You’d never choose a car based just on a name, so why choose audiences that way? Making targeting decisions based on segment names degrades the data’s quality.
Brands invest substantial time and effort developing names for their products in the hopes that these will be recalled and associated with their brand benefits: Swiffer, Candy Crush, Fanta. They might become a category definer (“Do you have a Band-Aid?”) or someday even be promoted to the hallowed ground of verbdom (“Why don’t you Google it?”).
Rarely do brands expect you to evaluate and make a decision based solely on the name. But what if you did make choices based only on names? What would the names tell you that would be useful?
Try this exercise. Based solely on the name, which car is the fastest: Evora, Charade, Aventador, 675LT, Macan? Which pharmaceutical is best for treating migraines: Kenalog, Zomig, Gleevec, or Indocin? Who is our best babysitter: Sinclair, Isabel, Lilly, or Kaylie?
Trying to evaluate something based on the name alone is absurd enough in day-to-day life; now imagine trying that in the complex world of ad targeting.
Which audience is best for promoting Organic Baby Yogurt: Grocery Moms, Yoga Moms, Lululemon Range Rover Moms, or Organic Moms? All you know is the name of the segments, so what should you do?
There’s a good chance you’re looking at the list and thinking to yourself, “Grocery Moms sounds relevant but doesn’t scream organic. Yoga Moms could be organic, but if you’re in class, you’re probably beyond baby food. Organic Moms is clear, but broader than babies. Lululemon Range Rover Moms probably want organic, won’t care about the cost and will tell their friends, for some word of mouth.”
If all you know is the name of these segments, can you possibly make a solid decision about which one is best? You’d never buy a car, pick a drug, or select a babysitter based on the name alone.
But so much audience targeting actually is done this way. It’s one of the reasons data segments are gaining a reputation as ineffective. In addition to variable data quality, when all that advertisers have to go on is the name of the segment, they’re forced to make decisions without substantive information.
Brands take a great deal of care understanding their target audience. They hire firms like LRW and GFK and IRI to do very involved and expensive work understanding the signals that indicate whether someone is in target for their products and how those signals can be activated by certain messages.
When it comes to turning that analysis into buyable media signals, however, a planner or ad-ops person has limited options. One approach is to run a finger down the list of segment names and choose the ones that sounds most like the people in the client brief. Or you can break down that complicated persona from the audience segmentation study into broad fundamental components (age, gender, income, etc.), and then worry about setting the targeting machinery to find the greatest combination of those signals.
Both approaches rely on data of variable quality and assumptions about what defines an audience. Next time you’re looking at audience segments, look past the names and instead consider:
- What does this segment represent? What is the description given by the segment creators?
- Why does this segment exist? Who owns the data and what are their motivations? Is it a publisher looking for an alternate revenue stream? A data provider focused on volume? A niche provider with limited scale looking for very high CPMs? Is the data being resold?
- How was this segment built? Whose data is being used to build this segment? Do you trust in the data? Does it make sense as the core of your audience? How was this segment derived? Almost all segments are modeled, and that doesn’t make them bad. You should, however, understand how segments are built and be confident in the underlying data and approach used.
When you find a segment that works for a specific objective, use the name for what names are best: identifying, locating and recalling.
(Oh, and for those of you wondering: The Aventador is the fastest car; only Zomig is for migraines; and while we like all of our babysitters, Isabel is the best. As for Organic Baby Yogurt, I’m really not sure, although I do see a lot of real world Lululemon Range Rover Moms at Whole Foods.)
Aidan Cardella is chief product officer at ChoiceStream
Email marketing is nothing new, but as our ability to harness big data improves, so does email’s potential as a marketing channel. In this article, we discuss data-driven personalization, what it means for email marketing, and how to deliver 1:1 communications at scale.
Two weeks ago, Foursquare announced what could be the most important component of its data business: the Pilgrim SDK. So what does it do, and what does it mean for location-based marketing?
Silicon Valley loves fancy job titles. It’s just something we do, and software and technology lend themselves to it. But it’s not always helpful.
Earlier this week Metro Bank became the first financial institution to offer a gender neutral option for both its staff and customers.