Google’s Visualization API

Launched yesterday, the visualization API is closely tied to iGoogle and Google Docs.

A fun game to play whenever Google launches something: “How Will This Help To Make Advertising More Relevant?” For example, the Google Maps API provides geographic location. Ad Manager provides results of ads ran by other parties.

I doubt that the visualization API is there just to make Google Docs or iGoogle a bit better. So what does the Visualization API add to advertising effectiveness? Could the new structured data brought in from third parties improve ad relevance? The API will support custom data sources in the future, so perhaps this is a way to gain access to data that can be reused in ads themselves?

Examples of visualizations are in the gallery.

Starbucks’ New Plan

Brand autopsy covers the Starbucks’ new plan, as disclosed at their 2008 shareholder meeting. Apart from trying to compete with McDonald’s, the company launched a platform that solicits ideas from customers. Yes, even you can go to MyStarbucksIdeas.com and submit your great idea.

Here is the problem though: a premium brand that asks for ideas inevitably almost exclusively gets the ideas of giving out  stuff for free. Free wifi, a punch-out coffee card to get a free coffee every 10 purchases, free cup of coffee when you buy a coffee beans, free drink on your birthday, and so on.

Acting out on these ideas is simple brand suicide. Just focus on quality and on ways to showcase quality.

Usage of Context in Online Advertising

Last week’s Economist has an interesting article about call center software.

One trick, which can shave seconds off each call, is to take into account “context data”—such as the caller’s location, the local weather, the time of day and records of their recent transactions—when routing a call […] Context data may also include information gleaned from news reports. Storms, sporting events and transport strikes, for example, can affect callers’ moods and purchasing dispositions

I haven’t seen any good, successful examples of usage of “offline” semantic context data in online ads. Can the tone and manner of messaging be adjusted based on the weather? Should different sets of ads run depending on what’s on TV in that geographic location at that particular moment? I can imagine the combination of daypart targeting with deeper semantic data making a big difference in some situations.

Interesting to see the call center industry go through the same optimization hell as the online ad industry.

Google Launches Ad Manager, Completes SkyNet

Today the news of Google’s imminent release of their free ad serving platform, Ad Manager, hit the web. Imminent, because for years now the commoditization of ad delivery for publishers has been in the air. And yet surprisingly, at the time of this writing this isn’t even a topic on TechMeme. A short post on TechCrunch completely misses the point, claiming this move to be Google’s entry into the ad management game.

Saying that is like saying that the acquisition of Urchin was Google’s entry into web analytics. It is, of course, everything but that. The parallel is interesting to make. Both services require significant investment, and both will be offered completely free to the publishers. Most importantly, both serve the unique purpose of collecting data on traffic, ad performance, and audience in order to maximize Google’s ad revenue.

Although at first sight this seems like yet another announcement of a small app, this move is fundamentally important because it is clearly putting Google in a blatant conflict of interest. Ad Manager’s promise is essentially the ability to use it as you please, which may include ads from third-party networks and exchanges. However, witnessing the performance of all ad inventory on a site, including ads served by other sources, puts the small publisher’s revenue directly at the mercy of Google’s SkyNet. Since the publisher’s AdSense revenue share is completely arbitrary, the revenue given back may simply match the eCPM being attained with other networks.

In addition, running ads from other networks and exchanges may serve as an optimization test for Google’s own ad products. If before Google had third-party ad performance results only from Analytics, now they complete the picture by serving the inventory first-hand. This means that a site that runs, say, a credit card ad from ValueClick without much success, automatically discloses to Google this information, allowing it to make the decision of not serving credit card ads on that same site in the future. In other words, they will use others’ money to make their product more lucrative for themselves. And of course, since the acquisition of DoubleClick cleared, AdSense is just the beginning of it.

Interestingly enough, Google Analytics now shows me this gem of a message:

At least they have the decency to ask.

The exact same strategy is being pursued in traditional advertising. A few days ago, Google’s Tim Armstrong announced the upcoming “dashboard” allowing advertisers and agencies to track the performance of all their online and offline initiatives — including results from Google’s competitors (and I won’t even get into their partnership with Publicis).
For other companies working in the space of ad delivery and optimization, such as the recently funded OpenX and PubMatic, as well as my so far unannounced project, this shows that we are on the right path. It makes little sense for publishers to put wolves in charge of guarding sheep.

This area is ripe with opportunities, but as always, openness and disclosure are key.

Mad Men

The show is just so damn good. Please go torr rent it.

John Hagel: Shift Happens

The basic paradox of the Internet can be framed very simply: The very platform that makes advertising both more relevant and more measurable is the same platform that longer-term will challenge and ultimately undermine the basic role of advertising in communicating with customers.

Edge Perspectives with John Hagel: Shift Happens – The Future of Advertising