitors are looking for.

Have you suffered a loss in traffic, which leads you to believe these pages my be an issue?

Ensure your product pages titles, are structured differently to the category pages, to ensure that a product page, won't be fighting for traffic against a category page.

[Robert Charlton]

Yes, a PS to my comments above, lest they aren't clear... I agree with realmaverick. You definitely do need to have those product pages, and don't noindex them. Noindexing is a temporary bandaid at best, and I don't recommend it.

Keep the product pages, differentiate those that are for different products and perhaps combine some that aren't (eg, those that are for basically the same product but just different heights or colors)... and make the links to the product pages as accessible as possible from your category pages.

Ensure your product pages titles, are structured differently to the category pages, to ensure that a product page, won't be fighting for traffic against a category page.

An important point. I try to make pages more specific the deeper they are. The anchor text that's on category pages linking to the product pages should be less specific than the titles and headings on the product pages themselves. The product pages are where the greatest differentiation needs to take place.

At the same time, to address your intitial concern, make sure that you're not building separate product pages for things like different heights and colors, where the product is otherwise identical. Amazon may be able to get away with separate product level pages for different colors, eg, but most sites probably can't.

If your CMS lets you go a hierarchy level deeper than your product pages to target different minor differences, that's another story... but most CMSs generally don't permit such sub-product pages, and most sites don't have the link juice to support it.

[Planet13]

Wow, these are some great responses. thanks for the input.

@ realmaverick

If your website is anything like mine, despite category pages individually receiving most traffic, the product pages traffic, when added up together, far exceeds the category pages.

I will have to double check. At first glance, when I look at the top 25 or 50 pages in GA, it would appear that the category, or that article pages, are the most popular landing pages. But I will have to add up the numbers for the product pages and see how they compare.

Have you suffered a loss in traffic, which leads you to believe these pages my be an issue?

No, not so much a loss in traffic per se, although there are a few instances where amazon, ebay, or even ehow results have jumped over mine. but it is hard to determine whether that is my site following out of favor with google or whether it is googles current fascination with mega brand sites that is causing this.

But it is more geared toward the problem of why SOME certain category pages rank in the top 3, and why SOME of them rank in the 60s, or 70s, or 80s. or out of the top 100 altogether. I am trying to find a way to get those low ranking pages to rank significantly higher.

@ Robert Charlton

Keep the product pages, differentiate those that are for different products and perhaps combine some that aren't (eg, those that are for basically the same product but just different heights or colors)... and make the links to the product pages as accessible as possible from your category pages.

thanks for the tip. I can certainly do that. My only concern is that what a human considers significantly different and what an alogrithim considers significantly different are often two separate animals...

An important point. I try to make pages more specific the deeper they are. The anchor text that's on category pages linking to the product pages should be less specific than the titles and headings on the product pages themselves. The product pages are where the greatest differentiation needs to take place.

Hmm... To me that seems ALMOST like a contradiction, but that might be because I am misinterpreting your use of the word "differentiation" here.

When I see that, I think that means the individual product pages should have MORE TEXT than the category level page (which is hierarchically directly above it). Maybe I am not interpreting that correctly?

I guess the challenge for me is if ALL widgets have certain parameters / attributes (size, material composition, weight, color, tensile strength, operating temperature, etc), how does one differentiate them significantly without simply stating, "this widget is blue and has a tensile strength of XYZ"?

Thanks in advance.

[Robert Charlton]

My only concern is that what a human considers significantly different and what an alogrithim considers significantly different are often two separate animals...

Search algorithms attempt to align themselves with human expectations and behavior. Google may not have nailed it exactly for each of the c25,000,000,000 ever-changing pages it's got in the index, and it still can't understand language, but it is trying to understand human behavior, not to disregard it.

To use your example of Iron Widgets... are people really searching for 15 Inch Tall Iron Widgets, vs say 14 Inch Tall Iron Widgets? Or are they looking for other distinctions?

My guess is that if you were selling hard drives, customers would be looking at the difference between 250-Gb hard drives vs 500-Gb drives, or with tires they may be looking at physical dimensions and properties and brands. Chances are with some kinds of widgets, perhaps yours, they may be looking at distinctions other than just size.

Are you offering any kind of verbal descriptions of your widgets and how they might be used... or are your widgets so specific that there's no point saying anything about them except how tall they are and what they're made of?

If the distinctions are purely visual, are you offering different views of your products on your site, perhaps with jquery viewers? Though that's not text, there is an engagement factor that Google might well consider. You may have to think outside the box, though, to figure out what's distinct about each of your products.

...The product pages are where the greatest differentiation needs to take place.

Hmm... To me that seems ALMOST like a contradiction, but that might be because I am misinterpreting your use of the word "differentiation" here.

When I see that, I think that means the individual product pages should have MORE TEXT than the category level page (which is hierarchically directly above it). Maybe I am not interpreting that correctly?

Yes, you seem to be misinterpreting at least what I'm intending to say.

Product pages should have more descriptive text about the products, but... without being too sparse... they're specific to specific products. The category pages are basically site maps with some introductory text and then text fleshing them out so they're not just lists. They are your tables of contents. They distribute visitors and search engines to your product pages. Some category pages are liable to rank, but because they're more general than your product pages, their likely target terms are more competitive.

Different sites are structured differently... so don't just copy this scheme... but if you look at WebmasterWorld, and the various pages leading to the Google pages, you'll see that....

- on WebmasterWorld, the front page is the default index page. Note that the distribution for search engines and for visitors is directly parallel. The page is essentially fairly even-handed about how it distributes link juice and directs the visitor, but, because there's a time dimension to the discussions here, it emphasizes things that are currently important... again for both search engines and visitors.

- Home (http://www.webmasterworld.com/home.htm) is a site map to smooth out the flow...

- Google ( http://www.webmasterworld.com/category30.htm ) is a main category page, basically a site map with a bit of descriptive text under each menu item...

- Google SEO News and Discussion ( http://www.webmasterworld.com/google/ ) is a sub-category page, listing all the threads...

- and each of the threads has, relatively speaking, a lot of text. These are the equivalent of your product pages.

You get inbound links to all of the above, either linked directly, or redistributed from direct links into the threads.

Note that ecommerce product pages are different from forum threads. Ecommerce is going to be less longtail than forums are, but ecommerce product pages are still longer tail than ecommerce category pages are.

You should have a rough idea in ecommerce what your product searches are likely to be. Because the searches are likely to be more specific in ecommerce than in forums, to compete you need to research what people search for and what they don't search for, and you should try to focus your vocabulary and content around product descriptive text (likely to be searched) and naturally co-occurring vocabulary.

Product pages should not be nearly so chatty as forum discussions, nor as expansively informative as Wikipedia pages... but they're still a lot less sparse than any of your category pages. This isn't to say that category pages should just be lists. It's a mistake, though, to think they should get into individual product details. They should not.

But if all of your product page details are identical, and there's nothing more, then either you're classifying things wrong, or you're missing some opportunities.

[Planet13]

Thank you, Robert:

There is a lot to digest there, so I will have to think about it. but I think the gist of it is that one really has to think outside of the box to make the descriptions (and the non textual content) as descriptive and unique as possible.

thanks again for the very detailed explanation.

[onebuyone]

Your most relevant category pages will takeover positions of products, but this won't change your traffic in any way.

Hmm... I am interested in why you say this. Could you elaborate further?

Between Panda refreshes/update when you take some position for some KW with page A and then you noindex it, some other page from your website will take this position even if it is not as relevant a page A.
New page will be gone with next Panda update/refresh.

[Aaranged]

Just from a logic perspective, does it make sense to hide the most detailed, actionable, (relatively) content-rich pages from the search engines? At at very basic level I'd say no.

And this strikes me as a classic bad tradeoff. That is, the price of improving the ranking of your category pages shouldn't be entirely borking your product detail pages. Put another way, chasing short tail keyword traffic shouldn't be done at the expense of your long tail traffic.

Furthermore, you may actually damage the rankings and keyword reach of those category pages for a number of reasons. If you have (sensible) breadcrumbs product pages going to be pushing and concentrating link juice upwards to their parents. If there's social interactions on your product pages they're lost to view. If there's spiderable UGC (like reviews)- often great, original, keyword-rich fodder - that's going to be invisible to the search engines. And, of course, fundamentally you're going to be losing the semantic strength of your product titles, descriptions and (perhaps) images and videos.

Some of this might be addressed with combining (no)index and (no)follow directives, but heavens knows only an SEO could up with something so convoluted for such basic pages (even with more complex day-to-day ecommerce situations like paginated results and URLs spawned by faceted browsing I generally think you're better off going big or going home - nix the pages altogether, or expose them and their linkages fully).

So, and interesting proposition but, yes, I think it is a crazy thought. :)

[Planet13]

Thank you both, onebuyone and Arranged for your helpful insights.

Furthermore, you may actually damage the rankings and keyword reach of those category pages for a number of reasons. If you have (sensible) breadcrumbs product pages going to be pushing and concentrating link juice upwards to their parents. If there's social interactions on your product pages they're lost to view. If there's spiderable UGC (like reviews)- often great, original, keyword-rich fodder - that's going to be invisible to the search engines. And, of course, fundamentally you're going to be losing the semantic strength of your product titles, descriptions and (perhaps) images and videos.

Those are all excellent points.

I think that the main problem is that my product pages are yet to be distinctive enough. They are neither truly distinctive from the other pages on my site, nor are they distinctive enough when compared to my competitors product pages.

<off_topic>
Any tips on how to get more user generated content for product pages is greatly appreciated. We only get like 5 product reviews PER YEAR, but they tend to be very good reviews.
</off_topic>

[tedster]

<off_topic>
How about direct emails requesting a review of the product, sent after a short period of time to those who made purchases?
</off_topic>

[Planet13]

@ tedster

How about direct emails requesting a review of the product, sent after a short period of time to those who made purchases?

that's the thing; I do send them out two weeks after the order but the feedback is almost minimal.

You would think that at the minimum we would get a couple of people who HATE us and we would at least get a couple of horrible reviews. Instead, we get nothing.

[Ahkamden]

For the off-topic, do you give them an incentive for filling out review? 10% off next purchase or something. Otherwise from consumer end, why waste the time I'm not getting anything for it.

[Seb7]

Could it be possible that meta noindexing product-level pages would actually boost ranking for category pages

- no.

I'm guessing you have links to the categories from the product pages; this is actually what is boosting your category pages.

I personally always recommend putting the entire product line online. But.. your titles, descriptions etc have to be directly relevant to the product, and web friendly. Linking correctly with category pages always works well.

(By web friendly; you need to think about what is the most probable search term someone would type as to find that particular product - its often not the same as the product description/name that the business has already for it. I often end up renaming 80% of the products before going live - and always had great success.)