Cross-platform media planning has been a hot topic at recent advertising and research conferences. As the ad spend moves online, buyers are more concerned that they are following their offline targets
accurately. This week and next we engage two of the major players in digital metrics, comScore and Nielesen, about some recent product announcements and how emerging tools can track consumers better
across media. This week, comScore's vice president of product management, Steve Dennen, discusses the company's Segment Metrix offerings. Behavioral Insider: What is the
Segment Metrix suite?
It is really a platform that supports different services. At a high level, Segment Metrix allows publishers and advertisers to better
target their audiences online. Just as today in a service like Media Metrix, advertisers and publishers can be targeting audiences based on demographics.
We rolled out Segment Metrix so
targeting and profiling of audiences can be done based on online behaviors. So in terms of what it supports, there are three basic services. Segment Metrix H/M/L - think about it as site level
behavioral targeting. We have taken our 120+ content categories we report on -- news sports, health, finance etc. -- and segmented the audience into heavy, medium and light users. Heavy is the top
20%, medium the next 30% and light the bottom 50%. That is based on duration spent consuming content on the sites.
The H/M/L piece we launched last year, and then earlier this quarter we
announced Segment Metrics for PersonicX. This helps bridge the gap between offline and online media planning. We partnered with Acxiom, the company behind the PersonicX. We overlaid that segmentation
scheme on our panel and we report out online activity based on the PersonicX segmentation scheme. So planners who traditionally use this kind of schematic for offline media planning can now look at
those same audiences online. If they are looking at doing a combined online/offline plan they can find [the same target] online.BI: What is an example of an offline segment
behaving in an interesting way online?
We profiled Lowes.com vs. HomeDepot.com. So something like the PersonicX segment called Country Ways indexed at a 229 on Lowes.com
and then at 160 on Home Depot. A lot of times planners have a certain scheme or targets they traditionally go after and they want to look up those targets and see what types of things online those
same segments do. BI: Publishers also can use this to identify new client bases.
A health site has a laundry list of advertisers related to health, but
they can use Segment Metrix H/M/L to identify some of these non-endemic categories of content that visitors to their site also tend to consume. They may index very highly on people who are heavy
consumers of business and finance information. This is a story a health site can build as a tool for their ad sales team. BI: What is the third piece of the Segment Metrix
The third part is custom segmentation. If a publisher has a certain way of thinking about or segmenting their audience internally, we can apply that scheme
to our panel and report out the same traffic for that segmentation scheme. It provides a tool for them to look at this segmentation model they tend to work on internally and how that bears out in the
Media Metrix data.
A publisher may want to create a segment to look at different levels of loyalty to their site versus a set of defined competitors. Maybe one segment is those who are
visiting their site 50% of the time, versus competitors who visit a competitor's at least 50% of the time, and then maybe a neutral group. So they can then study how those different segments
behave differently online and think about ways to reach and target those other segments to grow their audience.BI: Explain the methodology and data collection.
Segment Metrix H/M/L is against the panel on which Media Metrix reports. Think of it as putting a different lens on that same data set. We look at category by category, the
visitation in that category from our panel and then segmenting that population into heavy, medium and light segments. BI: Tell me about the distinctions among
heavy, medium and light users.
At a high level, it is pretty close to an 80/20 rule where the top 20% of the online viewers are accounting for 70% of the consumption. There
is a pretty distinct heavy audience. If you think about an electronics manufacturer or camera manufacturer, they would use the heavy and light segments differently. Let's say they have a pretty
high-end camera coming out and need to target that online. They can look at the photography category and the heavy viewers of that and base their planning decisions on using that segment. Whereas
maybe they conversely are coming out with a lower end or family fun camera. That is going to be more of the casual camera audience. And in that case they would look at light photography content
consumers. We do have advertising effectiveness capabilities where we can look at a test and control group exposed to an ad. BI: What methods underlie the PersonicX
It is a geo-demo based segmentation scheme. It applies the philosophy that birds of a feather flock together. So there are block groups of households that are
deemed to have common and consistent qualities from a demographic perspective. Because we are a panel and we know where our panelists reside, we are able to align them with the appropriate PersonicX
code for their block group and we aggregate that data up. Then we can report on all the online behavior based on that attribute. BI: Is there a next natural step for the modeling?
We are looking at building out our portfolio of third-party segmentation schemes. PersonicX was the first. We're looking at doing similar relationships with other
vendors who provide similar segmentation schemes. BI: Cross-platform planning seems to be in the air.
It's a common theme from the
overarching goal of getting more offline ad dollars online. There's been a lot of discussion in the industry about GRPs for online measurement and how it's valuable in putting online
measurement in terms that offline people are familiar with. BI: How much do we see into audience duplication and achieving unduplicated reach across platforms?
That is the holy grail of cross media measurement. That is an area we as a company are looking to move towards in the future with integrating online audiences with offline audiences. But in
terms of what the tool can tell you today it doesn't yet provide that level of insight. Cross media measurement is of high importance for a lot of the traditional advertisers, and we are focused
on trying to bridge that gap.