@fspaggiari

In a few days, we will attend the Technology For Marketing & Advertising conference in London where we will launch a new feature that allows our users to directly try our contextual advertising web service semantic analysis.

We’ve been testing our new “copy, paste and test it” functionality (that’s how we call it internally), and we’ve invited users like you to take part in our tests (you may have already received our invitation or read this other post.)

To test the new functionality, visit www.admantx.com, and enter a URL in the blue search box:

Once you have clicked “Go” after copying and pasting your URL…

… ADmantX will display the results of its semantic page-level analysis and contextualization, showing categories, topics and emotions present in the page content.

What do you think? Contact us or request a demo now if you would like to discuss your experience with us or to understand more.

If you are planning to visit the TFM&A show, we hope you will join us on the exhibit floor in booth #213!

@fspaggiari

This year, we’ve thrown our hat into the ring for the “European Seal of e-Excellence 2012”, which rewards ICT and Digital Media companies with an excellent track record in innovation marketing.

Background on the ADmantX Semantic Advertising Display Service
The holy grail of matching ads to a page has been described as knowing “the right consumer, at the right time, at the right frame of mind”. Behavioral Targeting and Audience Buying can certainly take care of the right consumer part of the equation. IP sniffing and a few other tricks more or less take care of time and place. But if you are concerned about brand safety and improved response, you are missing a key metric: context.

When you hear “contextual” as a targeting approach remember this is really nothing more than a dressed up way of saying keywords. And keywords are notoriously bad at true contextual understanding.

The use of semantic technology to establish context is really about correcting and expanding on keyword understanding. Both correcting (disambiguating common, multi-definition words for their in-sentence meaning) and expanding (looking for the story-arc across all the content) is a classic segmentation scheme. The deeper the segments, the more potential matches, the better the accuracy of the ad match.

Take this accuracy-based approach and extend it now into the realm of emotions, buyer intentions, people, places, things, brands, standard categories and dynamic categories and you have a very rich data stream that is in fact the consumers’ frame-of-mind in the final moments before the ad match is made. Combining a semantic set of data with audience data and you will get the 40% improved engagement.

For additional information on the ADmantX Semantic Advertising Display Solutions, refer to our Website, in which we offer some additional insight into the problems that the ADmantX solves as well as practical uses for our service, and try the power of semantic targeting right now.

As you can imagine, it would be an honor to be selected from among a long list of entries for this, one of the most prestigious European and global in the digital space and online marketing (and of course should we win the award overall you will be the first to know.)

FINGERS CROSSED :-)

The other day we put some of our core semantic analysis capability on the front page of our website for all to see and enjoy.  This post provides some guidance in using this new feature. 

Enter www.admantx.com and you will see a box and a button provided.  Please only use Chrome or Firefox browsers.  Internet Explorer doesn’t yet work.  Type or paste a webpage address (URL) into the box provided and Press the Go button.  I used this URL to produce the analysis you see below.  http://msn.foxsports.com/tennis/story/Djokovic-in-line-to-make-history-at-French-Open-89464073

You will be presented with 5 channels of data about the URL you enter.  Entities or small self-contained pieces of information about people, places and things will be underlined in the recreated text of the story.  Entities can also be things like companies, brands, holidays, money, measurements, time, etc.; about 65 different entity types in all.  Graphically on the right you see standard categories; the extended IAB compliant way to describe the whole story.  You see the emotions readers are likely to have when they consume the story and thus establish their frame-of-mind just prior to matching and serving an ad.  You see custom matching algorithms which are really recombination’s of entities, categories and emotions into something wholly new and unique that attend to advertisers campaign objectives.  Lastly, you see related articles – the most closely aligned, similar articles that provide “stickiness” to a page. 

The small + sign in the upper right provides all the same data but in more detailed.  Click the plus sign to see it.  In a production environment the data is delivered in JSON format.  

Some of our direct competitors do this but not many.  I suspect for those who do not openly reveal how their technology works it is because either it is a blackbox, difficult, demands lots of customization and/or simply doesn’t work very well.  OpenAmplify and ContextWeb are two companies that do go public so kudos to them.  In this semantic game it takes guts to “go public”. 

OpenAmplify gives away an API here to process text or a URL .  I did this with many URLs and with the same URL about tennis as noted above.  The outcome was basically the same each time.  Lots and lots of XML markup but in the end not much more than a categorization of Sports and Tennis.  No identification of sentiment or emotion as we do, no identification of geography though that is clearly present and no identification of the many people present. OpenAmplify claims lots of interesting outcomes but I did not see them.  The full XML outcome is at the end of this post.

ContextWeb, like us, makes the process of testing easier by pasting a URL right on their website here.  Again I did this many times but also the Tennis URL from above.  Again the main categories are Sports and Tennis.  But now some more of the “entities” that trigger Sports and Tennis are shown including lots of the people and things that make up tennis.  So this is an improvement but then I see that somehow Football, Basketball, and College Sports are also identified.  Looking carefully at the webpage you will see sidebar stories along these lines.  So the ContextWeb is mixing up the main story with other stories.

A screen shot of the ContextWeb output is below.

Add as many URLs as you like to see how wide ranging and accurate ADmantX can be.  Have fun.  And let us know what you think by emailing me directly at baker@admantx.com  Thanks.

Open Amplify XML for URL =

http://msn.foxsports.com/tennis/story/Djokovic-in-line-to-make-history-at-French-Open-89464073

<?xml version=”1.0” encoding=”UTF-8”?>

-<ns1:AmplifyResponse xmlns:ns1=”http://amplify.hapax.com”>-<AmplifyReturn>-<Topics>-<Domains>-<DomainResult>-<Domain><Name>Sports</Name><Value>4.00</Value></Domain>-<Subdomains>-<DomainResult>-<Domain><Name>Tennis</Name><Value>4.00</Value></Domain>-<Subdomains>-<DomainResult>-<Domain><Name>tennis</Name><Value>2.00</Value></Domain><Subdomains xsi:nil=”1” xmlns:xsi=”http://www.w3.org/2001/XMLSchema-instance”/></DomainResult></Subdomains></DomainResult></Subdomains></DomainResult></Domains>-<TopTopics>-<TopicResult>-<Topic><Name>Djokovic-in-line-to-make-history-at-French-Open-89464073</Name><Value>4.00</Value></Topic>-<NamedEntityType>-<Result><Name>Type</Name><Value>NamedEntity</Value></Result>-<Result><Name>SubType</Name><Value>unknown</Value></Result></NamedEntityType>-<TopicInstance>-<Result><Name>Djokovic-in-line-to-make-history-at-French-Open-89464073</Name><Value>4.00</Value></Result></TopicInstance>-<Polarity>-<Min><Name>Neutral</Name><Value>0.00</Value></Min>-<Mean><Name>Neutral</Name><Value>0.00</Value></Mean>-<Max><Name>Neutral</Name><Value>0.00</Value></Max></Polarity>-<OfferingGuidance><Name>Not At All</Name><Value>1.00</Value></OfferingGuidance>-<RequestingGuidance><Name>Not At All</Name><Value>1.00</Value></RequestingGuidance></TopicResult>-<TopicResult>-<Topic><Name>msn.foxsports.com</Name><Value>2.00</Value></Topic><NamedEntityType xsi:nil=”1” xmlns:xsi=”http://www.w3.org/2001/XMLSchema-instance”/>-<TopicInstance>-<Result><Name>msn.foxsports.com</Name><Value>2.00</Value></Result></TopicInstance>-<Polarity>-<Min><Name>Neutral</Name><Value>0.00</Value></Min>-<Mean><Name>Neutral</Name><Value>0.00</Value></Mean>-<Max><Name>Neutral</Name><Value>0.00</Value></Max></Polarity>-<OfferingGuidance><Name>Not At All</Name><Value>1.00</Value></OfferingGuidance>-<RequestingGuidance><Name>Not At All</Name><Value>1.00</Value></RequestingGuidance></TopicResult>-<TopicResult>-<Topic><Name>story</Name><Value>2.00</Value></Topic><NamedEntityType xsi:nil=”1” xmlns:xsi=”http://www.w3.org/2001/XMLSchema-instance”/>-<TopicInstance>-<Result><Name>story</Name><Value>2.00</Value></Result></TopicInstance>-<Polarity>-<Min><Name>Neutral</Name><Value>0.00</Value></Min>-<Mean><Name>Neutral</Name><Value>0.00</Value></Mean>-<Max><Name>Neutral</Name><Value>0.00</Value></Max></Polarity>-<OfferingGuidance><Name>Not At All</Name><Value>1.00</Value></OfferingGuidance>-<RequestingGuidance><Name>Not At All</Name><Value>1.00</Value></RequestingGuidance></TopicResult>-<TopicResult>-<Topic><Name>tennis</Name><Value>2.00</Value></Topic><NamedEntityType xsi:nil=”1” xmlns:xsi=”http://www.w3.org/2001/XMLSchema-instance”/>-<TopicInstance>-<Result><Name>tennis</Name><Value>2.00</Value></Result></TopicInstance>-<Polarity>-<Min><Name>Neutral</Name><Value>0.00</Value></Min>-<Mean><Name>Neutral</Name><Value>0.00</Value></Mean>-<Max><Name>Neutral</Name><Value>0.00</Value></Max></Polarity>-<OfferingGuidance><Name>Not At All</Name><Value>1.00</Value></OfferingGuidance>-<RequestingGuidance><Name>Not At All</Name><Value>1.00</Value></RequestingGuidance></TopicResult></TopTopics>-<ProperNouns>-<TopicResult>-<Topic><Name>Djokovic-in-line-to-make-history-at-French-Open-89464073</Name><Value>4.00</Value></Topic>-<NamedEntityType>-<Result><Name>Type</Name><Value>NamedEntity</Value></Result>-<Result><Name>SubType</Name><Value>unknown</Value></Result></NamedEntityType>-<TopicInstance>-<Result><Name>Djokovic-in-line-to-make-history-at-French-Open-89464073</Name><Value>4.00</Value></Result></TopicInstance>-<Polarity>-<Min><Name>Neutral</Name><Value>0.00</Value></Min>-<Mean><Name>Neutral</Name><Value>0.00</Value></Mean>-<Max><Name>Neutral</Name><Value>0.00</Value></Max></Polarity>-<OfferingGuidance><Name>Not At All</Name><Value>1.00</Value></OfferingGuidance>-<RequestingGuidance><Name>Not At All</Name><Value>1.00</Value></RequestingGuidance></TopicResult></ProperNouns><Locations/></Topics>-<Actions><TopActions/></Actions>-<TopicIntentions>-<Domains>-<DomainResult>-<Domain><Name>Sports</Name><Value>4.00</Value></Domain>-<Subdomains>-<DomainResult>-<Domain><Name>Tennis</Name><Value>4.00</Value></Domain>-<Subdomains>-<DomainResult>-<Domain><Name>tennis</Name><Value>2.00</Value></Domain><Subdomains xsi:nil=”1” xmlns:xsi=”http://www.w3.org/2001/XMLSchema-instance”/></DomainResult></Subdomains></DomainResult></Subdomains></DomainResult></Domains>-<TopTopics>-<TopicIntentionResult>-<Topic><Name>Djokovic-in-line-to-make-history-at-French-Open-89464073</Name><Value>4.00</Value></Topic>-<NamedEntityType>-<Result><Name>Type</Name><Value>NamedEntity</Value></Result>-<Result><Name>SubType</Name><Value>unknown</Value></Result></NamedEntityType>-<TopicInstance>-<Result><Name>Djokovic-in-line-to-make-history-at-French-Open-89464073</Name><Value>4.00</Value></Result></TopicInstance>-<Polarity>-<Min><Name>Neutral</Name><Value>0.00</Value></Min>-<Mean><Name>Neutral</Name><Value>0.00</Value></Mean>-<Max><Name>Neutral</Name><Value>0.00</Value></Max></Polarity>-<OfferingGuidance><Name>Not At All</Name><Value>1.00</Value></OfferingGuidance>-<RequestingGuidance><Name>Not At All</Name><Value>1.00</Value></RequestingGuidance><TopicIntention/></TopicIntentionResult>-<TopicIntentionResult>-<Topic><Name>msn.foxsports.com</Name><Value>2.00</Value></Topic><NamedEntityType xsi:nil=”1” xmlns:xsi=”http://www.w3.org/2001/XMLSchema-instance”/>-<TopicInstance>-<Result><Name>msn.foxsports.com</Name><Value>2.00</Value></Result></TopicInstance>-<Polarity>-<Min><Name>Neutral</Name><Value>0.00</Value></Min>-<Mean><Name>Neutral</Name><Value>0.00</Value></Mean>-<Max><Name>Neutral</Name><Value>0.00</Value></Max></Polarity>-<OfferingGuidance><Name>Not At All</Name><Value>1.00</Value></OfferingGuidance>-<RequestingGuidance><Name>Not At All</Name><Value>1.00</Value></RequestingGuidance><TopicIntention/></TopicIntentionResult>-<TopicIntentionResult>-<Topic><Name>story</Name><Value>2.00</Value></Topic><NamedEntityType xsi:nil=”1” xmlns:xsi=”http://www.w3.org/2001/XMLSchema-instance”/>-<TopicInstance>-<Result><Name>story</Name><Value>2.00</Value></Result></TopicInstance>-<Polarity>-<Min><Name>Neutral</Name><Value>0.00</Value></Min>-<Mean><Name>Neutral</Name><Value>0.00</Value></Mean>-<Max><Name>Neutral</Name><Value>0.00</Value></Max></Polarity>-<OfferingGuidance><Name>Not At All</Name><Value>1.00</Value></OfferingGuidance>-<RequestingGuidance><Name>Not At All</Name><Value>1.00</Value></RequestingGuidance><TopicIntention/></TopicIntentionResult>-<TopicIntentionResult>-<Topic><Name>tennis</Name><Value>2.00</Value></Topic><NamedEntityType xsi:nil=”1” xmlns:xsi=”http://www.w3.org/2001/XMLSchema-instance”/>-<TopicInstance>-<Result><Name>tennis</Name><Value>2.00</Value></Result></TopicInstance>-<Polarity>-<Min><Name>Neutral</Name><Value>0.00</Value></Min>-<Mean><Name>Neutral</Name><Value>0.00</Value></Mean>-<Max><Name>Neutral</Name><Value>0.00</Value></Max></Polarity>-<OfferingGuidance><Name>Not At All</Name><Value>1.00</Value></OfferingGuidance>-<RequestingGuidance><Name>Not At All</Name><Value>1.00</Value></RequestingGuidance><TopicIntention/></TopicIntentionResult></TopTopics>-<ProperNouns>-<TopicIntentionResult>-<Topic><Name>Djokovic-in-line-to-make-history-at-French-Open-89464073</Name><Value>4.00</Value></Topic>-<NamedEntityType>-<Result><Name>Type</Name><Value>NamedEntity</Value></Result>-<Result><Name>SubType</Name><Value>unknown</Value></Result></NamedEntityType>-<TopicInstance>-<Result><Name>Djokovic-in-line-to-make-history-at-French-Open-89464073</Name><Value>4.00</Value></Result></TopicInstance>-<Polarity>-<Min><Name>Neutral</Name><Value>0.00</Value></Min>-<Mean><Name>Neutral</Name><Value>0.00</Value></Mean>-<Max><Name>Neutral</Name><Value>0.00</Value></Max></Polarity>-<OfferingGuidance><Name>Not At All</Name><Value>1.00</Value></OfferingGuidance>-<RequestingGuidance><Name>Not At All</Name><Value>1.00</Value></RequestingGuidance><TopicIntention/></TopicIntentionResult></ProperNouns><Locations/></TopicIntentions>-<Demographics>-<Age><Name>Senior</Name><Value>0.10</Value></Age>-<Gender><Name>Neutral</Name><Value>0.00</Value></Gender>-<Education><Name>Undecided</Name><Value>0.00</Value></Education>-<Language><Name>Undecided</Name><Value>0.00</Value></Language></Demographics>-<Styles>-<Polarity>-<Min><Name>Neutral</Name><Value>0.00</Value></Min>-<Mean><Name>Neutral</Name><Value>0.00</Value></Mean>-<Max><Name>Neutral</Name><Value>0.00</Value></Max></Polarity>-<OfferingGuidance><Name>Not At All</Name><Value>1.00</Value></OfferingGuidance>-<RequestingGuidance><Name>Not At All</Name><Value>1.00</Value></RequestingGuidance>-<Decisiveness><Name>Low</Name><Value>1.00</Value></Decisiveness>-<TemporalityResult>-<Temporality><Name>NA</Name><Value>0.00</Value></Temporality></TemporalityResult>-<Slang><Name>No Slang</Name><Value>0.00</Value></Slang>-<Flamboyance><Name>Not Flamboyant</Name><Value>1.00</Value></Flamboyance>-<Contrast><Name>Not At All</Name><Value>0.00</Value></Contrast></Styles><Search xsi:nil=”1” xmlns:xsi=”http://www.w3.org/2001/XMLSchema-instance”/>-<RequestInfo><RequestID>4c52ef59-4865-42eb-9651-408c1c7efb28</RequestID>-<paramList>-<Param><Name>ipaddress</Name><Value>10.210.110.156</Value></Param>-<Param><Name>analysis</Name><Value>Amplify</Value></Param>-<Param><Name>formatrequest</Name><Value>xml</Value></Param>-<Param><Name>cache</Name><Value>enable</Value></Param>-<Param><Name>polarityProcess</Name><Value>chunked</Value></Param>-<Param><Name>optimiseresptime</Name><Value>enable</Value></Param>-<Param><Name>maxTopicResults</Name><Value>0</Value></Param>-<Param><Name>maxActionResults</Name><Value>0</Value></Param>-<Param><Name>sortOrder</Name><Value xsi:nil=”1” xmlns:xsi=”http://www.w3.org/2001/XMLSchema-instance”/></Param>-<Param><Name>spiderrequest</Name><Value>N</Value></Param>-<Param><Name>ampliverseID</Name><Value xsi:nil=”1” xmlns:xsi=”http://www.w3.org/2001/XMLSchema-instance”/></Param>-<Param><Name>scoring</Name><Value xsi:nil=”1” xmlns:xsi=”http://www.w3.org/2001/XMLSchema-instance”/></Param>-<Param><Name>textsizestripped</Name><Value>95</Value></Param>-<Param><Name>textsizeunstripped</Name><Value>95</Value></Param></paramList>-<Version><VersionNumber>2.1</VersionNumber><BuildDate>Nov 29 2011</BuildDate></Version></RequestInfo></AmplifyReturn></ns1:AmplifyResponse>

@BrookeAker

When it comes to responding quickly to market shifts and user needs, ADmantX reacts quickly.

We announced this morningtwo new features that allow us to offer more effective contextual advertising: our IAB-compliant taxonomy enriched with new categories, including emotions, according to IAB standards; and  an enhanced user interface to plan, test and execute ads.

Our membership with IAB gives us an important vantage point from which to view the ever evolving requirements of the online/digital advertising community, as well as the opportunity to leverage their expertise to bring even greater value to our customers. In this era of contextual advertising, being able to automatically understand content and the emotions readers experience when consuming content is a real, measureable advantage. By making available a wider range or categories, the new ADmantX taxonomy includes over 900 categories, with 100 specifically for the emotions readers experience in consuming content.

Leveraging emotions, ADmantX automatically understands whether content is positive or negative, and the behaviors or motivations it elicits in readers—critical information for advertisers who want to encourage brand engagement, but also to ensure brand safety. Using semantic technology, ADmantx can also distinguish between polarity and emotional nuances to provide greater clarity into user intentions, preferences and reactions.

Our new interface also supports the need for greater precision using semantic analysis to process online content. In seconds, ADmantX converts the html page into plain text—including pictures and graphic information—and immediately extracts four main channels of targeting data from the content: Categories, Topics, Entities (people, places and things) and Emotions. The new interface allows ADmantX users to visualize the results of their query and easily understand what the URL page is talking about (topics) and how it is being discussed (emotions).

For example, a document can talk about “badminton” not only because the word is present in the text, but also because it is semantically linked to other words related to the sport. Simply put, a category such as “badminton” is its own semantic universe, and includes all the elements (words, verbs, expressions, etc.) that belong to the “badminton” universe.

With this new feature, related articles or advertisements previously added to an external ad server are also identified to make pages more sticky.  These new features help our customers reach their online advertising targets more effectively and allows us to assure ad buyers significant improvements in ad engagement through appropriate and targeted ad placement.

@BrookeAker

“It is one of the great marvels of life that, across languages, culture and history, it is possible, with sufficient knowledge, effort and insight, to truly understand the meanings of other peoples emotions and mental states”

- Richard A. Shweder

Yes it is a marvel.  Who of us hasn’t travel to another country, struggled with language but seemed able to communicate in any case using gestures, facial expressions and the like.  Emotions are universal.  Everyone expresses anger and joy the same way.

Most targeting techniques cannot do this but they should. 

There is a profound shortcoming of behavioral targeting.  Namely - that the data is old.  In all decisions information that is new is more valuable than information that is old.  Click histories, demographics, and purchase histories are stable and good but not a current understanding of human emotion at the time an ad is to be selected and placed next to content.  

But an excellent proxy is what consumers are reading. 

In fact for several years now there have been many efforts to “detect sentiment” by analyzing what is written. But sentiment pales by comparison to the richness and variety of human emotions. It is relatively easy to detect what is “positive” or “negative”.  But much harder to detect the difference between joy and success - universally positive emotions. 

This deep understanding of emotions is a hallmark of what we at ADmantX have created the last several years.  A rich portrait of human emotions during content consumption, and done fast enough to influence the selection and placement of online ads.  Just what they taught in ad school - “know the consumer, know them in the moment, so you can serve them well”.

Semantic processing of content generates the sufficient knowledge, effort and insight to supplement the hole in behavioral targeting.  Use it well and gain for your customers.  

  @BrookeAker

Economic theory holds that segmenting markets is always a good thing for firms to do and to use those segments for marketing, sales and advertising purposes.  I set out to find out if the theory holds in the digital ad world.  I wanted to know this since our technology – as lots of others in the online ad world – is predicated on greater segmentation in the match between ads, content and consumers.  I found some interesting results. 

But let’s establish some definitions first;

A market segment is a sub-set of a market made up of people or organizations with one or more characteristics that cause them to demand similar product and/or services based on qualities of those products such as price or function. A true market segment meets all of the following criteria: it is distinct from other segments (different segments have different needs), it is homogeneous within the segment (exhibits common needs); it responds similarly to a market stimulus, and it can be reached by a market intervention. The term is also used when consumers with identical product and/or service needs are divided up into groups so they can be charged different amounts for the services.

Groups arranged by affinity can be charged more than the group as a whole due to a variety of factors – income, lifestyle, exclusivity or product attributes tailored to the groups specifics needs. This seems intuitive and based in common experience.  Talbots clothing style, store placement and many other factors indicate it is an upscale women’s segment in comparison to Marshal’s as a moderate priced outlet segment of shoppers.

There is science behind this.  For the deeply academic this paper can be deciphered to see this.  For the rest of us here are the abstracted highlights;

1. The per-customer cost of advertising to different market segments is higher than advertising to a whole un-segmented group.  But the prices you can charge to the segments more than offset the added costs.
2. The volume of advertising has an added effect of making the advertising more effective and this an advantage over competitors.
3. But advertising communicates across segments and competitors where over time an equilibrium can develop where segmenting advantages disappear.  So increasing the frequency of advertising changes to segments can help maintain advantages.

So the digital ad world is like the traditional ad world in that segmentation allows for higher prices and profits.  But the digital ad world also has two advantages the traditional ad world does not.  First is the volume.  In digital the ad volume is nearly limitless and the marginal cost of adding ad volume approaches zero.  Second, the ease and the cost of changes in digital ads are commensurately low to nothing.

Getting better segments for your ad campaign pays off.   The more segments the better.  And the digital ad world now has the technology to provide this segmentation and it’s benefits at a far greater scale and at a lower cost than traditional advertising.  Use it to your brands benefit.

  @scagliarini

I have spent the last few days in London attending the Ad Trading Summit and ad:tech London. These experiences confirmed that the online advertising sector is going through a continuous transformation with no end in sight. Here are some common threads that emerged from these events:

1. Awareness about ad trading, and the related topic of targeting, is growing. The Summit had more people in the audience than expected, and at ad:tech, presentations about these topics (mine included) were crowded with attendees. We cannot yet say that we’re moving towards mass adoption of ad trading, as people aren’t sure if this is the future, or just a fad. This requires, for the foreseeable future, that providers offering these platforms and services must continue to focus on evangelizing the market. It’s therefore safe to say that in terms of significant revenue from mass adoption, we’re not quite there yet.
   
   
2. Targeting is very important for agencies and brands, but other players in the ecosystem (traditional ad networks and publishers, for example) do not feel the urgency to provide a full set of targeting data. While I understand the reasons for this resistance (“we already have so many things to do and are struggling to keep up with the competition”), I suppose that this means that data providers should still be very careful about over extending themselves or investing too much before the demand is there (there are still signs that the demand will become huge.)
   
   
3. Lots of confusion around DSPs and trading platforms in general. All of the buzz and the capital flowing into these companies seem to suggest that they will be the next big thing, but there is not yet proof that relevant publishers are joining the game. Publishers don’t want to repeat the mistake they made a few years ago providing free access to online content. They see ad trading platforms driving further commoditization, driving CPM to the floor and reducing revenue—and therefore are very careful.  There is some evidence that private exchanges / DSPs are an attempt to address this concern.
   
   
4. There is the perception among users (and brands, for that matter) that behavioral targeting is really impacting privacy. Independently, whether the EEC (and the US government) will pass strict legislation to limit these practices, there is a significant risk that these perceptions could cause a backlash against the companies that rely heavily on, or solely on these practices (especially for retargeting, which personally I feel is a really spooky/bad online experience).
   
   
5. Contextual advertising is gaining traction but if companies providing this service focus too much on brand safety, then we are just participating in a race to the bottom. What we should do is to present agencies and brands with a more broad set of features to help them innovate how they present their messages to consumers. Once agencies and brands fully grasp that today they are limited in what they can do, they will demand that ad networks and publishers provide new ways to deliver their messages. It will be more than brand safety, more than topic matching. It will be about matching the message to attributes that are by nature more important for brands: feelings, sentiment, intentions and behaviors. At least this is what we at ADmantX are betting on :-)

 @BrookeAker

On Monday April 11 in San Francisco I will participate in a panel discussion at the AdWeb 3.0 conference entitled MONETIZATION DISCUSSION: DISRUPTIVE/ENABLING TECHNOLOGIES, SOLUTIONS, BUSINESS MODELS AND ROI. Some heavy hitters will be with me on the panel including Tony Katsur, General Manager at MediaMath, Dan Beltramo - CEO and Co-Founder at Vizu and Matt Meyer, VP at the Rubicon Project.  As the agenda page states this panel is about;

… the ways and means that ad networks, data services, technologies and solutions provide buyers and sellers in the advertising marketplace with the tools needed to accurately target and deliver to customers campaigns in the most efficient, relevant manner possible. This includes data collection, analysis, digital buyer/seller ad networks and systems from the perspective of agencies, brand managers, publishers, technologists, researchers and solution providers.

So that is a mouthful.  But it really means that there is a lot going on day to day to improve the way we match ads to content.  The mantra from all panel members could be summed as “Speed it up, make it better and prove it works”. Of course along the way this all looks somewhat disruptive. Or as I have previously written there is some creative destruction afoot in this mantra. Again, to borrow from Joseph Schumpeter, the 1940’s Austrian-American Economist, who borrowed and retooled the idea of creative destruction from Karl Marx in which he states…

The opening up of new markets and the organizational development from the craft shop and factory to such concerns as US Steel illustrate the process of industrial mutation that incessantly revolutionizes the economic structure from within, incessantly destroying the old one, incessantly creating a new one … The process must be seen in its role in the perennial gale of creative destruction; it cannot be understood on the hypothesis that there is a perennial lull.

All the math and high speed exchange of ads, content and their match is good creative destruction. So long as creativity also goes back into the creative destruction too.

And so it is with advertising. There has been plenty of destruction but a great deal of creation as well.  Ad Networks, Ad Exchanges, Supply Side Platforms, Demand Side Platforms, Real-Time Bidding to name just a few. Efficiency, cost cutting, throughput, and massive impressions count as the benefits of this new digital ad world creation. All to the good.

One main point I will argue on the panel is that what has been creatively destroyed is creativity. By letting the last step in the digital ad process be the limiting and dumbing down of the words that describe the display we have robbed the creative spark out of the process. No more word-smithing. No more deep thinking about ways to describe a product or a feature that can capture the imagination of the consumer. Only bland keywords – the fewer in number the better; the wider the appeal the better.  

At ADmantX we have reinstated creativity to its rightful place while using technology. We believe semantic technology that aids the modern ad team can do this. So when the display ad is loaded onto the server and keywords are generated we get in the middle. ADmantX reinterprets, corrects, and expands the kinds of descriptors that ensures brand safety, accuracy, but most importantly creativity and an emotional connection with the consumer. Just like the magic you see in MadMen when Don Draper transforms the cold flat description of the slide projector into a carousel (here).  

 @BrookeAker

The snowball started  around the first of December last year.  Here we are in April 2011 and like the laws of snowballing compound interest the do-not-track efforts of the FTC, browser companies and major publishers has doubled and doubled again in short order. 

  

“The important thing is to find wet snow and a really long hill.” 

-Warren Buffett

 In the last week alone Google was sued by the FTC and ordered to “Opt-In” not just add the option to “Opt-out”, and the Associated Press - AP declared they would be the first to use Do-Not-Track on Large Scale.  Snowballs this big, rolling this fast do damage when they hit something.  They are about to hit the behavioral tracking data providers and the ad networks, exchanges and others they serve.  

The self-regulatory efforts seem too little too late by comparison.  When the head of the FTC states that “Privacy by Design” should rule he is saying he expects self-regulation to fail.  That is backed up when the also order the largest entry point on the web to provide as a default position opt-in.  The order has the potential to vacate in one swipe of the government arm efforts by Microsoft’s Do-Not-Track Option in IE and Firefox’s Do-Not-Track Option. 

Microsoft seems to be taking an aggressive lead, interestingly, over most other browser makers to be out in front on the issue.  They even have produced this tool to test your browser(s) for “Do-Not-Track”.  Looks like a smart competitive move to me especially when Safari and Chrome are late or no-shows for the party.

Even big companies want more privacy making the snowball unstoppable.  When the regulators are on one side proposing a flurry of new bills on privacy, the publishers and companies you serve are on the other side asking for better privacy and the technology (browser) companies you must go through reach customers are adding privacy capabilities you’re surrounded. 

Better it seems to go with the flow and start looking for alternative ways to maintain the good (the benefits of targeting) and dump the bad (invading consumers’ privacy in order to achieve targeting).  What is left is to target based on the content the consumer is reading.

When done at depth there is no privacy invasion, the targeting is precise, and brand protection is achieved.  At ADmantX we are already getting an increasing number of calls asking about the use of semantic targeting as a hedge against the above industry dynamics. It’s time to get out of the way of the rolling, growing snowball and onto the right side of history.

AP first to use Do-Not-Track on Large Scale

Big Companies Want More Privacy

FTC: “Privacy by Design” should rule

Flurry of New Bills on Privacy

Google ordered to “Opt-In”

Microsoft IE includes Do-Not-Track Option

Firefox includes Do-Not-Track Option

Test Your Browser(s) for “Do-Not-Track”

  @scagliarini

After running a successful beta test for the last few months, today we will launch our page-level semantic data targeting web service www.admantx.com .  I am not writing this post to explain what admantx is because you can just read our Public Release   I would like to focus more on the three most important things we have learned, among others, in the beta test.

a)    There is a significant misconception in the market that one targeting mechanism competes directly with the others. It is not rare to hear from prospects “I am using behavioral so I am not interested”. The reality is that targeting mechanism should be chosen based on the campaign you want to launch. We have developed a system that makes it easy to choose if and for which campaign the advertiser wants to chose semantic targeting. We think that this is helping customers because it gives them more choices at any time (at least until other targeting less privacy compliant tools are legal of course) J  It is also true that behavioral covers only half the targeting equation.  Using semantics means you are targeting the content as well.  So both play a role. 

b)      Our unique interface taught users to appreciate the business value of being able to create dynamic targets inside their inventory in real time. Creating an admant (e.g. a meta-category) really means identifying a content slice in high demand for a particular ad in a particular moment.  This maximizes the value of the inventory.  This is something new and definitely different from what people are used to in finding targets where the activity is delegated to remote back offices, code writing and mathematicians.  It took time for beta users to fully understand the value of being able to access an interface and use their expertise about what matters for a brand to do in a very simple way what you thought could be only be done by the geek crowd.  Not surprisingly, they fell in love with the chance to maintain creative control. 

c)       ADmantX offers a knowledge/expertise sharing mechanism. We chose to make this available not because “social is cool” but because we believe that brand managers should be able to influence not only what to publish but also “where to publish” a certain campaign, and we think that they should do this in real time.  Social media moves too fast to do anything else.  In addition, we think that this capability to design the perfect match between content and brand is scarce and we would not be surprised if this scarce resource will become highly valuable for brands. We know it will not happen right away but we thought it was worth to create the conditions for this to happen.

So this is what we learned during the beta test. I am sure we will learn more starting today… and I’am looking forward to it.