Responding to a post stating that Google makes use of forbidden cloaking techniques itself, googleguy admits it and promises an "exemplar punishement": banning its own pages from its own search engine for cloaking.
Cloaking, also know as stealth, is a technique used by some Web sites to deliver one page to a search engine for indexing while serving an entirely different page to everyone else.
This is clearly forbidden by Google and many other search engines guidelines and it sounds really odd (but does not really surprise me) that Google itself went against its own policy.
The funny thing, in my opinion, is the action that they are going to take against themselves, in order to fix the matter.
Googleguy says:
To be consistent with our guidelines, we’re removing these pages from our index. I think the pages are already gone from most of our data centers–a search like [site:google.com/support] didn’t return any of these pages when I checked. Once the pages are fully changed, people will have to follow the same procedure that anyone else would (email webmaster at google.com with the subject “Reinclusion request” to explain the situation)
I say it is funny because in many cases like this one, Google would ban an entire website from its search engine, but banning Google from Google i guess it wouldn't work really nicely: Internet has already had is "big Bang" a few years ago ;-) and quite a bunch of supernovas...but the time for a blackhole still has to come...
GG replies to a member reporting a site for which Google shows backlinks that don't exist.
Link to quoteHuh. Send me a spam report with "oaktown" in the comments and I'll root around to see.
This one is interesting, but not surprising. GG describes describes Google's stemming as "some stemming" and then says that it is better to have the actual word on the page. Does that mean that webmasters still have to optimize for all word forms?
Link to quoteIndustry-specific terms, slang terms, precise terms, and sloppy terms. Check out your logs to see what users are typing. If users are typing
configure my widget
compile my widget
then you'll want to make sure you also mention widget configuration, reconfiguring your widget, widget compile errors, widget customization, etc. The spread of words that 100 people will use when they have the same information need is staggering. Be thinking about what queries your missing because your page has the word "configure" but not the word "configuration." Note that Google does some of this semantic analysis (e.g. some stemming), but it's better to find the words from your own pages instead of Google having to guess that your "compile" page also deals with "recompilation."
GG replies to a question about pages and pagerank.
Link to quoteYes, a one-page site can have high PageRank. Generally sites with content do better than sites with little content, and of course sites with lots of pages usually have more words that can match users' queries. Things like tutorials and 101-style intros and information are especially good. Oh, and glossaries. Glossaries are a great way to define words that users will type when they want the info on your site.
DiscussGG lists a number of competitor search engines, and one former competitor (Kaltix was acquired by Google). Personally, I hope Google does not take the Microsoft-approach to smaller, up-and-coming competitors.
Link to quoteIITian, I almost said something about having 20 major players would be better. I think there's a ton of room for different companies in different niches, and it's good to see different philosophies and the approaches that come out of that. Kartoo/Mooter/Grokker hit different UI aspects, Teoma/Vivisimo lean toward different ways of clustering, Yahoo espouses PFI as an insurance policy to guarantee crawling, Feedster/Moreover and a ton of others lean more toward blogs or news, Eurekster is trying search+social networking, Turbo10 has got customization of which databases to search, Nutch goes for the completely open-source approach, Dipsie is evidently trying to crawl into forms and the deep web, Kaltix was playing with personalization. It's a really interesting time, and I'm glad that lots of companies are tackling different parts of search.
DiscussGG replies in the thread about way everyone picks on Google.
Link to quotePersonally, this is one of my favorite threads in several weeks. Lots of food for thought for me at least. Scarecrow, I know that you and I have different takes on many things, but I appreciate your thoughts--especially when you put it with a little less acerbicism (?). I also felt better that you started up other *-watch.org sites; it makes things feel a little less personal if you take on everyone. :) Liane, I had no idea that that happened--I hope you're doing okay now? I'm cleaning my place, and I have to say that your two posts make me feel so much happier as I'm going around picking up today. I feel like the guy at the bottom of the pile who can see a little bit of daylight as the top person climbs off. :)
DiscussGG talks about Google Priorities.
Link to quoteHarryM, I really thought your list (and the ordering of that list was really interesting). It's true that a search engine has to make money or it won't be a viable service after long. I was thinking about this thread last night when I was going to sleep, and I think the thing that set Google apart was that we tried our utmost to align #1 and #2 from your list, so that improving things for users would also be good for Google as a business. Things like text ads instead of banners, marking the ads clearly, and trying to make the ads relevant by taking clickthrough into account are the sorts of things that make users happier and return to search again, so it helps set up a virtuous cycle. I think it's interesting that you listed "keeping webmasters content" near the bottom of the top 10. It's certainly true that one could make that principle the top one. The difficulty you'd have is that most webmasters are content when they're in the top spot, and not everybody can be there. That leads to systems like Overture, where you bid to see who is the most willing to be in the top slot.
Anyway, that's a really interesting way to express possible different priorities, and I think the ordering of priorities like that could generate a lot of philosophies of many different search engines. Do you mind if I play around with those a little bit? I might email a few other Googlers about this thread and talk a little bit about the orderings.
Sorry for the long post. Liane, thanks for your post. I remember talking to you a few months ago, so it was good to hear your thoughts about that period now.
GG describes the latest Google features as non-commercial.
Link to quote"the Pope installing condom machines in the Vatican"
It's strange, because I feel like about two weeks ago we launched numrange search, images in news results, web alerts, the UI redesign, and a demo of personalized search--all of which I think of as non-commercial changes and improvements.
Anyway, time to head home for the night. Catch you tomorrow..
GG advises on techniques to ensure indexing.
Link to quoteI recommend a tree-like structure where you can always get from the higher levels down to the more specific levels. Other structures work too, o' course, but when in doubt I'd go with a site map and a simple linking structure so that you can reach every page via the root page, or a page that the root page can reach.
GG reponds to the question 'why is everybody picking on Google?'.
Link to quoteThat's a good question, Brett. I know that the atmosphere inside Google hasn't changed in any substantial way, e.g. become money-grubbin' or caring less about our users or their privacy--we're still working hard to give high-quality search results back to users. And if you go into Google on a Saturday like today, you'll see quite a few engineers there.
I also think launches like Gmail do show that we want to organize many different kinds of information, not just the web, and I can see where some folks or companies would want to see Google only do search over the web. One good thing is that I think there's lots of good competition going on in search right now, and that makes everybody work harder to give better results.
GoogleGuy replies to a member who reported a problem get a news sites indexed by Google News after changing IPs. Later in the thread, the problem was resolved. GG responds.
Link to quoteI didn't do anything special, but glad that it started working for you. If you see problems repeat, let us know (hopefully with enough specifics to debug..)
best,
GG
GoogleGuy responds to a suggestion of a Local PageRank factor fo local searches. It makes good sense that geographically-local links should carry more weight for Local searches.
Link to quote"With the growing importance (in theory if not yet reality) on local searches, do you think it's possible for a local page rank in tandem with a world page rank?"
I dunno. Cool idea though, SEOMike.
GG replies to a member who pointed a key difference between GG's explanation of Google policy of manually editing search results and that of David Krane, the Google spokesperson who said that Google are "not able" to make manual changes to results. It may be dwelling on semantics, but "not able" was clearly a poor choice of words, because it simply is not true -- "not willing, for the following reasons..." would have been more truthful.
Link to quoteYup, and that's the answer I'm trying to give (except don't forget the third tiny category of security-related stuff such as credit card numbers or social security numbers posted up on the web). David Krane doesn't work in my building (I'm not in PR), but I saw him today at lunch and we bemoaned the misunderstanding together--he was talking about not editing the results for that specific search. So he said thanks for trying to clear it up a little more. Thanks for expressing it in nice bold though, kaled. :)
I'll try to make sure we express it more clearly in the future.
GG explains his personal take on the "hand-editing" of search results.
Link to quote"GG, it was nice to see your participation in addressing this issue. "
And bear in mind that this is just my personal take, based on my experience of working at Google. One of the reasons I wanted to join WebmasterWorld was to be able to explain Google's stances on questions in greater detail, dispel misconceptions, etc. So I wanted to explain the context of the original quote, and describe why Google tries not to take manual action other than in cases of spam or when we're legally required to do so.
I've said many times on this forum that Google prefers to write new algorithms to improve our quality instead of trying to fix problems by hand. I do believe most users would prefer a search engine that tries to avoid hand editing their search results. I remember once noticing that for a rather obscene query, another search engine had hard-wired the first result for the search to return Google's executive biographies page. I also remember when another well-known search engine hard-wired it so that a search for "Google" would return the text "Google: The Inferior Search Engine" before the regular results started. I'm sure the people at those search engines considered those to be funny pranks, but I find that sort of hand manipulation of results more disturbing, personally. If those results were selected by hand, what other searches might be chosen with bias like that? Both of those search engines are no longer around at this point.
By the way, danny makes an excellent point in message #51, because this actually happened to Google as well a few years ago. Do folks think that for the search "scientology," we should have hand edited the results to remove an anti-scientology site from the #1 result? I do think many users who step into our shoes and try to come up with a policy on requests for results to be hand-edited can at least see the potential difficulties. Google tries to be fair and consistent by limiting the situations in which we take manual action.
DiscussFor the second time in recent days, Google points to the original PageRank papers (usually he steers cleer of PR threads).
Link to quoteIf you go back to the original PageRank papers, these links were treated as uniform--on-site vs. off-site didn't matter.
Discuss