Usually the Supplemental Result is attached to content from an older version of the page. Indeed, it can show words in the snippet that no longer actually exist on the page, but are still stored in Google’s “snippet database” (my terminology), and still rank for those words, even though those words are NOT seen in the attached cache copy.

This behaviour has been around for several years, but has eveolved in the last few months and is continuing to evolve. There is a ton of info on WMW threads in the last two years on this topic.

http://www.sugarrae.com/blog/sugarrae-flash-game/

and pointed links to it from my blog. It only took 3 days for that page to be re-included in the main index.

Coincidentally, the page’s cache is exactly 3 days old.

BTW, I don’t think anyone is claiming PageRank is the only factor. But if you got a TBPR 2 site that doesn’t stick, the reason is obvious.

Also keep in mind, even if you don’t do any kind of link building, if domains linking to domains linking to you publish more pages, delete pages, gain links, or have shifts in how Google evaluates their links, the amount of PageRank landing on your site can shift on a daily basis.

For more supplemental info, read my latest post:

Got Supplementals? PageRank is Only the Beginning

Personally I’ve always regarded it as an issue affected by both domain and page/query. Though as the nature of how/why a page appears in the supps has changed, the presence of a page or group of pages in the supps still appears to me to be related to both domain issues and page/query specifics.

I own sites that have lots of pages with decent TBPR that are in supp, for reasons more along the lines of DD pointed out. Earlier this year, we ran another test, and broke a bunch out just to see if we could easily, and we did. The issue, and what we did to resolve it, had nothing to do with PR (though PR can and does play a role).

supplemental index is solely a function of link strength, or how much PageRank it has at any given moment

Now, again, as I said in my post, I haven’t watched supplemental very closely in a long time. That said, we had a site we overhauled about two months ago with about 65% of its pages in supplemental. We changed the content – kept it on theme, but changed all of the wording and gave everything unique titles and metas, along with new content on the page itself. These pages, as they began to get re-crawled, popped from holding the supplemental tags. We didn’t embark on a new link campaign. That said, we did change the internal linking structure to make more sense to users and give us more value with the spiders. As more and more pages dropped the supplemental tag, the site kept moving up and up in ranks. We also have other sites utilizing the same internal linking structure, with the same average strength in links that have a large portion of pages still in supplemental (older sites we haven’t touched in a long time). So, something else for me to think about.

I’m definitely not saying links has nothing to do with it – but I’m hesitant to agree to it being the sole/only factor without some of my own testing. Nothing personal – I’m the first person to say “test everything you hear”. As I mentioned, I haven’t paid much attention to this in the last year, though I’ve read a lot of posts complaining about it. I never thought it would be a problem that I’m a strong and agressive link developer, but apparently it has kept me from having to pay much attention to supplemental if your take on it is correct (links being the sole factor) – and I’m not agreeing to agree or disagree at 2 am. :P

Also, Michael – am I to take it that in your opinion, supplemental is acting the same today as it was say – two years ago?

It makes sense, but I think you missed what I was saying. Pages can exist in both the supplemental index and non supplemental index at the same time. It’s not that the page has been tagged both ways, it’s that there are actually two copies of the same page.

Pages don’t get tagged as supplemental per se, it’s not a field in the database that acts as a flag… it is a separate database (well, set of servers holding the database) altogether. Whether or not a page exists in the supplemental index is solely a function of link strength, or how much PageRank it has at any given moment.

It isn’t hard for many pages on your site to be considered as too close to the others for your domain name.

As far as I know that is a misconception. The concept of “too close” doesn’t feed into whether or not a page is supplemental (I know, many people will swear otherwise). If you see pages in the supplemental index, then at some point Google determined that there wasn’t enough PageRank to warrant them being in the main index. It should be the case that once it moves them into the main index, it no longer returns those supp pages… but the Google datacenters are never 100% in sync, as we all know.

I think what you are seeing now is pages that were purely supplemental from the get go, which is normal depending on your site structure. For most blogs, you wouldn’t see this behavior, since the newest content is found from a link near the top of the homepage… in fact, you would tend to see the reverse, with the older a post gets the further it gets from the main link juice, and the more danger it has of going supp. But with regular sites, unless you take care to have multiple focused entry pages, each with their own incoming link supply, and make sure to build off of those nodes, then often times the newer pages start out deep to begin with.