The new metrics landscape

Every year, as we look ahead to metrics needs, the list is long. This year, the list seems especially long due to emerging media technologies. It seems like each new medium brings with it its own aspects to measure.

As I have written before, all media will be digital in the future. This also means that all media will be trackable. It won't be very far off before TV, radio, print and outdoor have digital tracking capabilities in place. Whether this turns all media into direct response vehicles remains to be seen. However, we do know that those on the client side want to measure ROI wherever they can.

Increased bandwidth and storage implications
Digitisation will be enhanced by increased bandwidth, cheaper storage (an increasing number of vendors have announced terabyte-sized hard drives at affordable prices), multiple processors on a single chip (we already have quad, how high can this go?), massively parallel computing like that deployed effectively by Google, and capturing unused capacity through peer-to-peer computing where every machine on the net becomes a server. All of this will facilitate even greater consumer control and more consumer-generated content followed by consumer-generated advertising.

The increased power will also change the world of media research. Companies like Quantcast are already talking about census rather than sample for their research model. This means visibility of millions rather than tens of thousands of sites that comScore and NetRatings currently provide. It also could mean the application of concepts like cloud computing -- using massive computing power to slice and dice web analytics. Modeling of campaign effectiveness and econometrics modelling, formerly dependent upon market research sampling, will be outmoded by companies gathering digital information on a census basis.

Multiple attribution protocol
DoubleClick and Atlas have both been working on a process commonly known as multiple attribution protocol (MAP). As storage is cheap, we can now keep in the data cube all digital interactions of a consumer with a campaign. Every view, every click, every visit to a site, whether click-through or view-through. One of the reasons search is so highly regarded is monetisation. For many campaigns, the consumer sees a number of ads, visits the site and considers offers, but in the end, types in the product or service name at a search box to purchase or perform the advertised action. It has been well documented that if you just limit your campaign to search, less volume will result. Even if the search is clearly the best performing on a cost basis, it's not always scalable. MAP reporting and algorithms will permit us to take into account the underlying banner and other campaigns that happen prior to the final action. With this methodology, we'll be able to consider other actions besides the last click for partial attribution. DoubleClick has a reporting process in release now and Atlas is said to be releasing reporting and algorithms into beta in Q1 of this year.

David L. Smith is CEO and founder of Mediasmith, IncRead full bio.

 

Comments