I like Tom Hespos. He gets Behavioral Targeting. He’s still married to the event-based application of it, but then again he is working within the confines of what is practical and available today. That which is made available by the primary players like TACODA, Advertising.com and DoubleClick (Boomerang).
Tom put out an article earlier this week in iMediaConnection, “Using BT to Dodge the Undesirable” with which he addressed the idea of reverse-engineering BT to eliminate people who are undesirable through online advertising. Conceptually he did an interesting job of presenting the idea, but I am not sure that with the technology he is proposing to leverage, it is necessarily possible. Or maybe it’s possible but not really reliable.
What am I talking about? I will assume that you’ve read the article, if not, click on the link above and read it. Pay homage to Tom and then come back and read this post…
Event-based targeting presumes that you can make assumptions based on previously tracked events. If an individual lands on certain page that is associated with certain events, then that browser can be associated with those events going forward. For example if the page is a thank-you page following a transaction, then it may be assumed that the individual is a customer. Well with Tom’s example if the page that is landed on is a transaction-cancellation page, then that individual could be cookied and identified as a non-customer and categorized as an undesirable target for future advertising.
But that is a real stretch in terms of assumptions don’t you think? What if the individual is a previous customer? What if that individual is a highly valued, previous customer who just abandoned that particular transaction? If you cookie them based on a particular event and then use BT to target them based on that single event, you may be eliminating very strong prospects. This is one of the inherent weaknesses of BT. It is single-event based.
Another weakness of BT is the basic cookie technology foundation that it is built on – third party. See some of my previous posts that talk about third-party cookies. For example “How Does Re-Targeting Work?” explains the way TruEffect’s DirectServe or First Party Ad Serving uses First Party Cookies to track behavior and target users with online advertising. The “Universal Cookie” introduces the idea of serving ads out of a first party domain. And “What’s Wrong with Behavioral Targeting” cites a statistic that over 40% of the third-party cookies get deleted by ad-ware and anti-spyware software each month. That means that event-based behavioral targeting could be effective 60% of the time with the other 40% of the people already having the third-party tracking cookie deleted.
BT as an event-based targeting solution remains marginally effective in comparison to customer re-targeting. Tom’s idea of segmenting undesirable customers is a huge idea when you look at it, however, through the eyes of customer-retargeting. Having the ability to know who is a customer and re-target them is a value proposition well established by this blog. Tagging people with first-party cookies based on a low-value customer status introduces the opportunity to message to that audience differently as well. You can not control who your advertisements will see. But you can control what advertisements they will see.
Sure networks may try to avoid displaying your ad to certain people, by playing some other unfortunate advertiser’s ad to that individual through BT, but that individual may be receptive to a different type of ad from you if you given the opportunity to distinguish them accordingly.
There is really no such thing as a bad customer. That is an oxymoron. If they are a customer, they have value. Marketing conversion theory tells you that if you can call someone a customer, you have the ability to increase their spend potential. If you have contact information and some transactional knowledge you have the ability to test messaging and incrementally improve that relationship. Customer re-targeting online is a mechanism of doing that. BT does not allow you to do that. BT is a keep or eliminate proposition and it does not take into consideration whether someone is actually your customer. It only takes into consideration whether someone experienced an event on your web site previously. And that is not enough information to base decisions on.
_uacct = “UA-980395-1”;
urchinTracker();