Pandora Advertisers Use Lookalike Modeling to Prospect New Customers

Today’s marketers face numerous challenges when trying to get their message heard. Not only does it need to stand out from the rest, but it needs to stand out to the right audience if it has a chance of moving the needle for a brand.

On top of that, research has revealed that the average attention span for a human is only eight seconds--even shorter than that of a goldfish! And in today’s connected world, those eight seconds are often spread across multiple content sources and devices. As a society, we have adapted to this by becoming expert multitaskers--particularly among younger generations like Millennials and Gen Z.

But with so much vying for our attention, it’s often the most targeted, personalized messages that actually break through the noise.

Personalization Begins With Data

With over 80 million listeners coming to Pandora every month, our brand partners (like brands everywhere!) rely on sophisticated targeting to ensure their message reaches the right ears. And that is why data is so critical to marketing success.

Many marketers already get this. They know the importance of harnessing Big Data to achieve maximum impact--which is why some brands are savvy enough to capture their own 1st-party data. And if they do, it’s imperative that it doesn’t just sit stale on the shelf. Targeting technologies, like lookalike modeling, now enable brands to make use of this data in more ways.

Lookalike modeling leverages a brand’s existing CRM database to find new prospects. The modeling we do at Pandora doesn’t stop at the standard “age, gender, geo” parameters, but also takes into account behavior indicators like music preference. This approach to prospecting doesn’t sacrifice scale either. In fact, lookalike modeling dramatically increases scale without sacrificing similarity to the original data set.

We’ve seen success with implementing this technology for our advertisers. It’s shown serious promise in delivering custom, qualified listeners that expand a brand’s reach.

SoFi Drives New Leads with Lookalike Modeling

SoFi, a new kind of finance company, came to Pandora wanting to generate new leads. They desired to reach a wider audience of post-grad Millennials potentially interested in refinancing student debt. Together, Pandora and SoFi developed a campaign utilizing key advertising products to help raise awareness of their offerings. In addition to lookalike modeling, the media mix included Display Everywhere, Mobile Welcome Interstitial and Video Everywhere. Taking a sophisticated targeting approach allowed SoFi to model against their robust CRM database, creating an expanded target audience of qualified Pandora listeners who shared similar behaviors as their known customers.

Lookalike targeting provided SoFi with a custom audience of 2 million new prospects on Pandora--much larger than the initially provided data set...and it worked. The campaign successfully drove 18% higher leads and 93% higher conversions than the A/B tested campaign that targeted a broader audience of college grads. SoFi’s campaign demonstrated the power of data and targeting to deliver a message that resonated with the right listeners.

To learn more about how SoFi drove new customer acquisition with lookalike audience targeting, check out the full Success Story here: