----------------------------------------------------------------- SearchReturn Discussion List "Understanding Internet Search Technology" ----------------------------------------------------------------- Moderator: Published by: Detlef Johnson SearchReturn http://www.searchreturn.com ----------------------------------------------------------------- February 21, 2006 SearchReturn Issue #040 ----------------------------------------------------------------- SEND POSTS: ----------------------------------------------------------------- Refer a friend: http://www.searchreturn.com/subscribe.shtml ----------------------------------------------------------------- .....IN THIS DIGEST..... // -- CONTINUING DISCUSSION -- // "WebSideStory Release" ~ Mike Banks Valentine "ACLU, Google & The DOJ" ~ SearchReturn // -- ESSENTIAL NEWS -- // "Krugle Indexes Code" "Local Search to Triple" ----------------------------------------------------------------- // -- CONTINUING DISCUSSION -- // ----------------------------------------------------------------- ==> TOPIC: WEBSIDESTORY RELEASE "Search engine referrals convert at a rate twice that of other sources." From: Mike Banks Valentine http://www.searchreturn.com/digest/refs040.shtml Here we go again with bogus numbers and so-called "conversion rates" studies without looking at ROI. In a laughable footnote, WebSideStory notes quietly: Study examines only the conversion rate, not the ROI generated by each major search engine So why is conversion percentage relevant when total numbers of search engine *referred* visitors is not included in the calculations of conversion percentages? The entire thing becomes absurd once total numbers of referrals *within specific time frames* are known. I'm horrible at math, but fortunately I have a calculator on my computer here. It tells me that if Google sends 100 visitors to those sites in this study, almost 4 of them buy something. Whereas AOL gets more than 6 people buying something for every 100 that come from that search engine. We'll forget for a moment that those searches are actually still done on Google servers. Now those 6 AOL visitors took about 10 times as long to show up on the retailers sites from AOL searches because Google sends easily ten times the referred traffic as AOL. My calculator is telling me that, in the same amount of elapsed time, Google sent 400 buyers and AOL sent 6 buyers. Hmmm, I think the conversion percentage is kinda irrelevant once the time factor is considered. It would take AOL referred visitors about 66 times as long to convert at that better rate - *once they show up*. They trickle in from AOL at a rate equivalent to a faucet drip, while Google sends a blast of visitors at a full flow rate. It makes me crazy each time these "studies" are done which measure things nearly superfluous to ROI and the bottom line. I believe WebSideStory is looking for a little favor from AOL - or perhaps a little cash? MSN is right behind AOL as usual, in this silly "study" which has no concept of time or volume, and purposely ignores ROI by their own admission. Mike Banks Valentine http://RealitySEO.com ----------------------------------------------------------------- ==> TOPIC: ACLU, GOOGLE & THE DOJ http://news.bbc.co.uk/2/hi/technology/4731640.stm Essentials: Google is fighting hard against the DOJ subpoena. The US government wants a week's worth of Google search records, and website addresses for a million random sites from of their index. This is going to be a hard one for Google to fight. The premise for their case is as weak as the subpoena is silly. There is little doubt the DOJ is being obtuse about this request. Their plan is to prove the ease of porn findability on the Internet. Obviously, this can be proved by any hack without the records Did they ever think to try a search or two in the court room? Regardless, it is going to be difficult for Google to show the DOJ is overreaching. Is user privacy at stake? Do the records reveal Google trade secrets? No and no. Why then would Google fight this? The explanation that makes sense is protecting porn ad revenue as we thought eariler. If the DOJ began forcing American porn to load an age proofing mechanism, no pornographer could possibly comply. Even if they could, it would be bad for business. Since we're not talking about global governance in this case, American pornographers would have to relocate off shore to avoid age proofing. The whole thing would be a costly fiasco for them. Picking their battles, Google would rather fight the DOJ here than if the DOJ were to ask about records to help them in a criminal case. Statements show they would turn over records (including personally identifying information) if it helped law enforcement or national security. That is understandable, but it doesn't get cheers from Googlers to stick it to the man like a good fight against the DOJ does. This case is not about privacy or trade secrets or anything noble like that. It is much more likely Google is protecting their porn advertising revenue. The ACLU states that in response, (in case Google loses), they would be forced to request records that prove the data is not indicative of real searching activity. They would subpoena Google to learn how Google works, and show the influence of automated queries undermining what the DOJ could do with the data. Good plan. Google has more incentive to fight the battle too, and can enlist the ACLU to help their cause. It's about the free speach and the first amendment. The thing that doesn't add up is the ACLU might as well subpoena the engines that already surrendered data. Instead, they say they have no need or desire to obtain the information. The thing is, automated queries are not a Google specific problem. They are saying this to the court that is narrowly looking at the Google matter. If they needed it, the ACLU would subpoena data. The threat of an ACLU subpoena is what they need to help Google with its case (they will probably lose anyway). If they win it would help both Google and the ACLU together. The ACLU will probably win its case against the DOJ and COPA in the end, regardless of what happens to Google. http://news.com.com/2100-1028_3-6041223.html ----------------------------------------------------------------- // -- ESSENTIAL NEWS -- // ----------------------------------------------------------------- ==> TOPIC: Krugle Indexes Code http://developers.slashdot.org/article.pl?sid=06/02/17/2027211 Essentials: The search volume for code is enormous. The results are often mixed, and contain webpage information and not always reliable code. It will be interesting to see how well Krugle really works, considering that code very often can take on a personal flair and style. Open source tries to be agnostic in this regard, so it may work well. The index will be made up of open source code from common repositories and collaborative projects, starting with a reported 3-5 terabytes of code. ----------------------------------------------------------------- ==> TOPIC: Local Search to Triple http://www.clickz.com/news/article.php/3585511 Essentials: The local search market is set to more than triple over the next four years. This is growing at a rate more than twice as fast as physical Yellow Pages over the same period. Assuming the mobile search market is growing, if there is a real breakthrough for mobile phone consumers, demand for local search will be enormous (possibly outstripping this estimate). ----------------------------------------------------------------- Stay Tuned. Got feedback?: http://www.searchreturn.com/feedback.shtml Archives: http://www.searchreturn.com/digest-archive.shtml Alternate formats: http://www.searchreturn.com/info-formats.shtml Manage Subscriptions: http://www.searchreturn.com/help/manage-subs.shtml Problems unsubscribing? Contact the postmaster: mailto:postmaster@searchreturn.com Information on how to sponsor this publication: http://www.searchreturn.com/help/advertise.shtml Published by SearchReturn http://www.searchreturn.com Website Membership: http://www.searchreturn.com/register.shtml The contents of the digest do not necessarily reflect the opinions of SearchReturn or Detlef Johnson. SearchReturn and Detlef Johnson make no warranties, either expressed or implied, about the truth or accuracy of the contents of the SearchReturn Digest. Copyright 2006 Detlef Johnson. All Rights Reserved. -----------------------------------------------------------------