Breakout: Audience Extension

– trying to streamline the strategy for audience extension currently at Triad, especially going into 2014 – defining turnaround times, etc.

 

why use audience extension?

– high sell-thru rate of the site

– non-guaranteed buys (not enough inventory on own site) on Pandora

 

issues:

– biggest issue with audience extension is with inventory forecasting 

– in DFP, you have all of your forecasted segments that come from the DMP, DFP is looking at limiting the scope to only those that are explicitly targeted campaigns 

– video in audience extension is tough, especially with making sure that your ads are in brand safe environment

– blended rate slows down the billing process because they need separate lines in the IOs. (The Weather Company)

– are agencies coming back with their own reporting for doubleverify that we cross-reference? should publishers start paying their verification fees? the room was mixed about paying for these fees. but publishers need to define what their stance is and band together.

– is the ROI truly there for audience extension at scale? could be lower value and more work involved with this

– is audience extension sustainable?

 

best practices:

– work with the audience ext. companies on an IO basis so you can test out their platform and have something to compare against. 

– need to come up with a white/black list to understand where your audience extension is running thru. should you start off blocking everything and then deciding which you would like to accept

– adding minimum spend to these buys

– a lot of other publishers are still trying to refine audience extension strategy and starting to train their sales.

– predictive modeling to understand scale of audience/lookalikes

– define the audience segments you want to go to market with, and don’t let sales come up with very granular segments

 

what is a comfortable percentage of sales that should be off-site? if we’re underselling the main site and heavy-ing up on audience extension, it can be a problem. 

 

Key Takeaways / Action items: 

– test and vet out the vendors, so you can get a sense of the landscape and see what works for you

– control & security are concerns for being in a brand-safe environment

– Ask yourself: If you’re not selling out of your own inventory, should you still do audience extension?  Know you reasons and drivers for offering AE and understand what it means to you from a revenue POV

– AE is Inefficient Revenue – lower value, more work.  It’s a high cost of execution – minimum spends on contracts, map out processes and resources.  

          how are the ad ops’ team structured to handle audience extension? are there separate teams? etc.

– Predictive modeling can help publishers find lookalike segments and gauge performance as an alternative to AE