This post outlines the components necessary to enable effective music-connected targeting.
Any company hoping to succeed at music-connected advertising will need to continually dedicate resources to building a flexible music-connected-keyword-suggestion-tool (similar to the Google Keyword Tool, but for music) that uses a combination of all the methods outlined below.
In addition, labels and artists need to target on a geographic basis to succeed. However geo-targeting constrains impression flow. More on this later.
How Google Enables Targeting... In really simple terms, Google indexes all the pages on the web by putting every word, and every phrase, on every indexable page, into a giant keyword dictionary. Google then enables advertisers, through a refined keyword suggestion tool (rock music example below), to create a list of keywords that can be matched against this keyword dictionary. Google does not auto-magically process a few keywords into an unseen, mushroomed list of keywords to target (no human intervention required); instead Google relies on advertisers to curate the list of suggested keywords.
Music-connected targeting should work the same way. Advertisers should be enabled to pick and choose the list of music-connected keywords to target as they are suggested by an advanced music-connected-keyword-suggestion-tool.
A powerful music-connected-keyword-suggestion-tool that enables optimal targeting should derive keyword suggestions from the methods (1 through 5) outlined below.
Method 1) Similar Song Keyword Suggestions
In the music business today, song promotion is a primary goal of music marketers. To derive a list of similar songs to target, an acoustic processing and matching machine would compare songs against a database of similarly processed songs to yield a constellation of similar songs. A refined keyword suggestion tool would enable advertisers to sort the suggestion list against multiple acoustic characteristics such as mood and tempo. The software to enable acoustic analysis is readily available. Song preview-play buttons next to song keyword suggestions are also necessary.
Method 2) Similar Artist Keyword Suggestions
A list of similar artists could also be derived from acoustic analysis or via social recommendation. Successfully suggesting similar artists, songs or videos via social recommendation technology relies upon obtaining a significant flow of up-to-date end-user preference data. Social recommendation algorithms have been readily available on the web as open source for 4 to 5 years now, and large accumulators of preference data (e.g.: Apple and Amazon) already have refined algorithms. Moreover the remaining source of preference data (outside of Spotify) are much smaller and are headed toward consolidation or extinction. However, the processing of aggregate Twitter data may yield a trough of user preference data.
Method 3) Within-Genre(s) Keyword Suggestions
The within-genre method relies upon a data source such as Wikipedia to extract a list of artists or titles within a genre. A within-genre suggestion tool will often either 1) suggest a very long list of obscure names and titles to select from, and/or 2) may not be up to date enough to satisfy the needs of a music advertiser. Further testing combined with advertiser curation needs to be done.
Method 4) Similar Lyrics
A music-connected keyword suggestion tool would analyze a song's lyrics against an extensive dictionary of existing lyrical phrases, and then suggest other songs and lyrics that have similar phrasing. Lyrics are consistently one of the top searched-for items on the Internet. Lyric matching is proven and available technology. Lyric matching may also yield a very long list of obscure names and titles to select from.
Method 5) Keyword (other) Suggestions
In all reality, vanilla keyword targeting (like Google enables) against words or phrases that the advertiser thinks will resonate with any given music fan have proven to be effective. Songs cross-appeal and resonate with music fans for a myriad of social, musical, lyrical and cultural reasons. When considering the breath of the English dictionary, the artist or the songwriter will always know better than a machine which keywords may succeed and fail. Third-party keyword suggestion tools may facilitate music-connected keyword generation? This needs to be researched and tested.