Welcome back to aimClear’s coverage of #SESCHI in the Windy City! Day 1 of Search Engine Strategies Chicago flew by with a flourish of excitement revolving around a host of untamed, wily marketers frolicking throughout the conference rooms. Day 2 is well underway, beginning with, among other events, Crossing the Digital Divide: The Leap from Search to Display. This lively session featured speakers Alan Chapell, President of Chapell & Associates, Dax Hamman, Chief Revenue Officer of Chango, and Marc Poirier, Co-Founder and CMO of Acquisio. Keep on reading to see what we talked about!

Moderated by Jon Myers of the SES Advisory Board and Yahoo! UK and Ireland, the panel jumped a discussion about the arrival of new display optimization technologies and next-generation ad exchanges and a super fresh perspective on an “outdated” media model. Demand-side platforms give advertisers more control over both pricing and targeting, and they aid in the creation of advertiser-specific algorithms through their raw power.

Taking the podium first was Alan, who is often referred to as “the online privacy guy.” He likened his explanation of online behavioral advertising to a quote from the movie J. Edgar:

“If you have the data, you have the power.”

Defining online behavioral advertising as “the process by which someone’s usage of interactive media is collected over time and across websites used to deliver a more tailored or relevant advertisement downstream,” Alan noted. The traditional advertising behavioral model was an ad network model. Formerly, a network had a relationship with advertisers on one side and publishers on the other. It was a fairly straightforward model, but the magical elixir that made this model a raging success was that the networks had all of the data. Surprise, surprise! When an entity has the data, they are in the position to deliver more relevant advertising to the user.

In the e-mail marketing world, the advertiser owns the data. Demand-side platforms (DSPs) are systems that give digital advertisers the power to manage several ad exchange and data exchange accounts through a single interface. In fact, Alan concludes that DSPs might be closer to the traditional email model in the sense that the advertiser is using the DSP and therefore he has the most data in this model.

Alan described how data management platforms have tools so the advertiser can utilize data and tailor content for advertising purposes. Some advertisers sell data in these open exchanges because they have so much of it. Historically speaking, behavioral advertising has been a creature of the direct marketing world. Behavioral advertising metrics have been utilized by direct marketers, but now a lot of brands are more interested in reaching a certain demographic at a certain date and time, especially with the online video model. So brands are using online data bits to increase their metrics around engagement. The idea behind engagement is to decipher how much an end user is fiddling around with the ads. Does the end user have a higher recollection of the brand than they did before being exposed to brand?

Food for thought indeed.

Next to take the stage was Dax Hamman who promised to answer the two most common questions people ask him:

  1. Where the heck do I begin?
  2. How do I measure the results?

DSPs originally started out as a technology to access your media. Now they are becoming agencies, buying and selling data, and there are a whole lot of other companies that fall into this category. If you’re a big agency or a brand with a big online budget for behavioral targeting, Dax recommended going in and working in one of these yourself so you can see what is going on in that channel firsthand. If you’re a smaller company, you don’t necessarily need to work with a DSP because there are other options out there.

Wait…you can do WHAT?
You can now target individuals instead of crowds. The more closely you target an individual, the more accurate you will be. What’s happening with an exchange is that you’re buying in an auction environment like you do in search and finding individual cookies.

This panel of speakers must be movie buffs, and I like it! Hamman described DSPs by comparing it to The Matrix -it’s an entirely new way to view the world if you think of the world as code, not as reality. DSPs make more sense when you think of the data points as 1’s and 0’s.

Code from The Matrix

Movie references aside, Dax dove into the details we really want to hear by fulfilling his promise to answer where a marketer should even start. Concise and to the point, Hamman recommends retargeting as a starting point. People think of site retargeting when they hear the word “retargeting,” but it’s actually more complex. There are 2 types of retargeting.

  1. Retargeting on-site events
  2. Retargeting off-site events

Site retargeting is like an onsite conversion or optimization technique. It’s important that you think more granularly, think about where you put those pixels on the product pages. You can put an exclusion pixel to exclude people you don’t want to target, for example on the Career Page. You can target from your emails in order to segment your audience.

An off-site event would be considered a case where someone does a search on Yahoo!, Google or Bing and then doesn’t visit your site, but there is still the intent they’re interested in – this is search retargeting. There is also social retargeting, in which they would take a sample of your audience from a cookie and find another group like them.

“It shows we are all really annoyingly predictable.” -Dax

When you distribute content, send it to the world with the retargeting pixels attached. Search and site retargeting are both talking to markers of intent. Site retargeting is not about prospecting.

How Do You Measure All Of This?
Of course, there are the usual impressions, clicks, CTRs, conversions, conversion rates, etc., but you need to have additional metrics. One option is to build an attribution model, but it’s a pain in the neck. More often than not, it’s too difficult to figure out. As a matter of fact, strike this idea out right now because it is so difficult and there is no right answer. Instead, try a PSA View-thru Reporter a Media Overlap Report:

  • Public Service Announcement Study
    What’s the value of the post impression? Take one half of your audience and expose them to one group of ads and expose the other half to different ads in order to keep them segmented. You’re measuring the uplift from these 2 groups. In regards to post impression revenue, we know some of them would have happened anyway because we traced the control group.
  • The Media Overlap Report
    “All you have to do” is have a unique order ID every time a conversion happens. Make sure it’s going into your Google Analytics so you can pull it out again – make sure it’s going into your ad server. Pull your report at the end of the month. What order IDs were for Search and what order IDs were for Display?

DSPs are the gateway to buying out of the 1’s and 0’s. Site retargeting is very easy to set up so visitors you bring to your site are better-treated and you can increase conversions while search retargeting brings in new people because of their intent.

Last but not least was Marc Poiroer who described his own experience with DSPs. He maintained that it is possible to have success sending people to your own pages. Acquisio had a mature PPC campaign running for years which:

  • Generated leads at $200 each
  • 20 leads per month
  • Reached maximum value
  • Used last click attribution model

Their campaign had reached a plateau. To test the integration with the DSP, they put aside $250,000 to increase the volume and try to keep it within reason. Starting in February, they didn’t pick up speed until May. The campaign was managed internally at first and is still going today. Here are the initial results:

  • Company generated loads of conversions, but CPA went through the roof many times because they had many running at same time and some were more high-performing.
  • CPA was almost at $600 which they can’t support, so they brought it back down by controlling frequency cap, websites they’re exposed on, removing stuff that was not working, etc.

Marc noted that site retargeting works quite well. By excluding the careers section and the management team page and by tagging more marketing pages, they had a site list using RTB to focus on sites that marketers go to.

Search retargeting is for people who searched and landed on your website, so that’s your search data. Brand awareness shot through the roof! Impressions without a click generated the most conversions as this is the single most frequent path people take. Search path was the second most popular and they more or less doubled their lead volume, maintained their CPA, increased brand awareness, and were able to scale volume massively. But along with increased volume comes dropping quality – the cost of scaling is high. Marc decided to ignore view throughs and used a last click model.

Building an attribution model proved to be a somewhat unconquerable demon for Marc because his team didn’t know what to do with it – there is no real clear answer.

Not quite like riding a bike, huh?
Marc contemplated some of the challenges with tracking these metrics. First, they didn’t know what to make of view through conversions, and there were difficulties among various tracking systems. The data is not the same across the board depending on the tracking tool you use, so whose data do you trust? It’s important to remove ads from sites you don’t want to be associated with. The end of the quarter always proves to be rather difficult because that’s the time when there is a mad rush to burn through all the budgets. There are so many competitors at this time so it’s impossible to compete at the end of the quarter.

As for attribution, how do you give credit? There is a massive impact on branding because all of a sudden, people are now searching for their brand a lot more. Despite the massive budget Marc had at end of quarter, he couldn’t burn through all of it. As a small B2B advertiser, he only targets 30-50k people. Ya win some, you lose some…at least they didn’t go OVER-budget! Last but not least, Marc explained the more flimsy you get with how you target, the worse your results… so… DON’T BE FLIMSY!

That’s all for now, folks. Thanks to the speakers for a great session, and such an in-depth look at a true darling of the marketing world, digital display! Be sure to continue following us aimClearians on Twitter at @merrymorud @ericasendros and @meganlichty.

photo credit: danzden
photo credit: cydonian.com