In the email space, we always try to encourage our clients to send relevant content. The tools exist to send personalized messages based on data including demographic, behavior, or activity. From a consumer standpoint, it is extremely frustrating when we KNOW that marketers have the tools and data to send relevant messages, and are too lazy to take the time. If I purchased a Printer Model 432, why are you trying to offer me an upgrade package for model X, or ink refill for model Y.

If the tools are there to make it simple in the email space, can you imagine what it is like for a broadcaster to serve up relevant content? Well one broadcaster that I respect is doing such that: Pandora

Pandora has been around for a while, and a few weeks ago, I had the unique opportunity to meet Tim, the Founder. It started in 2000 with the Music Genome Project (MGP), in an attempt to discover, classify, catalogue, and expose local artists. Today, they employee 40 full time music experts, who review and classify 140 genes (30 on vocals alone) to every song in their library. Their collection is already up to 500,000 songs.

Pandora is the web user interface for the MGP, and subscribers (free) can enter in an artist or song that they like, and the system will serve up other songs that it predicts the user will enjoy. As songs are streamed, you can thumb up/down selections, it in/turn learns your preferences, and adjusts the play-list as your personal radio station.

So why is this important, and more specifically, in the context of marketing?

1. The consumer actively participates in being targeted with content.

2. Why is this model only being used with music?

3. If we know all the attributes around the content, how does that affect the advertising?

First of all, most people that use the web understand to some extent that they are being tracked. Browsers (consumers) either choose to accept it, simply ignore it, and constantly make decisions to exchange data for content. When targeting recipients with email marketing, we do not actually come out and say, “I know you went to this page on my website, so here is an offer,” but something a bit more discrete. My point: Pandora users are willingly targeting themselves with relevant content.

Why is this not being used in other industries? Netflix knows your movie history, and regardless of whether you liked or disliked particular movies, can calculate a trend of your preferences. Tivo knows the TV shows you record, so why don’t they serve up a personalized TV Station. MSNBC knows the type of news that you read, so couldn’t they adjust the content based on your history? Imagine if news articles were assigned attributes similar to songs (NOTE: I know you are thinking “tags,” but I feel that attributes take it one step further), and I could subscribe to an RSS feed for only news articles that related to dogs, that contain a valuable life lesson. Or maybe filter out my news from anything that deals with Murder and/or Space Travel. Furthermore, since Pandora chooses the attributes and serves up the content, they have been able to avoid the “rich get richer” phenomenon of many “most viewed/top rated/top linked” social engines, and encourage discovery.

A down to earth example is a restaurant model that I have been thinking about for a while. Based on things I like, do not like, allergies, what I had for dinner/breakfast, or my mood, the chef serves up a relevant meal at a relevant price (prix fix preferably). I get to skip that entire process of waiting for a menu, reading it, considering what I am really in the mood for, balancing my desire against the cost, and wondering if it will ultimately meet my expectations (albeit, I skip that process in exchange for a bit of preference privacy….).

If a consumers experience is enriched, life made simpler and more enjoyable, then my goodness maybe they are more open to targeting than we can imagine. Let the segmenting begin.

Finally, can advertising keep up? Pandora has painstakingly identified attributes around specific music, and those attributes have been associated with consumer profiles (at least playlists/ stations). Since that is the case, can advertisers take the same complexity of data, and serve up adverts with offers and creative both tailored to the attributes of that play-list? I will be watching my Rage Against the Machine vs. Madonna stations to find out.