This week's Webmasters hangout has a few more points than normal, and worth noting, and these are John's reflections, in my words, and some of my comments, below the video.
Here's the hangout if you want to watch it.
You won't get any help from Google as far as quality algorithm is concerned. I guess we probably knew there'd be no specific help and assistance.
Google crawls CDN files, despite suggestions otherwise. Watch out for robots.txt blocking crawling.
Aggregation: From what John Mueller said, lots of aggregated content won't help your site for quality. There's no need to mark aggregated content up especially.
Scrapers, and canonicals: Would probably only help against incompetent scrapers. Incompetent scrapers are, according to John, recognised as just that. Although, he didn't mention detection, he just agreed that scrapers are just that, scrapers.
SSL ranking benefit? Where to put the certificate? John says, put it on the whole site, not just on the ecommerce checkout part of the site. John didn't say anything about a ranking benefit, only that it won't affect the site if the certificate is across the whole site.
Affiliate site ranking: Copy and pasting feeds is classed at low quality and thin content. Lots of affiliate sites out there, and lost are just lazy, copying content. Affiliate links don't count against a site.
URL consistency: make sure that all urls are consistent across the site, including mixed case urls. Keep it consistent, and any changes to make it consistent will just take time until it is re-crawled.
Analytics: Google doesn't use any analytics information in crawling, indexing or ranking a site. Some analytics information is used in general understanding, but not specific site's in general. There is an option in analytics to turn on/off the share option.
Googlebot: Pagination on long lists, next and previous button. Next button should not exist when it's the last page to avoid google clicking and listing the next, and next, and next page, etc., even if there is noting there. It could go into thousands.
Panda slow roll-out - it's not moving slow to confuse webmasters. Slowness of roll-out is mostly for technical reasons.
Hidden content on a page: If it's important, make is visible.
302 vs 303 redirect: Google only differentiates from temporary redirect or permanent redirect. Temporary redirects: google may retain original target in the index.
Hacked sites: Once a site is fixed after a hack, how long will it take to recover the rankings? Google recommends a 503 while fixing is in process. Google recognises that is temporary, and googlebot will come back later and will not change original pages. If a 404 is shown, googlebot will reindex what it finds. Don't leave a 503 on for too long, although specific time-scales when a server is down could be up to a few weeks.
Panda again: Google confirmed again that its a slow roll-out for technical reasons.
RSS feed: Use RSS feed for whatever purpose, and google will extract what it wants.
Knowledge graph markup: Typically allow two-three weeks for the KG markup to be reflected in changes.
Analytics data on spammy pages: Google blocking spammy data messing up analytics. No news on this, apart from it being cleaned up.
Google Panda: Is being run on a site-wide basis, so block low quality pages from being indexed if you don't want to affect your high quality pages.
Site query: Not accurate when site canonical is not consistent, or part of the site is https - use Search Console
User comments on a site: User comments can contribute to the quality of a site. Quality comments: find a way to highlight quality comments. This is interesting because high quality comments ought to be featured. Imagine a click through to see additional comments, should they be low quality. Sounds a little like on-the-page cloaking to me.
Search console accuracy was queried - not sure if there was a specific answer.
Search console data: Extend data to more than three months? Adding additional storage, and retaining "snappy" performance is a challenge and requires a re-think of design. Short term, no intention to enable more than three months. Download the data if you want to retain older information.
Search analytics API may soon be finalised so that someone may develop a way to automatically download the data.
Thanks - it took a while to document it, but it was worthwhile as it was one of the better hangouts. Sometimes, the hangouts are not saying anything we didn't already know, or are answers that aren't really answers.
Thank you. Okay, so who is willing to test the affiliate links don't affect a site bit? You go first, let me know. One? Two? Three? Ten? If there isn't a threshold, then please do tell. I've always been under the impression they hate affiliate link sites. I understand from a fair competitive standpoint that they would want to announce to the world that a webmaster can "have at 'er" with affiliate links. Imagine if they said affiliate links can work against you. Meanwhile of course, those Adsense ads will be fine. Yeah. Okay. But back to my main point. Who here is going free will and posting affiliate links at will? With complete trust that it's all fine and that there is no threshold? I'm curious because for me this IS a big deal. Wasn't there a term "affiliate sites"? I may watch the video for the affiliate link mention, but I'm curious who here trusts the word that affiliate links won't affect your site. I think honestly that Google could say anything, like post affiliate links on your site without concern, but there would be no real way of pinning drops in traffic to your affiliate link usage. Afterall, it could be one of 10,000 different tweaks and algos that are run over the course of a year.
Been running 8 affiliate sites for over 4 years now (few hundred pages each), many of my pages have 10+ affiliate links. I outrank Amazon and Wikipedia on a huge number of terms, gaining rankings every year. Don't know if affiliate links are a negative factor, but they certainly aren't one for me.
I doubt there's a magical number- more likely a ratio of content to affiliate links on a page. If you have 10 paragraphs about widgets then an affiliate link to Widgets R Us, I can't see how that would be a problem. One sentence about widgets copied from Wikipedia and 10 affiliate links- probably not okay.
Most likely quality comes into play as well- is the content valuable for visitors or just stuffed with keywords to get people to the page so they can link out? MFA can stand for "made for affiliates" just as easily as "made for AdSense."
Depends on whether your site looks like it's a fully fleshed out content site that just happens to have some affiliate links, or a primarily affiliate link site that just happens to have a bit of content on it. I've had no problem with the first, and I've tried the latter and it didn't work. So mostly, use common sense.
MrSavage, a site with affiliate links isn't necessarily an "affiliate site," any more than a site that runs AdSense ads (as many big-name media sites do) is a "Made for AdSense site." The proof of the pudding is in the eating.
For what it's worth, our site does very well in Google, and it has plenty of affiliate links. IMHO,a good test of whether a site is an "affiliate site" is to determine whether the content is original and would have intrinsic value to the reader if the affiliate links were removed.
Thanks for the responses regarding whether affiliate links in any quantity are a risky strategy. I may be leaning a bit too much on how Cutts mentioned "thin" affiliate type sites in the past. Unfortunately I've never felt that I could monetize affiliate links to the extent I wanted to because of risk to rankings in Google. Sure, my sites are thin to Wikipedia. But "thin" is subjective and thus, to me it's a risk. I do appreciate the insights and hopefully this will allow me to relax a bit on my more quality laden :-) sites.