IAB Ads: 'Advertising Is Creepy'

Faced with increasing pressure from Washington, the Interactive Advertising Bureau launched a public service campaign on Thursday aimed at educating consumers about behavioral targeting.

The online campaign, created pro bono by WPP’s Schematic, features rich media banner ads with copy like “Advertising is creepy” and “Hey, this banner can tell where you live. Mind if we come over and sell you stuff?”

More than one dozen publishers — including Microsoft, Google’s YouTube, and AOL — have committed to donate a combined 500 million impressions for the initiative.

The campaign comes as policymakers are questioning whether data collection by marketers violates consumers’ privacy. Rep. Rick Boucher (D-Va.) has said he plans to introduce a bill that could require Web companies to notify users about online ad targeting, and in some circumstances, obtain their explicit consent.

In addition, the Federal Trade Commission has criticized the industry for using dense privacy policies to inform people about behavioral targeting, or tracking people online and sending them ads based on sites visited.

In a meeting with reporters Thursday morning, IAB President and CEO Randall Rothenberg said one goal of the campaign is to address regulators’ concerns that consumers don’t understand behavioral advertising.

The ad units themselves offer information about online ad techniques. For instance, users who mouse over the “creepy” banner can pull down copy stating that companies don’t use “personally identifiable information” to determine which ads to serve.

Users who click through land on the IAB’s Privacy Matters page, which includes a description of various forms of online advertising, information about cookies (including Flash cookies) and links to opt-out pages.

The portion of the landing page devoted to cookies says they “contain data that allow a Web site to customize content and advertising to your interests but generally do not contain personally identifiable information.” A section with information about geotargeting states that an IP address “reveals nothing personal about you to marketers and websites.”

But privacy advocate Jeff Chester immediately raised questions about such statements. “They are ignoring the growing consensus that cookies and IP addresses are personally identifiable,” says Chester, executive director of the Center for Digital Democracy.

Cathy Dwyer, a privacy expert and professor of information systems at Pace University, also questions whether the banners’ headlines are too sophisticated to draw in users. “Even ‘creepy’ itself is a technical term,” she says, adding that it’s mainly industry insiders and observers who use that word in discussions about behavioral advertising.

The FTC said this year in its report about online behavioral targeting that non-personally identifiable information could be used to identify specific users.

In the past, industry groups and observers defined personally identifiable information as names, addresses, phone numbers or other information that could be used to contact an individual directly.

Critics recently moved away from that definition, in part because Web users have been identified based on supposedly anonymous data. The most famous example occurred in 2006, when AOL publicly released search logs showing users’ queries and “anonymized” IP addresses for more than 600,000 users, Within days, one “anonymized” user, Thelma Arnold, was profiled in The New York Times after reporters identified her based on her search queries.

Schematic CEO Trevor Kaufman told reporters Thursday that a test of the campaign in late October and early November yielded a click-through rate of 0.5%. The trial involved 7 million impressions, mainly served on Microsoft’s Hotmail.

Separately, the digital rights group Center for Democracy & Technology also launched a privacy campaign on Thursday — although with a different goal. The CDT is hoping to persuade users to lobby Congress for online privacy legislation. The Web site for the CDT’s “Take Back Your Privacy” campaign enables users to submit concerns directly to the FTC and to send emails to their lawmakers.

first seen on Media Post by Wendy Davis, Yesterday, 6:58 PM

What Does Apples to Apples mean?
Simply put “apples to apples” in this case means you are taking multiple creative that look exactly the same in every way, shape and form except one single variable and testing them against each other. This one single variable is the most important part of the testing methodology and could lead to 25-50% improvement in your marketing campaigns (see below).

Control Creative
So as you can see above all your creative remain the same except for that one variable. In this case the background square highlighting the silhouette of the girl changes and everything else stays the same. Although this seems like a simple example, the testing method is simple by nature. As you can see below one creative is the clear winner with an increase click through rate (CTR) of one hundred and thirty percent, not bad for your first test.

 
Statistically Relevant Sample Size
The sample size is the number of impressions you run against your creative. Different creative sizes have different statistical relevant thresholds. Here are the thresholds:

425×600 – 500,000 – 750,000 impressions
300×250 – 750,000 to 1 MM impressions
160×600 – 800,000 – 1 MM impressions
120×600 – 900,000 – 1.1 MM impressions
728×90 – 1 MM – 1.2 MM impressions
468×60 – 1 MM – 1.5 MM impressions

In this case you need to test all of your creative at the same volume of impressions. For example: If you have four different creatives measuring 300×250 in size you would run the control creative at 55% while running the other three at 15% till you reach the statistically relevant number above. The reason you keep the control running at a higher level of 55% is to maintain a successful campaign while you test. Just because your testing doesn’t mean you should sacrifice the already proven success of your control, you might end up finding out the control beats all of your tests in the end anyway (this happens quite often actually). If you can’t afford the amount of impressions it takes to run a 4 creative test then just reduce the amount of test creative you run at one time. This will allow you to garner results at a faster rate which will turn into dollars before you know it.

Call To Action
You need to test your call to action just as much as you test different layouts, background colors and button colors. Think about what you want your customer to do:
“Watch Now”
“Buy Now”
“Register Now”
You need to have your call to action very clear and concise so that the customer doesn’t have a chance to forget what you want them to do. Don’t forget when your testing your call to action, you can’t be testing anything else at the same time. You need to isolate one variable at a time or you will never know if it was that change that made the difference.

Landing Page
When you’re creative testing you should send all of your users to the same landing page which is consistent with your creative’s look and feel. Consistency in your messaging along with the same look and feel can increase your conversions up to 35%. Again don’t forget, you can only isolate one variable at a time. If you test landing pages at the same time you are testing creative you will never know if it was truly your creative that made people perform the action you are asking them to perform or was it the landing page? That’s why isolating this one variable at a time method works so well, you know what you changed and you know if that one thing made a difference.

I hope this simple “Apples to Apples” creative testing explanation helps you and your marketing efforts in the future. feel free to post any questions and happy testing.