Rethinking the Application of Behavioral Targeting

Posted by Julia Casale-Amorim

With the intriguing concept and reported benefits of behavioral targeting (BT) so deeply entrenched in many marketers’ media plans, complacency over how this controversial form of targeting is actually conducted and the implications it can have from a branding perspective are all too often getting lost in the hype (and routine).

The basic principal of any online targeting method, including behavioral targeting or retargeting, is relatively simple. Sadly, it is commonplace for industry sellers of this technology to purposely “overcomplicate” these capabilities in an effort to “pull the wool over the eyes” of unsuspecting newcomers to the digital medium.

The major auto makers are infamous for their early adoption and devoted use of BT. Let’s take a moment to examine why?

Expanded Reach: BT provides an opportunity to reach “qualified” users on sites outside of the automotive segment where demand for inventory consistently surpasses supply, and where reach potential is fixed.

Reach to In-Market Buyers: BT provides an opportunity to reach auto intenders through both focused targeting (specific make/model/category) and general targeting (not model-specific) strategies.

Strong Metrics/Proof of Performance: It works and it’s backed by impressive purchase intent and brand favorability metrics (e.g. Dynamic Logic Studies).

BT Effectiveness – Some Thoughts and Questions to Ponder

How Fresh is the BT Data Being Used to Target Your Campaigns?
According to a 2009 study conducted by TNS Media, compression of the automotive research process is a major new industry development:

  • 65% of in-market Vehicle Buyers (VB) indicated that they researched one month or less prior to purchase.
  • 59% of New Vehicle Buyers (NVB) indicated they researched for one month or less prior to purchase.
  • The closer to the time of purchase, the more effective media efforts were. In the one week leading up to purchase, 42% of NVB were referred by search to OEM sites compared with only 15% in the 3 months prior to purchase.

Today’s methods of behavioral targeting rely completely on the use of historical user data, which may not accurately reflect a consumer’s buying stage at the time of ad delivery. In the wake of a shortening automotive consideration window, how quickly does a BT vendor need to gather, process and refresh its profiles so that that data used to target against is accurate at the time of campaign delivery? What mechanism is used to eliminate consumers who have already purchased from the buyer pool?

What About Inventory Quality? Doesn’t that Matter Anymore?
Here’s another way to look at BT. BT can be a means to market “poor” inventory as “desirable” inventory through the exploitation of data gathered through controversial profiling tactics. Marketers put so much focus on BT that the entire concept of quality and inventory source is being lost. Inventory is not so cut and dried as BT vendors make it out to be. There are countless factors that contribute to the success of any single ad placement, which are often overlooked during insular BT discussions:

  • Is the ad even visible on the page (i.e. above the fold?) How can an ad reach the intended audience if it is not seen?
  • When was the ad delivered? If the user has been stationed in front of a website for two hours and has been exposed to hundreds of different ads, at which point he/she is then served a behaviorally-targeted ad, will it be remembered? Will it even be noticed?
  • What is being served (content or otherwise) alongside the ads? Are the ads in good company? Is the advertiser jeopardizing the perceived value and status of its own brand by being delivered on a website or alongside other advertisements that are not of the same caliber? Think “belly fat” ads on un-moderated generic message boards.

The Multi-User, Multi-Platform Conundrum
BT relies on the placement of a profiling cookie on a computer. That cookie is used to collect information about a user’s browsing behavior.

BT is an imperfect science. There is simply no way to substantiate and verify exactly which set of eyes is currently placed in front of any given computer screen at any given time. The notion of blindly placing media on a site based on the “user profile” of a computer is just that – blind and, at best, inefficient.

Proprietors of BT reportedly go to great lengths to establish profiles on individuals in an effort to predict their buying patterns and exploit them. The data collected on a user typically comes from a single or mere source and the sample size is usually far too small to provide an accurate portrait of any one person’s buying habits. The profile outcome is just too subjective and cannot be interpreted with a high degree of accuracy.

The assumption that is made by BT is that only a single user is responsible for the “behavior” of a solitary machine. That assumption immediately encounters issues when you consider that a vast number of users actually only rent computer time at internet cafes or that entire families may share access to the Web through a single household computer. How, then, can BT be truly qualified as a viable and accurate information gathering and targeting method?

Beyond Targeting: Where Does Optimization Fit into the BT Equation?
Even if BT were a perfect science (which it is not), the notion of optimizing your BT media spend is still a valid point. What does a BT vendor do to optimize the placement of your campaign and the performance of your media spend?

The Consumer Privacy Debate and Putting all of One’s Eggs in a Single Basket
BT is currently under close scrutiny by a number of federal regulatory agencies and can therefore be assured to be subject to certain government regulation in the near future. There is no saying how this may predict the method’s future (for good or for bad).

Having all of one’s eggs in the BT basket may not be the strongest position in which a marketer can reside. It is no secret that many media buyers are not presently concerned about BT privacy issues, and have no plans to curtail their use of BT until the FTC restricts it – but what will they do when that happens, especially if they have had no prior experience with BT-free alternatives?

Prepared planners may want to consider testing alternatives now in preparation for what the future will hold. Many of them will be pleasantly surprised to see that there is no performance difference between BT and many BT-free alternatives. It is important to note that dynamic response-based optimization combined with premium high-visibility inventory can actually be a far smarter (and superior) approach to BT – a fact that can only be verified through testing.

Am I Suggesting that BT is Bad?

BT is not always bad, and this is by no means what I am suggesting. What I am suggesting is that putting all of your spend into BT alone is bad. It’s risky – the practice may not be around forever. It’s tunnelled – there are qualified buyers and prospects online that BT does not provide access to. And, it’s potentially a barrier to improved performance since using only one tactic offers not point of reference or comparison.

One thought on “Rethinking the Application of Behavioral Targeting

  1. Nice article Julia and it’s very fair to offer a view that contrasts the active set of BT evangelists out there. I for one don’t have a problem with BT but I share the opinion that while heavily hyped the current methords have their limitations and surely are not for everybody.

    At Peerset we decided to evaluate the merits of focusing on psychographics(interests) vs. behavior (intent) and very much because of some of the concerns you highlighted in your blog post. Marketers should desire to build relationships with people NOT webpages and NOT machines so they should strive to understand the makeup of the complete person. We believe that learnings from social media now make this possible.

    What I hope for is that the privacy debate and FTC hoopla doesn’t paralyze the market and that as a community we continue to find ways to marry the tremendous experience from hundreds of years of tradiontal marketing and customer segmentation with high velocity data and expression that the internet has made possible to analyze.

    CEO, Peerset
    “the science of infinite connections”
    San Francisco and Toronto
    Twitter @peerset

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s