I recently bought advertising on a tool which gave Google AdWords ad impression estimates via the Google AdWords API. My ad dollars, and the Google AdWords search volume tool itself, were both rendered useless when Google decided they wanted to provide more consistent and accurate keyword data.

All quotes below are from the above linked Google Groups thread.

As Ben Michelson put it:

I believe this may be just the first phase of a new "less is more" concept.

I expect subsequent versions will alternately snip out or merge previously inaccurate fields, until finally (AdWords 1.0) the TrafficEstimatorService will be void of inaccuracy by providing no information whatsoever.

Robert, another programmer, was also thrilled by the recent "upgrade"

Well done, Google. I just want to release my first Adwords program - partly based on the ctr value. I work about two month for it. Why we should develop programs for Google, if Google changes the API every two month (see also KeywordService)?

Patrick Chanezon, who was hired by Google as an AdWords API evangelist, stated:

The algorithms used in the TrafficEstimator may return some results that do not match your quality expectations, but they are not skewed in any way.

And here I thought making something inaccurate was skewing it...

Inasisi ran through some examples of the intentional data skewing and said:

If it is not on purpose, I don't understand why Google is not correcting the huge skewness in its estimates and further remove the only good statistics that we had to access to. If Google felt the need to be consistent to both the API users and the advertisers who use the UI, then they should have provided more information on the UI instead of having to strip them from the API.

For being so concerned with efficient market theory and collecting so much data Google sure is greedy with their data. They sure expect marketers to trust them with a bunch for not even trusting marketers with something as trivial as search counts.

Yahoo! and eBay allow access to their old marketplace data because it helps drive up costs, commerce, profits, and makes a more efficient market. Why can't Google get a clue on this?

Would the Sandbox concept be more accurately named the TrustBox?

NFFC, Lots0, and Massa highlight in far more detail what I was hinting at in my recent sandbox post and what I was trying to say on my WMW panel. Like it or not, SEO is largely becoming a game of public relations in many competitive industries.

Other semi-recent posts about the shifting of bulk spam to trust related techniques:

• risk vs reward in hiring a cheap link monkey

• link spam detection based on mass estimation
• TrustRank
• buying old sites

For those who recently complained that link trading doesn't work, it is largely because most link exchange offers and opportunities are garbage.

A long time ago I thought that search algorithms were going to advance to the point that it would be easier to influence (barter / manipulate / become friends with / etc.) people than search algorithms. With algorithms like TrustRank and the viral nature of blogging you really don't need to seek the approval of all that many people to do well. Put another way, if Danny Sullivan likes and frequently references your SEO website then odds are Google will too. In every industry there are going to be a limited number of people like Danny.

BTW, Danny also recently posted about the move toward trusted links.

If you watch Google close enough you get a lot of good free tips on public relations. They are ALWAYS in the news. And it is on all fronts.

• controvercy (Google Print)

• business (why split your stock when a lofty price makes front page news in the business / finance section)
• technology (too numerous to mention)
• etc etc etc

I have not been able to get a screenshot, but at WMW Vegas I noticed that when Baked Jaked was looking at Google search results for [Gwen Stefani Tickets] that their were Google AdWords ads at the top, right, and bottom.

On this pannel sites are reviewed for what they could do to improve their SEO.

ArtInternationalWholesale.com

• use specific page titles on deep pages

• Tim Mayer recommends optimizing for image search since it has lots of traffic and few people optimize for it. Use proper file names, alt tags, and link at the images.
• duplicate content issues (individual product pages are so similar)
• Matt says they need to look harder at link quality
• has the site duplicated on the .co.uk

OnlineHighway or InformationHighway...something like that...I so could not see the URL

• 50,000 to 5,000 visitors per day on update Allegra

• using popunders is just as evil as popups
• unsure purpose of site by looking at a page
• used to have multiple location based URLs...301ed to one central domain. Matt Cutts recommended that.
• Baked Jake said it can take 2 weeks to 6 months for 301s to take effect

TicketsToGo.com

• TicketsToGo seems penalized in Google since October 2004

• also created TicketsToGo.net because
• duplicate content issues
• Jake recommends starting from the bottom up. Building links into some of the subject specific pages and then working your way up.
• target Geo specific concerts
• Matt Cutts said "tell me about your backlinks" ... uber spammy reciprocal linking campaign. said good news is no manual spam penalty, but few of the low quality links this site has are doing it any good.

BargainTravel.com

• Tim Mayer asked what is actually unique about your domain?

• Yahoo! looks to ensure that with travel that the travel box is owned by the domain, not an affiliate form. Would not recommend submitting to Yahoo! paid inclusion
• Matt pointed out bad cross industry linking between his own site (like mortgage and credit sites), but said there were some good links
• Tim recommends making the site more unique from page to page and cleaning up the navigation links. He thinks the site navigation being at the footer and the page content existing primarily of wildcard replace duplicate content makes him think the local pages are for search bots instead of users
• not only link to related pages about immunization, etc., but also create tables of the locatin based related information, etc.

MicroMatic.com

• home page title nice

• site looks good
• individual product pages have good data
  Matt Cutts calls some of their paid backlinks "painfully obvious" to most any search engine. Matt said those links are not hurting them, but they are not helping in Google.

• could probably be rather easy for a site like that to get many links from beer hobbyist sites

LendingTree.com

• question about looking at their sitewide links to IACI partner network

• instead of looking to rank for mortgage Matt recommends looking for 20 year mortgage loan, etc.
• Jake recommends geo targeted pages
• Matt recommends maybe adding more text, but they are already looking at ROI testing and that is why there is limited text
• internal links can help reinforce topics
• Matt said their cross network linking seems pretty organic / not with intent to spam. Note that in Google's spam review guidelines that IACI's travel sites were ones that were whitelisted examples for remote quality search raters

• mortgage calculator link on LendingTree built for a manipulation test on Google...that was the reasoning the guy said and Matt Cutts made a funny face
• Matt Cutts said the partner links section on IACI properties as a technique do not work in Google.
• Matt said the goal of engines is to detect and count editorial quality votes.

Why chose Yahoo!?

• provide control to publishers - not a black box

• quality network
• competitive revenue opportunity (over 100,000 ad buyers)
• opportunity to integrate with Yahoo! content & Yahoo! users
  custmoer service & community

Size of Yahoo! Publishing beta?
approximately 2,000 publishers

they just launched ads in rss feeds

• open to all beat participants

• diversifies rev ops
• aligns w growing shift to rss
• supports movabletype and wordpress
• ads optimized to drive revenue

Yahoo! stated some think 5-6% of web users use rss but Yahoo! research showed it was closer to 30% of web users.

Jen asked if Yahoo! has anything similar to Google AdWords smart pricing?

• not needed for the following reasons

• allows advertisers to bid separately for the different content channels
• Yahoo! is more selective with partners

Jen asked when Yahoo! Publisher would be global
likely early 2006

plans for an affiliate program?
want to work to lower bar to make it easier for publishers to make money and work with Yahoo!...will allow affiliate program and will likely eventually support cpm pricing

wide range of topics on one site...how to be relevant?
can target ads at page level, directory level, or site level...can allow page or directory to override the site level targeting

going to change rev share percentage after beta?
absolutely not, but eventually may use traffic quality to adjust click price

Will Yahoo! offer behavioral targeting on contextual ads?
no nearterm plans, but may eventually

Rate of revshare / how compare to Google AdWords?
Yahoo! does not share the revshare %. more interested in being competitive in allow you to monetize.
revshare by publisher will vary over time

may eventually say you are in x range... to get in another range you may need to (get more traffic higher quality clicks etc)

Jen said targeting was no good at start...now better...is it where it needs to be?
still working to improve...pleased with speed at which it is being made better

Will Yahoo! offer a premium publisher program?
may give advertisers more control over who working with. but even small publisher may be premium if quality targeted traffic etc

How long will Yahoo! publisher in beta?
maybe toward end of q1

Jen asked plan on cpm ads?
may add cpm cpa. yahoo already does cpm on internal network

Will Overture drop the minimum bid?
unsure.