THE SEARCH ENGINE REPORT
December 6, 1999 - Number 37
About The Report
The Search Engine Report is a monthly newsletter that covers developments with search engines and changes to the Search Engine Watch web site, http://searchenginewatch.com/.
The report has 104,000 subscribers. You may pass this newsletter on to others, as long either part is sent in its entirety.
If you enjoy this newsletter, consider showing your support by becoming a subscriber of the Search Engine Watch web site. It provides you with some extra benefits and access to some exclusive materials and articles. Details can be found at:
Please note that long URLs may break into two lines in some mail readers. Cut and paste, should this occur.
In This Issue
+ Search Engine Strategies Coverage
+ Snap Unveils LiveDirectory
+ RealNames Slipping Generics Into Specifics
+ Lycos Changes Listings Format
+ Alexa Releases Amazon Shopping Tool
+ Search Engine Lawsuits O'Plenty
+ Yahoo Changes Home Page
+ Google Loses Virginity But Gains A Virgin
+ Search Engine Articles
+ Subscribing/Unsubscribing Info
Lots of news this issue, so I'll just dive right in....after wishing you all Happy Holidays!
Search Engine News
Search Engine Strategies Coverage
I'd hoped Search Engine Strategies '99 would go well, especially since I had organized the speakers and the day's content. I was pretty sure it would be well received. But it was absolutely stunning to see so many people clearly enjoying themselves, getting questions answered and learning more about the often confusing subject of search engines. One attendee left me his business card with a message on the back that was all I could hope for: "I came in a Search Engine Boy, and I leave a Search Engine Man!"
Moreover, several of the search engine representatives themselves commented to me on how useful meeting up with webmasters and marketers was for them. After the sessions, I was pleased to see them all uniformly staying behind to answer questions.
My thanks to those search engines that participated, to our other excellent speakers, our sponsors, and to all the attendees that came. I was especially happy to meet so many subscribers in person.
Yes, to answer the mail I've been getting, we are planning a repeat -- this time for the East Coast, and probably toward late March. As soon as details are firmed up, I'll let everyone know.
Yes, if you were a conference attendee, the slides are being made available. Email notification will go out to all of you this week. If you are afraid we don't have your correct address, check the conference home page below to know when the email has gone out and how to tell us if you didn't receive it. Sorry, slides are only being made available to attendees.
For those who weren't there, About.com's Chris Sherman has done an outstanding job summarizing key sessions in a five-part series. It's even great for those who did attend, to help you remember and act on some of the information presented.
You'll also find discussions of the conference in the I-Search newsletter beginning just after Nov. 18, and the special Conference Edition contains the results of a meta tag poll that was presented. See the archives to find the past issues.
Search Engine Strategies '99 Home Page
Search Engine Strategies 99: Special Report
About.com Web Search Guide, Nov. 29, 1999
(Note: The server seemed to be down when I checked today, but if you don't get through, try later -- I'm sure it will come back up.)
Snap Unveils LiveDirectory
In the wake of moves by the Open Directory and Go to allow more user participation in constructing their directories, Snap has unveiled a new "LiveDirectory" system that allows sites to be added within minutes to its guide.
It's a significant move, because crawler-based search engines such as AltaVista and Go have moved away from the "instant add" features they once offered, due to abuse by spammers. But Snap's implementation, within a directory structure and with some submission tracking features, might prove more resistant to abuse and perhaps ultimately beneficial to webmasters and users.
"This is aimed at getting the newest sites in fast," said Paul Wood, Snap's Senior Product Manager for Search & Directory.
In the past, submissions were placed into a queue, for review by Snap editors. If approved, they were added to the Snap directory. Under the new system, submissions are made available nearly immediately in the "LiveDirectory" area of Snap. Over time, sites listed in this area that come to attention of editors are reviewed and, if deserving, promoted into Snap's "main" directory.
For instance, if you look for "coffee research," you'll see how the new LiveDirectory data fits into Snap's other search resources. First, you are shown any matching hits from the editor-reviewed main directory, under the heading "Top Web Sites." Next, any matching selections from the LiveDirectory appear in the "LiveDirectory Web Sites" area. After that come results from Inktomi's crawling of the web, in the "All Web Pages" section.
In the search above, examples from all three data sources appear on the same results page. In contrast, most popular searches will continue to be dominated by results from the main directory. For instance, a search for "coffee" brings up 11 pages of results just from the main directory alone. For this reason, spamming of the LiveDirectory system is likely to be more limited, as it offers far less chance of ranking well for popular terms.
On the other hand, for very specific queries, the system is now more likely to fall through to LiveDirectory information. As this has an element of human involvement, I suspect it will result in better relevancy than the former dependence on pure crawler-based information for backup. It's an interesting compromise between making the most use of your editors, allowing user participation and still providing comprehensive crawler-based coverage when needed.
As a searcher, you needn't do anything special to access the LiveDirectory information. Just search as normal, and if the main directory has no matches, relevant hits from the LiveDirectory will appear, as explained above. Snap may also offer a link at the bottom of the results page to "Find the latest listings" for your topic within the LiveDirectory.
If desired, you can bypass any main directory results and go straight to LiveDirectory answers. Simply do a search, then choose the "LiveDirectory" link in the navigation bar that appears just below the search box, on the results page. FYI, you can also use the other links to access results from other data sources. "All Web Pages" takes you to Inktomi-powered crawler results. "Pictures," "Audio" and "Video" take you to multimedia search results powered by Snap's editors, image search engine Ditto.com and Inktomi.
Why would you go directly to LiveDirectory information? You might wish to see the very latest entries for a particular topic, which haven't yet been reviewed by an editor. Just be aware that because they haven't been reviewed, the quality may be less than main directory listings. Also, at the moment, there is no way to browse LiveDirectory listings in the way you can browse main directory listings. They can only be accessed via the search results page.
Be aware that LiveDirectory was just recently integrated into the main site, so formatting and other ways of displaying the data may alter from as described in this article, as Snap fine-tunes the system.
This is the LiveDirectory's home page, where you can search for any sites already listed and access the Update feature.
RealNames Slipping Generics Into Specifics
RealNames has had a long-standing and laudable policy of not selling a generic term to any one company. For instance, Amazon.com couldn't purchase the name "books," nor could Ford purchase the name "cars." But in the past two months, the company has cut two significant deals that go against this policy, enabling both its partners and RealNames itself great benefits. It also appears to be backing away from its policy on generics altogether.
The slippage began in October, when a deal with MP3.com was announced. Users could now enter "mp3" followed by any topic such as "jazz" or "Alanis Morrisette" into a RealNames-enabled search engine to receive a link taking them directly to information about that topic at MP3.com.
For instance, go to AltaVista, enter "mp3 jazz," and you'll see a link just above the first numbered listing that says "Click on this Internet Keyword to go directly to the mp3 jazz Web site." That's a RealNames link, and if you click on it, you wind up at page within MP3.com that lists music genres.
This deal also had the convenient side-effect of turning the generic term "MP3" into a RealNames link leading directly to MP3.com. Again, go to AltaVista and enter "mp3," then click on the RealNames link. Voila -- now you're at MP3.com. This is even more significant than you might realize, as I'll explain below. But first, let's talk about last month's deal with jobs.com.
With MP3.com, the RealNames press release played up how the deal benefited Internet users, giving them the ability to lookup information about MP3 files by artist or genre. In contrast, I felt the Nov. 16 press release for jobs.com gave at least as much if not more weight to the benefits that jobs.com was gaining from the deal, rather than Internet users.
"Online recruitment company obtains exclusive use of Internet keyword 'jobs'," the subhead says, followed by this key statement in the second paragraph: "With the use of Internet Keywords, jobs.com expands its Internet reach to millions of job seekers who use search engines and browsers such as Alta Vista, GO Network, MSN and Internet Explorer 5.0 all of which recognize Internet Keywords. Now anyone can find jobs.com by simply typing the Internet Keyword jobs in the search command located on any of these prescribed systems.
Indeed, in one fell-swoop, jobs.com has essentially catapulted itself to the top of these major search engines for the popular generic term of "jobs," via the backdoor that RealNames offers.
We've had a limited form of this backdoor at AltaVista since RealNames was introduced there in May 1998. Remember that I said you couldn't register generics? Search for "books" at AltaVista, click on the RealNames link that appears, and you'll be taken to a list of RealNames that contain the word "books" in them. A significant number of people click on the RealNames link for generic terms, then to some of the sites listed. It's one reason why registering RealNames is a smart move. As a webmaster, I'm an unabashed fan of this backdoor.
But what happens at AltaVista isn't the case at MSN Search or Go. Both feature RealNames links as prominently at AltaVista, if not more so, but they will only feature the links if RealNames has an exact match for the search term. For instance, a search for "amazon.com books" generates the RealNames link at the top of MSN Search (with the little RN next to it) because Amazon has registered that exact name. But search for just "books," and no RealNames link appears, since there is no such name registered. LookSmart has also recently changed from what I'll call a "loose" to a "exact" use of RealNames data.
In my opinion, the exact use of RealNames data by search engines is better for users. It makes the link more valuable to users, because they're growing to understand that when they click on it, they are taken to one official web site associated with that term. In short, Go, MSN Search and LookSmart are doing it right, despite the fact that if they made the change to loose usage, they'd benefit financially. That's because they share with RealNames the revenue generated from every click on those RealNames links. For some high traffic corporate clients, RealNames is apparently charging up to 40 cents per click on its links.
Now you can see why the MP3.com and jobs.com deals are so significant -- because "mp3" and "jobs" are actually registered, they will cause RealNames links to appear at the top of even search engines that have an exact usage of RealNames information (note: this isn't happening for jobs.com at MSN Search yet, but it's what should happen unless MSN has stepped in to specifically override this behavior or, more likely, is using its own, slightly older RealNames database).
The deals make you wonder if this means any company with a generic .com address and enough money will be able to cut similar deals with RealNames. If so, it even more so greatly enhances the value of generic .com addresses. It certainly goes against everything RealNames stated about generics and specifically domain names using generics when it launched on AltaVista:
"Organizations that own domain names may want to enter the main text of a domain name as the Real Name Address. For example Mecklermedia, as the owner of internet.com, may want to use the Real Name Address 'Internet'. Since 'Internet' is a generic term, this would not be appropriate. However, 'Internet.com' can be entered as a valid Real Name Address. Users of the Real Name System can then type the domain name as a Real Name Address and arrive on the Mecklermedia site.
I almost hate to use that quote, since Search Engine Watch is published by Internet.com (which was spun off from Mecklermedia). But it was the example RealNames gave in its FAQ back in May 1998, and it speaks directly against the type of deals it has cut with MP3.com and jobs.com. Moreover, RealNames has consistently stuck with this policy.
"The RealNames service is designed to take Internet users directly to a specific web page, which means that certain types of words and phrases are unsuitable as RealNames. Generic or common terms and personal names, for example cars, books and flowers or John Doe, are not unique, and do not help to deliver users to a specific web page. We want to find a RealName for you that identifies your site AND distinguishes it from all other sites. For this reason, generic or common terms and personal names are not permitted as RealNames," stated the RealNames "Namespace" policy in Dec. 1998.
And here's a quote from the August 99 RealNames FAQ: "Generic terms will not be assigned to one entity."
Exceptions to this policy have been allowed before now. For instance, RealNames has resolved "Amazon" to Amazon.com and "Apple" to Apple.com for some time, even though both are generic or could refer to multiple companies. The reason has been that the user expectation on the Internet is so strong between those words and those companies that it was justifiable to make such a connection.
But do most users equate "MP3" with MP3.com? I would doubt it. I'm even more doubtful that even a significant minority of users expect that "jobs" should equal jobs.com. It's not the market leader for online jobs (that's apparently Monster.com, which just became AOL's job search engine partner in a $100 million deal). It wasn't the top rated recruitment site in a recent poll by PC Computer (HotJobs.com took that honor). Yes, jobs.com is one of the web's leading recruitment sites, but it's a tough sell to say users expect jobs.com when they search for "jobs."
You'll notice I quoted a lot from past statements on the RealNames site. That's because if you go exploring today, you discover the company is sending mixed messages regarding its policy on generics.
For instance, the current FAQ page says terms are rejected initially if they are generic, and indeed, if you try to register a generic term, the system tells you to try again. The Easy Overview page flat out says, "Make sure your chosen Internet Keyword complies with our Namespace policy. It must be appropriate and unique, it must link to a live site, and it cannot be a generic term."
In contrast, the all-important Namespace Policy no longer singles out generic terms as being against the rules. And in another FAQ, it's clear that generics are now very much for sale:
"Q: How much does a generic Internet Keyword cost?
A: Each generic term is subject to a negotiated agreement. The cost in every case is uniquely established."
Ironically, this clearest statement that RealNames has changed the rules mid-stream comes within the area about the Policy Advisory Board that it formed in July of this year. That board is supposed "to ensure the company is held accountable for its decisions to the Internet community as a whole and to introduce an element of self-regulation into the RealNames approval process," the RealNames FAQ stated earlier this year, when the board was introduced. While it is made of what sounds like independently-minded members, they all nonetheless have a stake in RealNames.
Elsewhere in the Policy Advisory Board area, we find more specific information relating to its approval of generics, oops -- I mean what are now "so-called" generics terms:
"In the case of so-called generic terms, we will only approve such a term as an Internet Keyword if there is a sufficient user expectation associated with it. However, we cannot assign generic Internet Keywords to just anybody. Apple is a generic term and it is Apple Computers Internet Keyword. MP3 is MP3.coms Internet Keyword. So long as a generic term is sufficiently branded by a subscriber, then we can assign the Internet Keyword to them. In general, generic terms will not be assigned where the user expectation is insufficient or where the subscriber will not be prepared to market the Internet Keyword to users widely enough to establish this required user-expectation.
So apparently, even if users right now don't generally equate jobs.com with the term jobs, the fact that jobs.com will spend oodles of money trying to create this perception (with no guarantee of success) is enough for RealNames to sell them or anyone else the generic term.
I wanted to talk with RealNames about this apparent change in direction and opening up of generics, so you could hear its point of view. However, the company is about to have its IPO shortly, and it's been advised not to speak to the press about anything, it appears. Companies about to go public are legally required to observe a "quiet period" before hitting the market. Oddly, the legal eagles advising RealNames seemed to have no problem the company issuing a very loud press release about the jobs.com deal, which sparked my initial questions about this whole affair.
"You're right. It is a bit odd to be able to issue a release but not conduct any interviews. However, those are the quiet period guidelines we've been working under, and my legal contact just reiterated today that we will not be able to respond to your inquiries related to jobs.com. Sorry I cannot be of more assistance. As soon as we're able we'd be happy to discuss these issues with you," said RealNames spokesperson Katie Greene.
Well, we can at least check RealNames own S-1 filing for that IPO to so if it has anything to say on the generic issue:
"In addition, it is important that our customers, users and distribution partners perceive us as a neutral third party in assigning Internet Keywords and in maintaining the integrity of our Internet Keyword data. This process will be particularly difficult for us if and when we assign so-called 'generic' Internet Keywords such as 'cars' to a particular party. Even more than common brand names, generic Internet Keywords could be appropriately assigned to multiple parties, leading to potential conflict with customers, partners and potential customers once assigned to only one party. If we are unable to establish ourselves as a neutral third party, it is possible that we would lose users, customers and distribution partners and lose the support of the Internet community more generally."
If RealNames is aiming for a "neutral third party" reputation, then it might consider rethinking more deals of the type described above and the new policy statements it has on its web site. They send out entirely the opposite impression.
RealNames Namespace Policy
You can also access the RealNames FAQ and Easy Overview from here.
Policy Advisory Board Area
Information about the board, and also where you can follow links to see the more specific comments about generics by following the "Policy FAQ" and "Approval Policy" links.
The Verdict: Online Recruiters
PC Computing, Dec. 1999
Rates online recruitment sites and shows HotJobs.com on top.
Major Consolidation Among Online Recruiters Predicted
Business Wire, Nov. 29, 1999
Names Monster.com as the online recruitment market leader.
RealNames S-1 Filing
Be forewarned -- it's almost 2 MBs in size.
`Quiet period' hotly discussed
San Jose Mercury News, Oct. 17, 1999
Discusses the difficult issues companies face when trying to honor the "Quiet Period" yet continue with business as usual.
Lycos Changes Listings Format
Lycos has altered the format of its search results page, in order to consistently present more information from different sources within the same space.
For some time, Lycos has used different search databases such as the Open Directory, its own crawler and news wires. However, in its old results format, it was possible for one of the data sources to crowd out the others. For instance, there could have been so many directory listings in response to a search that the news headlines area would never be displayed.
Now when you search, the results page will usually be divided into four sections: Popular, Web Sites, News Articles and Shopping, with listings in each area.
The "Popular" area shows either Lycos editor picks (such as in a search for "yosemite") or Direct Hit popularity results (such as a search for "yosemite camping"). Also, any link in this area that begins "START HERE" is a paid listing.
"Web Sites" shows information from the Lycos version of the Open Directory, plus finds from the Lycos web crawler. Generally, you can expect to see the usual directory categories first, followed by some individual web sites from the directory, then web page found by the crawler coming last. However, it is not always this way -- crawler-based results can and do sometimes come before directory listings.
For instance, try that "yosemite camping" search again. Results 1 through 8 come from the Lycos crawler, then result 9 is from the directory (you can tell because of the home category shown below the listing, rather than a URL for the crawler-based results).
"News Articles" pulls back relevant hits from news wires and a select group of news web sites that Lycos crawls. Finally, "Shopping" pulls back hits from the LYCOShop area, where Lycos offer online storefronts to merchants similar to Yahoo Store.
Be aware that because you get some results from each section, there's no longer a big "next" button at the bottom of the page to bring up more results. Instead, you need to make use of the "More" links that appear at the bottom of each section, to see more listings for each particular data set.
Alexa Releases Amazon Shopping Tool
When Amazon.com purchased Alexa earlier this year, many people wondered what use the online retailing giant would make of Alexa's skills and technology. The answer popped up last week, in Alexa's quiet rollout of zBubbles.
"Just as Alexa gives better information for site visitors, so does the shopping companion empower shoppers, which ties in nicely with Amazon.com mission to help people find and buy anything online," said Cynthia Lohr, Alexa's Director of Communications.
When installed, zBubbles appears as a tiny icon in the upper right-hand corner of your browser's page display area. It very much resembles a traffic signal tipped on its side, so that all three lights are horizontal -- hence the icon is known internally as the traffic light.
In the first light is a little Z, and it's the heart of zBubbles. It turns from the usual gray to red to tell you if it has shopping information about a page you are viewing. By clicking on the Z, you can see information about products listed on that particular page.
For instance, if I go to MicroWarehouse.com's home page, then click on the Z, it tells me information about the Palm V, one of the products listed on the page. I'm shown places to buy it and related products, as submitted by other zBubbles users. Conveniently for Amazon, I'm also shown its price for the Palm V (hey, I'd save $70! But would I get the free leather case that MicroWarehouse throws in?). Even better, especially for Amazon, is that I can easily add the item to an Amazon.com shopping cart right from within zBubbles.
Clearly, this is a powerful tool for Amazon to extend its shopping presence anywhere its affiliate links don't already reach (yes, there are such places -- think schools, universities or importantly, Amazon competitors). But less cynically, it is also designed to encourage Internet user participation.
For instance, in the MicroWarehouse example, no users have yet entered other places to purchase the Palm V or related products. Now I think that anyone with both a Palm and a Furby must have the program that lets you use the Palm to control your Furby (including the ability to turn the darn thing off). So via zBubbles, using the second traffic light that has a pen image in it, I quickly add my recommendation into a short form.
Now when I click on the Z, I see "Furby Remote Control" as one of the recommended products. Slick -- perhaps other users will benefit from my work! But this change is only related to that particular page -- if I go over to Palm.com, zBubbles displays only information that visitors to that page have left behind.
zBubbles also gives users another means of accessing its product information -- by placing little Zs in orange circles right on the page you are viewing, next to any words it knows about. For instance, if I go to the Yahoo home page, on top of the "n" in the word "Pokemon" is a Z. Clicking on it brings up zBubbles information.
These Zs look as if they are actually part of the page, as if the page author placed them there. In fact, just for fun, I went into Yahoo's Running Shoes category, then made Adidas related to the word "running shoes." Now Adidas, which wasn't listed in the category, has sort of gotten placement without any approval by Yahoo's editors.
OK, so it's not exactly the same as being listed in Yahoo -- and certainly only those with zBubbles would even see the Z. But as a webmaster, it concerns me that people can be marking up my pages like this, in a way that looks like it's something I've done or approved of. And if I'm concerned, I can imagine some highly competitive retail sites that will freak out when they discover this. Can you imagine Datek littering e*Trade's home page with Zs that tie the two sites together? Guerilla marketers, take note!
For this reason, I suspect that having Zs on the pages themselves, useful though they are, will have to go away. Or alternatively, Alexa may have to offer some type of opt-out for webmasters that don't want their pages littered this way. The alternative will probably be a large number of complaints and possibly even lawsuits. Now there would be mess to sort out in court.
Remember that third light? Clicking on it merely brings up the normal Alexa service -- zBubbles is currently bundled with it, though Lohr said the two applets could be separated, in the future.
Certainly expect zBubbles to begin changing almost immediately -- the program literally came out in beta just last week. At the top of my wish list is the ability to toggle zBubbles off and have it stay off even when I open new windows. Perhaps my Furby remote control might help...
Just follow the links for the Alexa/zBubbles program for IE5. It's not available for IE4 or Netscape users, but the regular Alexa program is.
Alexa Upgrades For Internet Explorer
Never heard of Alexa? It's a great tool -- this review explains more.
Furby Remote Control
Maybe the best thing I put on my Palm, after Qmate for Quicken.
Search Engine Lawsuits O'Plenty
There's been a lot of legal action revolving around search engines recently, and especially last week. Below is a summary of the various cases. As a disclaimer, I was the search engine expert for Terri Welles, in the first case described, and I'm working as a consultant for Carla Virga's legal team.
You may recall the case involving Playboy and Terri Welles, the 1981 Playmate of the Year. In it, Playboy sued Welles for among other things, trademark infringement. Her use of the terms "playboy" and "playmate" in her meta tags was one of the central points in Playboy's argument of infringement. In April 1998, the US District Court in San Diego refused to grant Playboy a preliminary injunction against Welles, finding she had used the terms appropriately and fairly to describe herself.
The case didn't stop there. Playboy filed amended complaints against Welles, claiming again trademark infringement, among other grievances. Welles filed a countersuit against Playboy, claiming that Playboy had harmed her reputation, hindered her business and was liable to her for damages. Last week, both cases were dismissed.
The case against Welles wasn't just about meta tag usage, but I will focus on that particular aspect, since it will be of interest to most readers. The key points in Welles' favor regarding the meta tag issue, from my reading of Judge Judith Keep's opinion, are as follows:
1) There was no likelihood of confusion between the Welles' site and the Playboy web sites.
2) Welles' site did not prevent significant numbers of people from reaching the Playboy web sites.
3) There was no evidence Welles intended to divert customers from the Playboy web sites to her site.
4) Overall, Welles' usage of the words in her meta tags was fair use and thus allowable.
"Finding that Ms. Welles' use of [Playboy's” trademarked terms in the metatags of her website is a fair use comports with the fact web users must utilize identifying words to find their intended site. Not all web searches utilizing the words 'Playboy,' 'Playmate,' and 'Playboy Playmate of the Year 1981' are intended to find 'Playboy' goods of the official 'Playboy' site. Plaintiff has not addressed the fact that Ms. Welles' fame and recognition derive from her popularity as a Playboy model and Playmate of the Year. If a consumer cannot remember her name, the logical way to find her site on the web is by using key words that identify her source of recognition to the public," wrote Judge Keep.
I don't think this ruling bodes well for Terminix, which in a different case, has filed suit against a woman who put up a protest site complaining about the pest control company. One would assume this will be an open-and-shut case -- Carla Virga's site is non-commercial, making it more a free speech case than a trademark infringement issue. Once again, trademark usage in meta tags is being hyped as some type of ultimate proof of wrong-doing. I think those filing these cases against people with strong claims, such as with Welles and Virga, are going to discover that trotting out the specter of "hidden" meta code won't make up for a weak case.
Meanwhile, back to Playboy. In another case earlier this year, Playboy sued Excite and Netscape for selling banner ads linked to its trademarks. Its request for a preliminary injunction was dismissed in July. The company appealed, and that appeal was just dismissed on Nov. 15.
Another case, this one against image search engine Ditto.com (formerly ArribaVista) by photographer Les Kelly, has resulted in a preliminary finding in mid-November that the service's display of thumbnail images does not infringe the copyright of artists whose work it finds by crawling the web. A final ruling may come down at anytime.
And finally, back to images -- this time the dispute between GoTo.com and Go.com. GoTo won a preliminary injunction forcing Go to remove its logo for being to similar to GoTo's. Then, after Go replaced its logo, it won a 30-day reprieve. Now the old logo is back, at least temporarily.
Terri Welles Lawsuit Page
Has past filings about the lawsuits -- will likely be updated.
Terminix Lawsuit Aims To Mute a Web Critic
Wall St. Journal, Dec. 3, 1999
Nice summary of the Terminix case, for those with subscription access to the Journal.
Carla Virga's Legal Brief
Carla Virga's Protest Site
Thumbnail not even a tiny infringement
National Law Journal, Dec. 6, 1999
Details of the image search lawsuit.
Disney's Go.com gets green light for 30 days
Reuters/Variety, Nov. 18, 1999
Disney: Giving up logo could cost $40 million
Bloomberg, Nov. 15, 1999
Search Engine Lawsuits
Links and information about most of the suits above, and other legal issues involving search engines, can be found on this page.
More About Meta Tag Lawsuits
Practical advice about dealing with trademarked terms in meta tags.
Yahoo Changes Home Page
While it's difficult to keep up with changes at its competitors, Yahoo has remained a bedrock of stability since its launch. That's why the addition of a new Yahoo Shopping box to the home page a week or so ago was so dramatic. I think it's only the second major change Yahoo has ever made to its home page -- or certainly since it became a commercial site. The box has links to retail offerings and appears between the search box and the directory topics.
Google Loses Virginity But Gains A Virgin
And so finally it's come -- advertising on Google. They promised something other than banners when the site came out of beta, and Google delivers on that promise. Ads are simply a big text link, within a blue box at the top of the search results page. I find them oddly refreshing -- maybe I'll even click on one :) Try a search for "shopping," if you haven't seen one yet.
Meanwhile, Google has cut a new deal to power searches at Virgin Net, an entertainment and leisure portal for the UK. As part of its Virgin Net service, Google has added a special index of 5 million UK web pages to its regular listings.
Try Google in the UK by using the "Internet Search" option to the right of the home page.
Search Engine Articles
Internet Labels Lose Meaning in Rush for Popular Addresses
New York Times, Nov. 29, 1999
I can remember in early 1995 when I had to justify to Network Solutions why I needed two similar domain names ending in .com and still wasn't granted the second one. Today, they'd swipe my card without a second thought. In fact, I was very surprised to discover earlier this year that in addition to .com, .net and .org were being pushed as well -- even to non-network companies and for profit organizations. So I was pleased to see this article addressing how the domains are being devalued. And what if we do ever get new domains like .firm? Only a fool would believe that corporations won't grab every variation of their name for every domain, just to protect themselves. In short, those new names won't solve anything.
You've submitted to all the search engines. Now what?
Web Developer's Virtual Library, Nov. 29, 1999
A nice, short guide to other key things you can do to promote your site in addition to search engines.
Search Engines, Portals, Directories - Oh My
ClickZ, Nov. 23,1999
Quick guide to buying banners at search engines.
Yahoo opts for Net censorship
Sydney Morning Herald, Nov. 4, 1999
Yahoo removes messages that had no complaints and which weren't incorrect, in an effort to be extra careful.
Direct Hit's Subversive Game Plan
Industry Standard, Nov. 1999
Excellent article on Direct Hit's balancing act between competing with its partners and supporting them.
On Caching Search Engine Results
University of Crete/ICS-FORTH, Jan. 13, 1999
All the major search engines cache results. In the most simplified description, this means that if you search for "travel," the search engine will first check fast RAM memory to see if it has already served a set of results to that query. If so, then it doesn't do the more time-extensive effort of checking against data stored on multiple hard drives, across multiple machines, to retrieve a fresh results list. Sound scary, like you might miss data? Don't worry -- with the number of queries that take place, it isn't much of an issue. This paper, which I recently came across, does a good job of describing caching in more detail. It gets technical deeper into the paper, but the introduction will be accessible to many people. Interesting findings of an analysis of about 1 million Excite queries from 1997 found that most queries are popular. That is, they are frequently repeated. The most popular term in the set was requested 2,219 times, down to the 1,000th most popular term being requested 27 times.
How do I unsubscribe?
+ Use the form at http://searchenginewatch.com/sereport/unsubscribe.html or follow the instructions at the very end of this email.
How do I see past issues?
+ Follow the links at http://searchenginewatch.com/sereport/
Is there an HTML version?
+ Yes, but not via email. View it online at
I didn't get Part 1 or 2. Can you resend it?
+ No, but you can view the entire issue online, via the link above.
How do I change my address?
+ Unsubscribe your old one, then subscribe the new one, using the links above.
I need human help with a list issue!
+ Write to email@example.com. DO NOT send messages regarding list management issues to Danny Sullivan. He does not deal with these.
I have feedback about an article!
+ I'd love to hear it. Use the form at http://searchenginewatch.com/about/contact.html.
This newsletter is Copyright (c) internet.com Corp, 1999