"noindex, follow" or "robots.txt" for thin content pages

[khi5]

Does anyone have any testing evidence what is better to use for pages with thin content, yet important pages to keep on a website? I am referring to content shared across multiple websites (such as e-commerce, real estate etc). Imagine a website with 300 high quality pages indexed and 5,000 thin product type pages, which are pages that would not generate relevant search traffic. Question goes: Does the interlinking value achieved by "noindex, follow" outweigh the negative of Google having to crawl all those "noindex" pages? With robots.txt one has Google's crawling focus on just the important pages that are indexed and that may give ranking a boost. Any experiments with insight to this would be great.

I do get the story about "make the pages unique", "get customer reviews and comments" etc....but the above question is the important question here.

[khi5]

trung.ngo - check out this article I posted http://www.blindfiveyearold.com/crawl-optimization

that's where I got my "inspiration" from to consider using robots.txt instead...

[khi5]

I am thinking if I exclude more thin pages from being crawled (robots.txt) that may be better than my current "noindex, follow" - the thin pages are already "noindex, follow".

You are saying "unless there's evidence that the pages are taking up too much of the crawl bandwidth, it doesn't seem like too much of an issue to me." - but how would I know this? Fair to assume for a website with 5,000 pages this is probably not an issue?

I am concerned with the "noindex, follow" Google may think "ahh, we have seen all this stuff before. Thanks for keeping out of our index, but we are still going to devalue your original content indexed pages because we crawl and see all this thin stuff." I am thinking with the robots.txt it would potentially be a stronger signal that could help my indexed pages. Or you think it is a minor and probably not relevant?

[trung.ngo]

Hello there,

Have you had any duplicate content or crawling issues in the past or is this more of a preventative measure? If the pages, as you put it, "would not generate relevant search traffic", then I would argue that it'd make sense to "noindex, follow" based on the assumption that the pages are not currently driving search traffic, and have no real potential to contribute significantly to brand discovery via a search engine in the future.

I wouldn't necessarily say that Google crawling your page more frequently would automatically give you a boost in rankings; it's more associated with whether or not they're crawling pages frequently enough to index updates to the pages. So unless there's evidence that the pages are taking up too much of the crawl bandwidth, it doesn't seem like too much of an issue to me.

All of this to say, take a look at the data to see if a real problem exists--whether crawl resources or duplicate content--before doing anything drastic. And, of course, also understand what you'll be losing by making the updates. If you do choose to prevent crawling via robots.txt and are at all concerned with the duplicate/thin content aspect, remember to implement a noindex and confirm that the pages are removed from search results before disallowing in robots.txt--otherwise, they'll remain indexed.

[khi5]

Hi Keri, There are some good comments but none really answer this question and that is why I am trying to approach from different angles. Maybe you can shed some light on this:
AJ Kohn wrote this great article: http://www.blindfiveyearold.com/crawl-optimization - he talks about using robots.txt to exclude thin content in order to increase frequency with qhich indexed content gets crawled, supposedly helping rankings. In this great whiteboard Friday, Rand suggests using "noindex, follow" - http://moz.rainyclouds.online/blog/handling-duplicate-content-across-large-numbers-of-urls.

I am trying to get more light on this (people who have experience with this), but struggle to get answers.

[KeriMorgret]

I noticed you had similar questions at http://moz.rainyclouds.online/community/q/unique-content-below-fold-better-move-above-fold and http://moz.rainyclouds.online/community/q/risk-using-nofollow-tag with several answers each, including some that were marked as Good Answer. Did any of those answers help to answer your question?