Forum Moderators: open
Today for example I was searching for "foo widget" It was great when you saw a site with the category next to it.. that way you could be fairly certain every site within that cat was going to be on a simlar theme.
What I now use is...
[ foo widget +"dmoz" ]. At leat that way it will display the ODP category within the results.
Anyone else still missing them.
Mack.
The line with the category path on it was of tremendous use, though: it showed that the site was included in the directory, showed what the topic of the site was, provided a link to the relevant part of the Google directory, and made it really convenient to search within the directory. The last two are especially sorely missed given that the directory tab has been removed from the Google pages (I wouldn't mind the extra click there one bit if I could still get to the Google directory categories on a search-by-search basis).
As mentioned earlier the similar pages option is a real NO NO and I never found a similar page by using it. Replacing that similar into directory will be a much better option.
However I think the size and datestamp are a good indicator for many users that are on dail-up. Even by using adsl I refuse to click links that indicate more than 100K.
Now we talk about virtual real estate. Has anyone tried to resize it's window from wide to narrow? It is worth a try. You will see that the real estate from the SERP are flexible where the real estate from the adwords are FIXED which really blows if you use the google deskbar and about 60 to 70 percent of the display is filled up with adwords :(
It's seldom that Google does something aggressively anti-useful. This is one time. LOL, it reminds me of some of the posts here on webmasterworld... all or nothingism. Just because two lines may be too much doesn't mean zero lines is the right "fix".
Still, I think this is all about the boondoggle known as personalized search. It currently isn't anything at all if you know about the Directory.
I used to click the links a lot (back in 2001 and 2002), but in 2003 the Google directory was so far behind what was published at dmoz.org (Google updated only in March and November), that I preferred to directly browse the categories at dmoz.org instead.
Maybe Google also had some issues with using the RDF data? For a while there were some encoding issues. The recent change to UTF-8 has brought a lot of benefits. There were initially tens of thousands of encoding errors uncovered, but hard work in the last few weeks has reduced the errors to almost nothing. Indeeed, the latest RDF has just 19 errors in it in over 1.8 GB (that's 1 800 MB) of site data, and the sites have already been corrected in the original data ready for the next dump.
GoogleGuy, get your techs to get a copy of the new RDF in a few weeks time, when hopefully it will be down to zero errors.
I must have used one of those links on about every other search.
If they really were removed based on click through data, then I'd say that the wrong conclusions were drawn from that statistic. Seeing a third-party categorisation with a significant number of listings is an extremely useful feature even for people who don't click on those links. They still get a much better idea what each site is really about.
I may have missed a few things, but this is the only example where I immediately noticed that Google had eliminated an obviously useful functionality from their site. I must say I really was surprised by that.
My guess is that Google is now marginalizing their ODP clone into oblivion, and will soon get rid of it altogether. It has to do with the culture of the people running Google. They are geeks that think "computers do it better", and a human edited directory having value just doesn't fit into their logic. Since "computers do it better", they just can't see how anything human edited can add "an obviously useful functionality" to their site. Sort of how Spock looked at Dr. McCoy.
Since the Directory links have been dropped, I find myself using the Toolbar directory icon more, but the serps are different... You are searching within the Directory, not within the Web. I like to be able to see the Directory category within regular serps.
Folks can still access the directory by clicking the "more >>" link on the front page though, so getting to the directory is only an extra click away.
"more" isn't very specific. Assuming you remember that the Directory exists, when you use this link, as Mozart said, you lose your search and you have to type it in all over.
If Google is using click-through stats to determine what features to keep, I'm wondering how "more" is going to survive. I'm frequently showing Web professionals the features behind the "more" link, and most of them have never clicked it.
I appreciate Google's traditionally lean look, but I've always felt that Google needs to work much harder to make some of its great features more accessible. This is demoting a big one.
How do you think they will monetize this?I dunno. I started thinking about the different models out there such as Y!, OPD, JoeAnt/Skaffe, etc, but I would expect something new.
Interesting thought. For this to work this possible new Google directory would need to have very high visibility from the front page. Otherwise why pay to be listed? And, what about non-commercial sites who wouldn't be able to pay?
Baloney. I was able to quickly disprove that with a few searches on non-competitive keywords that happen to be in the ODP category cat name, and that cat has a high Google PR. With the lack of much SEO for these keywords, if the above was true then I'd expect the SERPs to be totally dominated by the sites that are ODP listed. NOT the case.
Interior pages can get Directory listings, and they can rank for various terms. There isn't any different treatment.
Good point. Not only are there many cases where interior pages/directories also have an ODP listing, there are also cases where root isn't listed in the ODP at all, but a subpage/subsection is. This sort of thing can easily happen with amateur sites, where the webmaster isn't trying to have the whole site focused on one topic.
Interior pages can get Directory listings.
I don't think it's very common, though. (Not in my experience, anyway.)
It's highly unlikely that a site with, say, 100 or 500 subtopics would get 100 or 500 directory listings in DMOZ. Even if all those subtopics were listed, they wouldn't be listed by individual page--and pages are what Google indexes.
Google's fundamental design is the antithesis of dmoz. It appreciates that human input at the directory level is not efficient, since a robot can find and index far more web pages more quickly than the collective staff of dmoz. It makes more sense to invest labor in AI design and for the time being in removing pages designed to trick the automated system.
As Google Guy observed, references to dmoz in search results simply waste valuable real estate without meeting the user's expectations of finding information responsive to his search request. Most people never look at web directories and don't care whether a site is listed. They simply want an answer to a request for information.
As AI improves, reliance on dmoz and other directories will diminish, and the usefulness of Google search results can only improve. The fact that some person has listed a web page in a directory should have no relevance in determining which pages fulfill the user's request for information. Furthermore, almost any Google search now produces dozens of useless references to machine-generated, clone directories, some of which rank higher than legitimate websites. Hopefully the removal of references to dmoz listings is the first step toward removing the directory clutter from Google.
More backwards thinking. Google is Google in large part because they relied on the quality of the dmoz directory. The idea that AI is improving is just silly. Spam is improving. The web is exploding with low quality piffle. Trash upon trash upon trash, interlinking and mirroring and copying and duplicating. Human quality editing is becoming far more important with each passing day, and Google gets closer and closer to the necessity of hiring several dozens of humans to categorize and rate the quality resources on the Internet.
Perhaps in many years the AI will catch up, but bot technology is under seige from the piffle peddlers. The human element continues to be the single best signal of quality for the algorithm and it needs to be amped up even more... especially if persoanlized search ever is to exist.
since a robot can find and index far more web pages more quickly than the collective staff of dmoz
But this particular thread/argument isn't about speed and size of database. It's about implied, subjective quality. I may not have used the directory link often, but when I did, it was extremely useful. Chalk up my vote for bringing it back in some form.
Perhaps in many years the AI will catch up, but bot technology is under seige from the piffle peddlers. The human element continues to be the single best signal of quality for the algorithm and it needs to be amped up even more...
Google's algorithm already has a signal of quality based on "the human element": PageRank. It may not be perfect, but then, neither is a manually edited directory like Yahoo or DMOZ.
Because they know that they have fundamentally different design, purposes, and uses. And the carpenter who is very good with a saw ... still needs a drill sometimes.