• Register
• Log in

SEO Theory - SEO Theory and Analysis Blog

Algorithm analysis, Web community relationship analysis, SEO practices and techniques, industry news, etc.

• Home
• About SEO Theory
• SEO Glossary

Supplemental Days and Supplemental Nights

Posted by Michael Martinez on July 27, 2007 in Supplemental Pages

A few weeks ago Matt Cutts suggested that it might be a good idea for Google to remove the “Supplemental Pages” tag from search results. Although his proposal might have seemed like a good idea on the surface to me it seemed like it might be disingenuous if Google was going to continue treating pages in the Supplemental Results Index differently from the way it treats pages in the Main Web Index.

Since last fall three facts have surfaced about Supplemental Results pages that were all confirmed by Googlers:

1. Google has been using Internal PageRank to decide which pages are to be Supplemental or not
2. Google has not been indexing every word on Supplemental Results pages
3. Links from Supplemental Results Pages have not been passing anchor text

Google introduced the Supplemental Results Index a few years ago and apparently used it as a catchall for Web pages it wasn’t algorithmically sure what to do with. A lot of spam pages and many duplicate results pages ended up in the Supplemental Results Index. Over time, it became apparent to many people that Supplemental Results pages could still rank well in search results, although many business site operators were anxious about seeing the label next to their pages.

When Google launched the Bigdaddy Update in December 2005, Matt Cutts confirmed on his blog that Google was using two separate robots to crawl for the Main Web Index and the Supplemental Results Index. Googlebot-Supplemental didn’t get around to everyone’s pages as often as Googlebot tended to, but Google was being very picky about which pages Googlebot would not visit.

Google increased the Supplemental Results Index crawl rate as many people complained about their listings not updating. More recently Matt said again that Googlebot-Supplemental’s refresh rate had been increased. It now appears that Supplemental Results Pages are reindexed at least once a month. I have seen some pages refreshed twice in one month.

All well and good, in my opinion, but I have complained frequently and in as many places as possible about the fact that Google doesn’t parse (and index) every word on a Supplemental Results page. Matt even confirmed this at SMX Advanced 2007 (a fact that none of the SEO bloggers paid much if any attention to when reporting my question in their conference wrapups). Google was only indexing “important” words, Matt told us.

But what is an “important” word? How do you know which words you can use on a Supplemental Results Page and hope for a chance of being found in natural search results? Google’s segregation of the Web into content that is worthy of appearing in search results (Main Web Index) and content that is not worthy of appearing in search results (Supplemental Results Index) penalizes many innocent Web site operators who lack sufficient PageRank.

Internal PageRank, of course, comes from links — links, apparently, that are found only in the Main Web Index. So it would seem to those of us on the outside. The solution, several Googlers have said repeatedly, to having pages in the Supplemental Results Index is to get more “quality links”. Instead of helping Webmasters improve their search visibility, however, the Googlers only increased anxieties and inspired a massive wave of spam link drops.

What is a “quality link”? I have repeatedly and openly told people on numerous blogs and forums to get links from Main Web Index pages to help improve their search visibility. My hope was that people would find innovative ways to get links without resorting to spam. Well, this is the SEO community we’re dealing with — the solution to any problem has been (for years) “get more links, any links, links at any cost”. So naturally people increased their misguided linking efforts.

At the same time Google launched an all-out war against paid links, paid link brokers (can you see any sites named “text link ads” in this SERP?), and sellers of links. Many a Web site has been punished or singled out for ridicule and condemnation for selling links — not only by Google but by misguided Webmasters who have been helping Google police its index. Google’s mantra of “editorial choice” has become a joke because of this campaign against paid links.

“Just stop allowing links to pass anchor text,” I keep saying. Google, of course, refuses to consider the possibility that its practice of allowing Web sites to misrepresent the purpose of other Web sites is at the heart of all this trouble.

Link anchor text is not only the holy grail of SEO, it’s the holy grail of search and may, in fact, be part of the reason for why Google has lost search market share two months in a row (something that, contrary to media reports, has not happened before).

This is not about PageRank, it’s about manipulating search results with link anchor text. Millions of people alter the relevance scores for at least as many Web sites every day by pointing links with anchor text that is passed to those Web sites. SEOs have long known that link anchor text in no way provides any relevant data about what a Web page is relevant to. All it does is allow people to bypass on-page repetition filters. In effect, it makes spamming Google easier because you can hide your spam across hundreds or thousands of Web pages.

In Google (and other search engines that allow link anchor text to pass from page to page) all links are relevant regardless of where they reside. That is the way it’s always been since Larry Page and Sergey Brin first put together their lame idea of what would be the best anti-spam measures in Web search (they claimed their system could not be gamed).

So links pass anchor text even today, except for links on Supplemental Results Pages. Or has that changed?

It’s hard for me to be sure because when I see a page in the Supplemental Results Index I start pointing links at it from Main Web Index pages. I don’t like Supplemental Results Pages because they don’t rank for anything. But over the past few weeks I have started to notice Supplemental Results pages once again appearing in random query results.

Does that mean Google is now indexing Supplemental Results Page content? Does it mean that link anchor text now passes from Supplemental Results Pages to other pages?

Normally I wouldn’t mention something like this because if it’s true then knowing that Google has allowed such pages to appear in search results and pass value again gives me an advantage over everyone else who has been reading my and other people’s dire analyses of the Supplemental Results Index issue.

In fact, if historical trends hold up, it’s almost guaranteed that you’ll see people repeating what I and others have been saying about the Supplemental Results Index over the past six months for about the next two years (give or take). It will depend on what Matt Cutts says on his blog and whether Danny Sullivan picks up on it and repeats Matt’s message on SearchEngineLand.

But now is the time for people to start looking at their Supplemental Results Pages in a new way. Do your Supplemental pages come up for unique expressions in search results? Do your Supplemental pages pass anchor text? Is Google now handling your Supplemental pages differently?

I have been openly critical of Matt’s proposal that the “Supplemental Results” label vanish because I feel strongly that if Google doesn’t allow Supplemental Results Index pages to work the same way as Main Web Index pages then it would be misleading people into thinking the problem had been resolved. One would think that Google’s Universal Search algorithm might make the Supplemental Results Index obsolete.

Maybe that is what Matt has been suggesting, preparing people in the SEO community for life without a Supplemental Results Index.

It doesn’t have to mean that Google won’t trust a lot of pages. Maybe Internal PageRank still won’t pass as easily as link anchor text — which would be a stupid approach, but Google has been arguing against paid links for the wrong reason anyway. Most people don’t buy links for PageRank, they buy links for anchor text. Anchor text boosts your relevance scores and helps your pages perform better in search results.

Of course, paid links on sites that appear in the Main Web Index would also help Supplemental Pages get out of the Supplemental Results Index. From Google’s perspective, all they really did was take one problem (people buying links for anchor text) and turn it into two problems (people buying links for anchor text and people buying links to get their pages out of the Supplemental Results Index).

If they do away with the Supplemental Results Effect and allow all pages to appear in search results and pass anchor text then they reset the stage and go back to one problem. If they stop allowing pages to pass anchor text with their links they solve most of their spam problems almost overnight (and their relevance should improve drastically as well for many queries).

Of course, many people have been spoiled by link anchor text and they have designed Web pages that don’t contain relevant content (with respect to their targeted keywords). The screaming in SEO blogs and forums would be horrendous if Google plucked up the courage to do the right thing and stop passing anchor text.

I don’t expect that to happen any time soon, but I will keep an eye on Supplemental Results. I will also look to see if other people in the SEO community report similar apparent changes in how Google handles Supplemental Results Pages.

Frankly, I don’t mind the label as long as the pages can be found for queries and pass anchor text (yes, I use links to pass anchor text too — I just don’t depend solely on links).

Stay tuned.

Oh, one more thing. I’ll be away for the weekend so I may not post again until Sunday evening.