Ad targeting in Internet radio does not deliver for advertisers at the local level.  No it is not rooted in a problem with technology.  As recent announcements from Triton Digital and Abacast make clear, the technology exists.  However, since most station’s audiences are small and a significant amount is out of market listening (often over 40%) targeting will not yield impressions needed to generate significant revenue.  As a result targeting at the station level makes little sense other than possibly to guarantee to an advertiser an ad won’t be heard out of the metro.  Targeting is useful on a national basis as a combination of stations on a network can deliver results as long as the target is not too narrowly defined.  Pandora is probably the only service that can target locally without aggregation due to the size of its network.

Most radio stations do not collect listener data for their streams so the only targeting that can be done is geographic based off of an IP address.  Listeners will give up this data for something they value if it cannot be obtained from other sources, e.g. Facebook.  Some targeting is done based on station format, e.g. an AC station’s audience is primarily 25-54.  This is not always accurate.  IP addresses are not always correct as well which can result in lack of delivery of the campaign.

Mobile targeting is perhaps the holy grail as you can reach consumers closer to the point of purchase.  As with targeting to desktops the problem of scale is even a greater issue in mobile.  We have heard that Pandora is having difficulty monetizing its mobile audience which is 70% of  their total audience.  I don’t quite understand this as I would argue that engagement is greater with a mobile phone than on a desktop (you may leave your desk but typically you don’t leave your phone).  I believe that longer term we will see premium CPM’s for mobile.  Effective ad creative and proper delivery  will help.