Crunching Data to Create Custom Audiences

In these times of multiple metrics, wherein brands find themselves measuring everything from clicks and visits to Facebook fans, what would benefit digital marketers most? It isn’t fewer channels and touchpoints for consumer interactions. It’s a better way to identify the consumers who are most likely to make them.

This week, Boston-based ChoiceStream – a personalized marketing solutions company specializing in online recommendations – launched a new ad platform designed to help brands locate interested consumers. Crunch uses a combination of proprietary and third-party data to build predictive algorithms that can be used to seek out relevant campaign audiences online, but that’s only where the efforts begin. Where Crunch differentiates itself is in the next step: the application of the learnings gleaned from that data to broaden the loop and create custom audience segments ideally suited to each advertiser’s campaign.

The way decade-old ChoiceStream sees it, every brand has a base audience that it knows quite well. This is typically comprised of loyal, long-term users, along with obvious potential customers who meet a campaign’s criteria. Where the real value of Crunch lies for marketers is in identifying a brand’s secondary, supplementary audience segment. These consumers are markedly less obvious, and can vary dramatically based on external factors the brand may not even be aware of.

The analogy used internally at the company paints a clear picture of its approach. “CRUNCH is like a jet plane and the people at ChoiceStream are the pilots,” says Eric Bosco, ChoiceStream’s chief operating officer. “We could put a campaign on autopilot and have it run its course, or deliver the manual and say, ‘have fun.’ But we’re in the business of providing consultative services to our media partners.”

When an advertiser partners with Crunch, the first step is to compute the ideal audience based on how people are interacting with the advertiser’s ads. Success metrics are defined, and pixel technology is used to determine when a significant event (whether it’s a conversion or a particular behavior that’s exhibited) has occurred. Next, the collected data is used to guide ad distribution. According to Bosco, some audiences remain relevant over time, from campaign to campaign, and these are preserved in a data normal database, but audiences can also change over time because, as you might suspect, in display advertising, external factors have a very large relevance in the success of campaigns.

In one example of a Crunch campaign for a cable TV subscription service client, ChoiceStream was able to use its collected data to identity and prioritize Spanish speakers as a top-performing audience for the advertiser. Two weeks later, that segment was exhibiting less interest and was de-prioritized in favor of fitness enthusiasts. This is the kind of knowledge that’s difficult for planners and buyers to gain mid-campaign, even if there’s an army of them working toward the client’s goals. In fact, the significance of these audience segments might be something they never fully realize even after a campaign has come to a close.

“Everything we find out about a brand’s target audience becomes valuable insight that it can use for programs beyond digital advertising,” says Lori Trahan, chief revenue officer at ChoiceStream. “(But) advertisers can benefit from a longer engagement to keep the benefits going. We want to partner with our customers to reach the most qualified audience so that all programs can flourish.”

When, as a marketer, you have an endless number of objectives and countless approaches at your disposal for measuring your campaigns, a fast track to finding the right audience for your needs isn’t just a time and money-saver, it’s a life-saver for your brand. No matter the quality of your ad creative or the sophistication of your media buying strategy, it’s all for not if you don’t, in fact, know who you should be targeting. This is both one of the most essential and most elusive aspects of digital media planning. The more complex our campaigns become – those multiple metrics not withstanding – the more risks there are involved in executing them.

What we really need is to know for certain that we’re putting our ads in front of the right audience. If we can discover a new audience segment in the process, all the better.

Related reading

Vector graphic of a megaphone spewing out business themed items, such as a laptop, tablet, pen, @ symbol and smartphone