Tuesday, October 23, 2007

Behavioral Targeting at eMetrics

Behavioral Targeting (BT), as I predicted earlier this year, has become very common term among marketers. Look at any online marketing events and you will see more than one session on Behavioral Targeting. This was also evident at eMetrics in Washington DC, where a full track was devoted to Optimization and Behavioral Targeting and several other speakers, not in this track, also talked included Behavioral Targeting in their presentations.

In my presentation, I gave an overview of behavioral targeting, what it is, why should marketers get involved with it, the difference between On-Site and Network targeting, who the players are in the market for each type of targeting and also talked about the privacy issues surrounding Behavioral Targeting. I gave several examples showing how business can benefit from Behavioral Targeting. I also outlined a process that you need to follow if you decide to engage in Behavioral Targeting. This is one of the main things that I wanted my audience to take away from this presentation. This process is an outcome of conversations with several customers who have engaged in Behavioral Targeting. To make my point, I cited an example of customer who engaged in Behavioral Targeting network. After running BT ads on one of the leading BT networks, they called me complaining that they were not seeing the value in Behavioral Targeting. After few minutes of conversation it became clear to me that they did not have a proper process in place, they did not fully understand how BT worked and what to expect (one of the reasons of why BT did not work is cited in my previous post titled Size of Your Segment and Network Reach Matters in Behavioral Ad Targeting). Result? A lot of effort, time and money was wasted. So, if you are planning on getting involved with Behavioral Targeting make sure you have a proper process in place (I will outline our process in a future post). (Note: If you were, are or planning to use Behavioral Targeting I would like to talk to you, please contact me at batraonline at gmail.com)

At the end, I showed how on-site Behavioral Targeting can be accomplished without using any third party Behavioral Targeting tools. You can use any Web Analytics tool including Google Analytics, persistent cookies and some coding to get Behavioral Targeting going on your site. This is a great way to try on-site Behavioral Targeting before spending money on third party tools.

If you missed the presentation, don’t worry you will have more chances to hear it in near future. In November I will be doing a free Behavioral Targeting seminar in Seattle so if you are in Seattle area I would love to meet with you there. The final date of the seminar will be posted on this blog and also at ZeroDash1 events page. If you would like to schedule one in your city, email me at batraonline at gmail.com


  1. Reading your previous post on BT (linked to above) and Joseph's comments lead me to (once again) state that calling this type of adserving "Behavioral Targeting" is inappropriate, at least by established marketing standards. It's much more like psychographic targeting through traditional media, namely lifestyle publications.

    I know that's not your call and you are just using standard terminology, but I really think this issue deserves discussion.
    The crux of this matter is how we define "behavior".

    A behavior in offline marketing has a known value, often a purchase of some kind. Visiting a web site and looking at a cruise page doesn't strike me as a "behavior", it's more like picking up a travel magazine and leafing through it. That's an impression or maybe an experience, not a behavior.

    Offline, BT would target people who have previously bought cruises. This is an actual behavior with a known value; finding 750 of them would be fantastic because the expected value of this audience is very high relative to marketing cost.

    However, finding 750 of them who have leafed through a travel magazine would not be such a big deal. This is because the value of this experience is quite low, in terms of expected outcome of marketing to them.

    And that's because leafing though a magazine is not a "behavior" for marketing purposes, it's the consumption of media, which has a much lower expected value.

    So with a tip of the hat to Don Quixote, I know I'm not going to change the ad media juggernaut that is BT. But I think it's useful to understand that behavior in the classic sense is not being targeted here.

    Is BT advertising more targeted than general banner ads? Sure. BT is really what I thought (around 1999) banner ads were supposed to be all along. But somehow we missed that part and just went straight to calling any space on any web page "media".

    And now, finally, we can target some of that media to specific media consumption patterns across multiple web sites.

    That's what is now called BT.

  2. Tony Evans8:28 AM

    Sorry Jim, but I'm a bit confused about your take on behavioural targeting and wonder if you're missing the point. Yes, we can argue about the precise definition of the term behavioural, and maybe 'interest targeting' may be a closer fit to what actually happens. But that doesn’t take away the enormous advantages that behavioural targeting offers over traditional forms of advertising - both on and offline.

    The whole idea behind behavioural targeting networks is that, as users visit different sites in the network and look at pages with different types of content, an anonymous profile of their interests is built up.

    Then these users can be targeting with relevant advertising (relevant, because we know what they are interested in) on any page containing an ad space in the behavioural targeting network.

    The site and the context is irrelevant.

    By the way, you can also target people who have bought something (like a cruise), requested a brochure or done something else that has a 'known value' (your definition of behavioural). It's called re-targeting and has been used for years by online advertisers.

    As you say, finding 750 people interested in cruises (because they visited a cruise site) may not be that big a deal - but how about being able to target your cruises at 7,500 or even 75,000 potential customers? Sure beats throwing a load of ads on the web and hoping someone might be interested in them.


I would like to hear your comments and questions.