Supplemental Result

tag in "site:" search results

[dcheney]

I was doing a bit of research today for a term with few results. One site had 2 results so I hit the "More results from www.yyy.com" link. It came back with 8 results. The last of these entries has the odd bit.

Most results have something like this as their last line:
"www.yyy.org/blah/blah.html - 15k - Cached - Similar pages"

But the last result has:
"www.yyy.org/blah/blah2.html - 7k - Supplemental Result - Cached - Similar pages"

Anyone else seen this before?

This is purely a guess, but could this be a page that was found by freshbot but hasn't been fully incorporated into the google databases yet?

(I'd be happy to sticky the particulars of the search)

[dcheney]

Just to be clear - the "Supplemental Result" is just plain text - not a link

[Slade]

I just got this on a search I was doing for a SQL2K problem.
try: sysmergeextendedarticlesview

[moltar]

I found the following info on google.com

[google.com...]

Not much information I have to admit.

[dcheney]

Thanks Moltar - I guess its been around and I just never noticed it before.

[RonPK]

On the page moltar mentioned it says

Supplemental Result
Google augments results for difficult queries by searching a supplemental collection of more web pages. Results from this index are marked in green as "Supplemental."

So, now there are two kinds of web pages: normal, and other pages that form a 'supplemental collection'?

Dcheney, was your query 'difficult'?

[dcheney]

Yes, I suppose difficult was apt. The original query has 208 results. The site specific one had 8 results. I'd be happy to post the query, but I don't think I can because of the TOS.

Sadly the site in question (not mine) is offline - so I can't compare the PR of the various pages on the list of 8.

[markdidj]

I had two sites, one at .co.uk the other at .com. They were mirrored sites but I changed the .com about 4 months ago. I have been checking the results in Google for the change at .com but it's still indexing my old pages, now with the new "Supplemental Result". When can I expect these pages to change? How can G update the .com site's index without checking if the pages still exist? It's a bit annoying as to G it still looks as if I have a mirrored site, lowering by PR for the .co.uk.

[RonPK]

Putting 1+1 together, I'm getting the impression that the 'supplemental collection' is a set of mirrors (markdidj), pages with more or less identical content (slade), offline sites (dcheney) and missing pages (as mentioned in [webmasterworld.com...] ).

Let's hope GG drops by to shine some light...

[dcheney]

RonPK - in the case where I'm seeing it, all 8 are from the same offline site - only the last 1 appears as "Supplemental"

[mfagan]

on ResourceShelf [resourceshelf.com]:

* It's an experimental feature
* It augments search results for hard-to-answer queries, by searching a supplemental collection of web pages -- in addition to its main index of 3.3 billion web pages. Results from this index are marked "Supplemental" because they originate from a separate, experimental index that is only used to answer the most obscure and infrequent queries.
* I've [Gary Price] asked a few follow-up questions including: how do they determine what goes in this index and what is an obscure query? Stay tuned.

[PatrickDeese]

IMHO "supplemental results" are most likely hand tweaked results for queries that the Google algo couldn't produce good results for - perhaps even problematic searches that were reported via the "Dissatisfied with your search results?" link.

[robertskelton]

I noticed it for the first time just now, so of course I came straight here.

I was looking for information on AltaVista's Enterprise search - all the supplemental results are pages that are gone, and now re-direct to FAST's site.

It could be that there are some dead pages that Google has decided are important enough to keep indexed, so we can access the cached copy, and/or visit where they used to be.

[Allergic]

mfagan : You should also search on the first statement of Gary price :

1) On our recent post about Google size estimates being off, a Google spokesperson tells us what we for the most part already knew, "We looked at your queries and the bottom line is that Google's estimator is an estimate, not an exact number...we're working to making it more accurate." Like I said a few days ago, those of you who use Google page estimates as a way of determining popularity need to be very careful.

This was his recent post [resourceshelf.com] and check the Exemple 2!

But he did not investigate deeply enough like I post it here [webmasterworld.com] with no answer at all.

Here all the links and numbers [multiforum.info] (in french but it is mostly digits and links)!

[anallawalla]

I too noticed it for the first time today. I was looking for the exact match "miller heimann" which is a kind of sales training method. The results are clickable and I can't tell what is unusual about the pages that come up, other that this is a fairly obscure topic with only 31 total hits.

[fLaMiN]

same here, just noticed it then so decided to sign up and find out what it is ..

i searched for:- google "supplemental result"

the only result has the text next to it ..

ive never seen it before, even tho you have pointed out in the interpret section that it looks like it has been there for a while ..

must be doing something if all of a sudden this many people have noticed this change

[2_much]

Yeah, I just saw that today too, came to post, and here it was. Obviously they're testing something today. I was talking to a friend and we "googled" ourselves, and when I typed my full name in with quotes, the last result was "supplemental". The interesting thing is that the page marked as "supplemental" has been in the index for a really long time. The page does contain my name, like all the others. It doesn't have any linkto's (don't know about PR, no toolbar). Most of the others also don't have linkto's, so I have no idea what differentiates it - perhaps just the tag?

[GoogleGuy]

Hey, the supplemental results are a new experimental feature to augment the results for obscure queries. This is a new technology that can return more results for queries that for example have a small number of results. So it might not affect the results for a popular search, but for a researcher doing a more specific query, it can improve the recall of the results. The supplemental collection of pages has been collected from the web just like the 3.3 billion pages in Google's main index.

Hope that helps,
GoogleGuy

[RonPK]

Thanks, GoogleGuy. However I'm stuck with two questions:

As a user, how am I supposed to interpret these results? Are they less relevant? Or is Google telling me "your query returned 5 pages, and because it's such a difficult query we'll show you another page you might want to try"?

As a SE watcher, I wonder how G determines whether a page should go into the supplemental collection or not?

Sorry for the dead link in msg # 9. It pointed to an earlier posting with the same title as this one, and was deleted by an admin.

[dcheney]

GoogleGuy,

Thanks for the reply. I suppose its a trade secret, but you did skip the rather obvious question - what criteria determines whether a spidered page is placed into the main index vs. the supplemental index?

Obviously one generally wouldn't want their content placed into the second index since it will never be returned unless its an "obscure" search.

For a site like my own its not as big a deal since the majority of successful searches are for specific, and often fairly obscure, proper names.

[HitProf]

Cached version of the page is different from the live version.

The Supplemental result is already in the cached image but not in the text :) Page changed 27 August.

[added]In the Dutch version the text in the live version still doesn't match the image[/added]

[GoogleGuy]

Hey, pages get added to the supplemental index using automatic algorithms. You can imagine a lot of useful criteria, including that we saw a url during the main crawl but didn't have a have a chance to crawl it when we first saw it.

Think of this as icing on the cake. If there's an obscure search, we're willing to do extra work with this new experimental feature to turn up more results. The net outcome is more search results for people doing power searches.

[2_much]

Didn't Google already do this before? I think the results were displayed differently. If a page wasn't fully crawled, it would only show the URL in the SERP, not the title and ransom note description. Apart from this, what's the difference?

[robertskelton]

Many resulting pages have since changed or are now dead, and they tend to come from large sites, so I figure the supplemental results are from an old index, an index that looked at every page in large sites, rather than just most of them.

My guess: whenever Google has some spare spider power, and time up their sleeves, they go find some extra pages at big important sites (like blogging sites). The data is supplementary because they never know when these pages might be indexed again, if ever. They know the data is old, so they don't include it in the main results, unless it looks like it could help.

[rfgdxm1]

I just found 8 pages on my site that are coming up as "Supplemental Result" on searches. These are all in a directory that isn't linked to from my site, and possibly is orphan. This was never intended to be generally available to all. My guess is some time back someone posted the URL on a bulletin board, Google found it then, and even though these pages are now orphan, Google refuses to let them go.

[Morgan]

I found a couple on my own as well. They're pages that were indexed, but no longer have any links to them, but they still work. Probably just an easy way for Google to keep more information available even if nobody links to it anymore.

[thoken]

The explanation given by Google and Google Guy is perhaps the kind of cautious statement required when such a critical feature like ranking by automated analysing and comparing of page contents is introduced. To me [google.com...] seems typical.

[valeyard]

Hmm. Call me Mister Suspicious but I don't buy this.

I thought Google's goal was to index the entire web. If they do that then any query can be answered, there's no need for a "supplemental" index or some strange idea of an "obscure" query.

If I search for "snark jubjub -boojum" and there's only one match, so be it. As long as the page is indexed then I don't care if it Google thinks it "obscure".

With my user hat on I can't think of any good reason for differentiating between the main and supplemental index. The Google algo is designed to put "most relevant" answers at the top regardless.

So if it's not for the users' benefit it must be for Google's own benefit.

Sounds like an excuse for hiving off a chunk of the main index. Is this another pointer to capacity problems and people trying to invent ways round them?

[mcavic]

I just did a search resulting in two pages from my site in the top two positions. The second one is a supplemental result, and it's a page that has had a 301 redirect since at least last April. (And now it's 404).

I have also seen examples that lead me to believe that supplemental results are orphaned pages that no longer have any backlinks.

As for still-valid orphaned pages, I think this is a good thing. But for 301 or 404 pages, I don't see any value to the user - those pages should disappear forever, as soon as possible.

You can imagine a lot of useful criteria, including that we saw a url during the main crawl but didn't have a have a chance to crawl it when we first saw it.

This doesn't make sense to me. It implies that supplemental results are fresher, but it seems to me that they're more stale. Plus, lately I've seen new pages being added to the main index in a matter of days. So if a page misses out on being indexed, why whould it not just wait a few more days to be indexed? That's the whole point of fresh listings.

[GoogleGuy]

"Sounds like an excuse for hiving off a chunk of the main index. Is this another pointer to capacity problems and people trying to invent ways round them?"

Hi valeyard, and welcome to WebmasterWorld! The supplemental results are above and beyond the pages that we already search. So we're not taking away any docs--in fact, we're searching even more docs than before.