Over the past 18 months, online publishers have made great strides in increasing response rates, maximizing available inventory, and improving advertising revenue. Just take a look at the recent performance numbers from Yahoo and MSN as two examples of how the online media space is really turning the corner.
What’s behind the big improvement? My theory is a combination of seasoned media industry management and smart use of new technologies such as behavioral targeting are driving better audience management and delivering stronger results for marketers. And when marketers find value in a solution, the money begins to flow.
The two major suppliers in the site-centric audience management space are Revenue Science and Tacoda Systems. Why should an online agency professional care about the tools publishers use to secure and deliver inventory for clients? After all, if they improve performance, why worry how they do it?
If you really want to ensure you’re getting the best results, it pays to take time to understand how the tools work and how you can use them to your client’s advantage.
What’s Good Behavior?
Before we get into the details, it’s important to note there are two key aspects to behavioral targeting:
- Explicit: Based on the data a user provides with profile registration or recorded actions on a site, such as email newsletter sign-up
- Implied: Data derived from observing an user actions as she interacts with the site
Let me illustrate the difference. A grandmother visits Amazon.com to shop for a CD for her 13-year old grandson. She buys the latest from Modest Mouse. Amazon monitors Grandma’s explicit behavior and assumes she’s into alternative tunes. It targets her with concert tickets, but Grandma has no intention of attending a show. Unless Amazon gathers more in-depth data to truly understand the intent of her behavior, that ad impression for tickets is wasted.
The self-professed leading provider of behavioral targeting to online publishers, Revenue Science uses on-demand tools to identify, evaluate, and implement custom audience segments based on marketer-supplied profile criteria.
Note: The key to a successful effort is to ensure your RFP targeting criteria are clearly articulated so your sales rep secures the right audience slice.
What’s interesting about this product offering is Revenue Science only collects revenue from publishers if it delivers enhanced performance. There are no upfront costs.
By offering an end-to-end, enterprise-level application approach, Tacoda sells its services to publishers as a means of collecting audience data from existing data sources such as ad servers, content servers, email databases, e-commerce servers, CRM applications, and data analytic programs.
It merges these with site-generated data to provide marketers with a composite view of a site’s audience. Marketers can then slice and dice profiles based on demographic data, such as Zip Code, age, and gender, as well as audience site behavior, such as frequency of visits or likelihood to interact with online ads.
Consideration: Check the target audience profile data gathered from your client’s site to determine how closely you can match data points with the publisher site data. Pay careful attention to demo slices as they may measure info differently, such as Adults 25-54 versus Adults 25-49. There’s a big difference.
[Editor’s Note: Tacoda yesterday announced a new behavioral marketing product.]
Worth the Effort?
How well does behavioral targeting impact performance? Here’s a real-world example. Working with The Wall Street Journal Online on a campaign for a major client last fall, we implemented a side-by-side ad-delivery comparison by analyzing the level of brand impact measurement. A large volume of impressions was used with the control cell. The impressions were composed of regular ad placements in relevant content sections, while a separate cell of impressions was implemented with the behavioral targeting enhancement via Revenue Science.
We looked at such measurements as aided brand awareness, message association, brand favorability, and purchase consideration. The behavioral targeted cell outperformed the test cell in every instance. In the message association category, the lift ranged from 14 to 49 percent among a group of creative executions within the same campaign approach.
By adding a few profiling questions into the questionnaire we also found the behaviorally targeted cell delivered a more qualified audience in terms of income and product use frequency. Therefore, the behavioral aspect of the buy not only worked harder delivering the message, but we spent less money on a CPM basis to reach the core of our target. Sounds like a double dose of positive results, no?
The real challenge now is to ensure behavioral targeting criteria are correctly applied across all sites in the media plan. Be certain to know how a site does it to achieve the maximum return on investment (ROI).
This is one area in which some collective knowledge will help push the effort forward. What’s your experience with site-centric behavior targeting?