A study done by ContextWeb and Online Text Exchange found that ads placed contextually at the page level outperformed (in terms of branded recognition and ad recognition) both ads placed without regard to context, and ads placed contextually at the site level .

It shouldn’t come as a surprise that contextual ads (for instance, Dell or Intel ads embedded in an article on computers) outperform those embedded on a neutral page—reading an article on a specific topic demonstrates familiarity and affinity for the subject matter, so one would assume there is both greater recognition of the advertiser and intelligibility and appreciation for the proposition of the branded message among readers of specific topics.

What is surprising is that context-matching at the page level outperforms the same context-matching at the site level. Is it possible that on a niche site too many similar products drown each other out and dilute the strength of any specific brand’s message? On a site dedicated to computers, it might be easy to forget which manufacturer is offering a $500 rebate on laptops, because visitors were exposed to ads from Dell, Compaq, Gateway, Sony and Lenovo, for example.

In contrast, a general news site with only one tech feature might only feature one ad showcasing laptops, so its relative uniqueness among the pack might be retained better by the reader.

At any rate, the study seems to strengthen the argument that advertisers might be better served by spreading their spend more broadly, allowing contextual matching technology to serve their specific message/offer on pages that are topically similar, rather than concentrating on a narrower band of niche sites that more closely align with their product category.

This last week several publishers were reporting that their AdSense channels weren’t reporting accurate data.  Google was a little quiet about it, but today on the AdSense blog, they said that all is well again.  We’ve experimented with channels a lot, so we understand how important this data is to publishers.  Glad it’s fixed.

I want to thank everyone that has applied to the beta.  We hired a temporary designer to help with the comping process and today we are sending the first batch of people code to run the YieldBuild service.

Next week, we should be much faster at turning around the requests and we are working hard on a self-service tool.  Thanks again for applying to the beta and your patience as we improve our processes to serve you more quickly.

This has been a busy week for all of us working on YieldBuild. We’ve had over 600 publishers apply to the beta program and applications are still coming in more quickly than we can respond.

While we are hammering away on a self service solution, we are looking for some temporary design help to create mock-ups for all the new beta customers.

We need someone that will work from our office in San Francisco and who has their own computer and graphics software.  If you’re interested, please send an email to James@Hubpages.com.

The folks at Google understand the importance of formatting and placing ads.  Google is getting close to making a new tool for changing formatting options without having to go into pages and update the javascript.  This is a good idea and something we developed at YieldBuild.   We go much further, by just requiring a maximum height and width with the rest of the formatting optimization done from our custom algorithms. YieldBuild also supports placing many ad zones in the page so many more spots can be tested in combination to find the best performing layouts.

Sign up for the YieldBuild Beta.

Over the weekend Techcrunch wrote a piece on YieldBuild.  The response has been fantastic.  We are quickly working to get back to all the publishers that applied to our beta.

I had a little trouble responding to comments on the Techcrunch article because of Akismet flagging them as spam.  Mike dug a few of them out, but there was one more question I wanted to answer.

What stops the publisher for using YB for training and figuring out the optimal layout and quitting YB after that?

YieldBuild is a service that is constantly working to find better layouts.  After the initial training phase, YieldBuild will run several high yielding layouts that outperform the one best layout over time.  Also, different types of users respond to ads differently.  For example, someone that comes to a website in the morning from a search engine will have a different optimal layout than a person that directly navigates to a site in the afternoon.  YieldBuild automatically performs these types of optimizations.

Almost anyone with children will tell you that games, whether console, online or MMORPG, can capture the attention of kids and teenagers for hours every day, trumping television or Web surfing.

It should come as no surprise, then, that Google has built on its February purchase of Adscape and suggested, with recent interactive IP filings and an AdSense product rep speaking on monetization of gaming products next week, that it is going long in game advertising.

However, unlike Microsoft, which owns a popular console, and the Adscape deal, at $23 million was easily dwarfed by Microsoft’s purchase of Massive last year, estimated at $200 million, Google doesn’t have as strong an association with the industry as its chief rival. So, in keeping with Google’s tradition of capitalizing the long tail of publishers, Google will likely go after small, independent game developers that operate and sell beyond the scope of the big guys. And contextual ads might offer higher value than the banners currently typically served, targeted as they may be to the typical gamer demographic.

Game aggregators and distributors such as GameTap (Turner), Steam (Valve) and GameSpy (IGN/Newscorp) might also be attractive targets for big deals with Google.

PPC management firm SearchIgnite, in cooperation with investment bank RBC Capital markets, reports that Yahoo’s portion of text ad spend has increased by almost 2 basis points from July to September, while Google’s share remained flat and eCPM dropped.

Much of this is due to reinvigorated interest among marketers in allocating additional budget to Yahoo Publisher Network (YPN). Spend on Yahoo increased by 7.8% between the 2nd and 3rd quarters of this year, while it inched up only 0.8% on Google against an overall PPC market increase of 1.8%.

For a long time, AdSense has been the main player in town – its reach in search and across millions of websites is unchallenged, and because of fiercer competition among advertisers, it has generated the safest source of income for publishers looking to monetize across disparate topics.

However, as Google reaches full penetration across all advertising categories and its access to inventory becomes limited (despite recent forays into radio, print and mobile ads), prices climb enough to force advertisers to consider other avenues, like YPN. Yahoo has also done its part to help improve lead quality, leading to the sorts of results that build confidence among advertisers to increase spend.

Q4 is projected to see an even higher increase in PPC–who will get the lion’s share this time?

Venture Beat reported a new network for AdSense ads on Facebook today. They claim the ads can be purchased to only target Facebook. AdSense has had site targeted ads for some time, so we aren’t sure what’s different about these ads, since they appear contextually relevant to each application. Maybe Google is gearing up inventory for advertisers and will sell it soon.

But, we agree with Eric Schmidt’s comments about the number of pageviews generated by social networks. Therefore, it makes a ton of sense for AdSense to help publishers monetize their Facebook applications as well as other social networking sites.

The real problem facing social networking sites is monetization. Effective CPMs need to increase significantly. If AdSense can do this for Facebook apps, then it will be a big story.

Text ads, served by Microsoft were spotted on Digg yesterday.  JenSense did a nice job of getting under the covers and pulling out the code.  We also appreciate her comments about the formatting of the ads .

While we wouldn’t be too critical of the contextual matching, we do think that the formatting of the ads could certainly be improved.  After all, that’s our business.

If you look at the code a little more closely, there appears to be only one formatting setting at this point for the ad size.

<script type=”text/javascript”>

dapMgr.enableACB(“top_ad_msft”, false);

dapMgr.renderAd(“top_ad_msft”, “&PG=DIG728&AP=1390″,

728, 90);

</script >

</div><div class=”permalink-wrap”><div id=”contents”>

<div id=”wrapper”>

<div class=”sidebar-short”><div id=”item_ad_msft”

class=”item_ad_image”>

<script type=”text/javascript”>

dapMgr.enableACB(“item_ad_msft”, false);

dapMgr.renderAd(“item_ad_msft”, “&PG=DIG300&AP=1089″,

300, 250);

</script >

We tried to get the ads to come up on Digg, but didn’t get any.  It’s possible that they’re specifying the formatting server-side during provisioning which is then tied back to each ad via the location naming (“top_ad_msft”, “item_ad_msft”). This leaves the door open to more sophisticated formatting down the road. In some ways this may be a more future-proof approach.

So at this point we’d tentatively agree with the assessment that no formatting is being applied to these ads other than size.