Answers Served: ToneFuse Talks Using Music Interests for Targeting


conversationADOTAS – According to ToneFuse, Coldplay fans are 44% more likely than average to be in the market for travel services (probably to destinations as dull as Chris Martin’s vocal melodies — ZING!). John Lennon fans are 101% more likely than average to be pet owners — there must be a lot of cats out there named Yoko.

ToneFuse has discovered this data by adding a filter of music interests to analysis of third-party behavioral data for the 60 million uniques across its network of high-traffic music-related publishers. The results are quite interesting (especially to big music fans like myself), but also have made for the cookie-less ToneTargeting, ToneFuse’s ad targeting technology based on music interests.

ToneFuse - ToneTargeting - Nicki Minaj

CEO Val Katayev took the time to explain how ToneTargeting works as well as detail ToneFuse’s proprietary ad units and contextual targeting service.

ADOTAS – ToneFuse handles brand ad representation for high-quality music sites — can you tell us what you look for in your hand-picked publisher base and give us some examples?

KATAYEV: We look at many attributes such as making sure the property is brand-safe and reaches a substantial audience. However, a unique piece of the puzzle for ToneFuse is that we make sure that a site has clearly identifiable content (ie. artist names, song titles) in order for our platform to be able to determine visitor’s music interests.

What kind of media buyers are you typically working with?

In most cases at ToneFuse Music our team works with media buyers who are either from agencies or direct that represent major brands. Our clients typically look for highly engaging ways to reach a large audience of young adults in order to tell a story about a product and build awareness.

ToneTargeting doesn’t use cookies — how do you derive information about music fans for targeting purposes?

Rather than using cookies, we cross referenced all of our music data from 60 million U.S. users with third-party behavioral data. That gave us a massive database of information about each and every music interest. Thus we know what an average fan of Justin Bieber, U2 or Lady Gaga is more likely or less likely to be interested in. Since we have this information, our targeting decisions are made based on what content the user is checking out (i.e., Lady Gaga song facts) as opposed to looking at cookie data.

What are some of the stranger trends you’ve noticed?

The power of iTunes. Before The Beatles were available on iTunes, they weren’t even a Top 30 artist in our platform. Ever since their catalog became available on iTunes, they are regularly in the first to fifth position as artist most demanded on our platform. And this is based on 160 million users around the globe.

Given the fact that our audience is young, largely the 12-34 age group with 18-24-year-olds at its core, that goes to show the wide reaching power of Apple to control the mindset of this generation — even tweaking music interests of the youth to love a band from the 60s!

How do you measure the effectiveness of ToneTargeting? How does it compare to your behavioral targeting efforts?

These days there is no single measure of effectiveness and we look at everything including various types of engagement, video view-through rates and CTR (don’t kill the messenger, clients still look at clicks) among other elements.

We built ToneTargeting to be complementary to traditional behavioral targeting. ToneTargeting provides far more reach and no privacy concerns, but its accuracy is not as high as BT, thus the product is positioned between content targeting and behavioral targeting.

If I tell you my favorite bands are Mission of Burma and Hum, can you suggest what brands should target me? Or will you claim I made those band names up?

Stop making stuff up, Gavin!

First and foremost, our system must have enough data to make statistically relevant decisions. Meaning, an artist must be at least somewhat popular for ToneTargeting to make conclusions about its fan base. Here’s what our engine is telling us about fans of your favorite bands. You can decide what brands should be using ToneFuse to connect with those fans:

Mission of Burma fans index well above average on segments of Nightlife, Art, Job Seeking, Shopping for Video Games, Renters

Hum fans index well above average on segments of Auto Buyers, Comics as Hobby, Suburban Lifestyle

What separates Flextitials from other rich media display units?

Flextitials are different because as name suggests, it gives advertisers far more flexibility to use the browser as a canvas. The creative is no longer limited to a template of an ad unit. Out latest incarnation, the Flextitial Duo, allows advertisers to launch campaigns that combine interstitials, pre-rolls and other foreground or pre-roll creative along with wallpaper skins simultaneously. Another important aspect is that we can execute a campaign in less than hour to target these creative combinations across the entire ToneFuse Music platform that reaches a massive audience.

Can you explain how ClickFuse contextual targeting works?

Think of ClickFuse as Google Adsense for music. We knew that we can do a better job monetizing music traffic because music is harder to analyze with a “one size fits all” algorithm. Think of pages displaying content about Lady Gaga’s “Poker Face” and how Google would probably display poker-related ads, or Rihanna’s “Umbrella.” Our system would display ads related to “Poker Face” the song because of the deeper integration that our music site publishers have gone through.

Plus, since we’re directly integrated and know exact what content users are looking up, we can use ToneTargeting to display the right ads to them through the ClickFuse ad server.

Did you know that R.E.M. fans are 100% more likely than average to lose their religion?

In an effort to improve ToneTargeting, we are looking for statistical and analytical scientists to improve the algorithm. You’re welcome to apply!

How about that Lady Gaga fans are 20% more likely to get caught up in a bad romance — probably with someone who wears meat as clothing?

Well, after a breakup they should hook up with Nicki Minaj fans who are 72% more likely to be single (no, I’m not joking).

Or that 50% of Milli Vanilli fans will blame it on the rain while the other half will blame it on the stars?

Ok, you win. I’m officially lost for words.

Please stop me, I could do this all night…

Please stop, I said you win.


Please enter your comment!
Please enter your name here