XA.net (formerly CPM Advisors) announced the results of a recent campaign results using eXelate data and XA.net's platform, CPMatic.com. According to to the release, "A computer retailer accessing eXelate data on CPMatic.com using multi-day variable frequency caps saw a 92% ROI lift on the targeted campaign vs. a single-frequency-cap no-data control on the same site inventory." Read the release.
XA.net's CEO Rob Leathern discussed the results of the campaign and the XA.net platform.
AdExchanger.com: What is meant by variable frequency caps? And why is this an important factor in achieving favorable campaign results?
RL: It's a combination of varying the number of times we show the ads to a given user in the intender audience and also taking into account how many times the user is actually seeing the ad (versus it showing up on the page where they cannot see it). This is not something you can manage very easily manually, at scale. After a certain point too, we can be pretty sure that intenders are actually "already boughts" or "never wanteds" if you know what I mean.
Are there a limited number of sweet spots for using intender data? If so, why? Can in-market data work in areas other than consumer electronics, autos and travel, for example?
There are a variety of variables at play that determine whether in-market targeting makes sense and have led to these categories predominating so far online, including the length of the buying cycle, the average margin/transaction size, the ease of aggregating enough data in a segment, and the likelihood of getting the user to react to an ad message in that segment. One of the reasons the eXelate-CPMatic partnership is so interesting is because the self-service component of our system (and the per-use charging model on the data) makes it quick and easy to set up lots of small tests to assess the value of other types of data.
What differentiates the CPMatic platform from other demand-side platforms in the space?
Our platform has always been open for anyone to sign-up (and immediately use) which allows us to gather a lot of feedback to continually evolve it. We want the overtaxed marketing manager at a company or agency to see us as a single place they can get media, data, and consolidated reporting with a single IO / payment relationship. We make it easy to buy ANY online media with access to every data source. And our account team is here to support them every step of the way. We’re working to make advertising more powerful for advertisers and simpler to understand and manage for marketers. Our technology allows both small and large-scale media buyers the ability to create hundreds of campaigns and tags across every available supply channel in an easy and efficient manner.
Beyond showing potential customers how easy it is to buy eXelate data through CPMatic, why did you feel the need to create http://buy-exelate-data.com ?
Our team has built ad servers and complicated back-end systems, but designing and creating front-ends to demystify things while still allow for enough control is always especially challenging. I get to speak at industry events fairly often, and even within the industry it has been difficult to get a quorum of advertisers understanding retargeting and behavioral advertising, so we have been thinking about how we can explain things better and educate people about these opportunities. It's going to continue to take time and effort to explain the nuts and bolts stuff, especially as the noise level is high around more flighty concepts.
In the release, you note, "The eXelate-CPMatic integration enables use-based pricing which guarantees that the advertiser only pays for data when the desired user is reached." What has been the previous billing/pricing method? Why move to this model?
The two main models out there are pay for data as you use it (by targeting users), or pay for data that is put into a cookie pool that you control or have access to but have to build yourself... regardless of whether you are then able to reach that person later. There are pluses and minuses to both approaches for various parties in the industry, but it is clear that payment upon use is a much better model for the advertiser. I suspect that in the next 12-24 months we will see some more variability in the types of data business models that people try, but there will always be some tension between publisher/data creator, intermediary and advertiser on the flex between creation/usage and the tracking thereof.
By John Ebbert