Thanks Cyrus

If external links are a ranking signal, do you think there would be a difference in perceived value whether external links are noFollow or doFollow, or do we expect that to make little difference?

Cyrus, I was actually looking to answer the statement you mentioned, "though I don't believe we've ever studied the difference between followed and nofollowed in this regard".

We've a really popular post on our site which lists hundreds of Twitter chat hours and links to the Twitter hashtag and the host in each case (http://tillison.co.uk/blog/complete-twitter-chat-hours-directory/).

Across the team, we're disagreeing whether all external links in the post should be nofollow or whether they should remain untagged and therefore dofollow. On the one hand, it feels like we're leaking page equity through every link and want to retain it, of course. On the other, nofollow kinda feels like we trust none of those links and that the page may be less valuable to the Googlebot.

I'm working through the links making them all nofollow, but would be really interested in your perspective on it.

Thank you Sir. I think we reached the same conclusion.

By the way, the it was a just a simple example of the page hierarchy - we're not doing Horror Books

Good. Same page

I was looking in to rel=contents and those variations before, but I can't quite decide whether this is worth the effort or not.

e.g. There's a huge list of resources on a single page, segmented in to categories. The page is HUGE and takes ages to load, so I've been creating new pages for each segment and optimising those pages independently, but there is some common content with the primary page.

V1: Horror Novels page has a section for each author, each section lists all novels by that author.

V2: Each Author has a page which lists novels by that author, but links back to the Horror Novels page which is essentially an index of the Author pages. Would you also

Would you use rel=contents, rel=prev/next or a different approach in this case? From what I've read so far, there doesn't seem any "SEO value" in linking that way.

I guess we're trying to improve the UX through faster load times and segmenting the information in smaller chunks, but also presenting a number of pages to Google as a body of content rather than a single page without causing issues with duplicate or similar content - we just need to make sure that we're optimising it in the right way, of course.

Thanks Pete

So, for a more specific example, if an eCommerce store has an "email this product" page for each product (Magento seems to love doing this and creates a duplicate of the same email page for every product), would you recommend a canonical link for each of those pages to the main Contact page or canonically linking each page to each related product page?

From setup, I'd consider NoIndex on all of those pages anyway, but it's a bit late for that once a site has been live for years.

The email pages are obviously related to the product page, but the content there isn't anywhere near identical.

Or maybe there's a "more appropriate solution" that you alluded to?

Hey Pete

Can you explain, "you can't use rel=canonical on pages that aren't 100% duplicates" a little further please?

Do you mean that only duplicate pages should be canonicalised? Identical pages in two different sub-directories is fine, but two similar pages is not?