Let’s get this out of the way first: A solid micro-targeting strategy has to start with an infrastructure that can tie together multiple data sources (aka Big Data).
Micro-targeting is not new. Businesses have been doing this for ages because micro-targeting was the only targeting available: small, one-dimensional lists.
Today, micro-targeting is about pinpointing audience segments across massive volumes of massive volumes of data (no, not a typo). It’s costly and complicated, and the reality is there are very few companies executing data targeting strategies well.
Let’s put aside the nitty-gritty details of how to hire and build a database marketing organisation; which retargeting, marketing automation, or programmatic platforms have the best targeting capabilities; or whether you need Pigs, Hives, Hadoops, Sparks, or some gal named Cassandra on your team.
Instead, I thought it would be fun to look at pitfalls I’ve seen when I take off my marketer hat and step into my consumer shoes (Manolo Blahniks, of course). Some of my favourite brands are making common flubs in getting me to buy, and sites I visit regularly are missing the mark with content recommendations.
They’re simply not looking at the data in a way that accurately defines … me.
- That’s me today, but not me.
Once upon a time, I bought a gift for my nephew. Marketers now think I’m a parent of middle school-aged kids. The closest I’ve come to parenting is feeding mushy brown stuff and picking up the same after a critter that walks on all fours.
We get wiser as we age. Before defining someone as a parent or a ballet-loving foodie from implicit data, look for consistencies over time. Use real-time targeting for transactional behaviours that you can isolate aren’t behavioural anomalies or solutions to a one-time need.
If a potential customer falls out of a purchase flow, hitting them up with a better offer is a more effective, time-sensitive solution sure to bring in greater revenue.
- Kind of me, but not really.
I was fascinated by Marshawn Lynch’s choice not to talk to the press and how a person with sealed lips could drive such a media frenzy. I’m definitely not into sports. I’ve since seen an influx of advertising sports-related content suggestions and advertisements.
My one-time session with click after click after click on press around this American Football player has labeled me a jersey-wearing, foam finger-waving, Seattle Seahawks football fan.
Similar to making sure you consider the time element (which in fact can solve some of this issue), it’s easy to look at one small piece of data out of context with the big picture.
Also, keep in mind readers may come to you for one niche, but not others. If you’re marrying your data with third parties, also be aware that third-party data may dilute the picture of your true target audience.
- Been there, done that.
This week, I wore a new floral top. Guess what I saw on my computer screen? That exact floral top in an ad. Yes, I bought it. No, I’m not going to buy it again, and I’m certainly not going to buy it for my mom, or my brother, or my sister.
Nor am I going to sign up for another credit card in addition to the same credit card I already have. Nor am I going to get married again in less than one year and need to buy yet another wedding dress (my last name is not Kardashian, y’all).
I’ve seen this error over and over and over again and it is certainly a huge money sinker. It’s also very easy to solve. Even if you don’t think Facebook is your friend, retargeting pixels are.
- Definitely not me.
I recently received a piece of direct mail from a major insurance provider. The call-to-action on the envelope addressed to me is strong: UNA GRAN OPORTUNIDAD DE AHORRAR EN TU SEGURO DE AUTO!
The problem: I have four years of education in Spanish language, but by no means am I fluent in the language. I live in Los Angeles and my neighbourhood naturally has a high percentage of Spanish speakers, but I’m baffled by this clearly erroneous classification of who I am.
My guess is it is not because I am a frequenter of Mexican restaurants, or that I recently traveled to Spain, but that I somehow inadvertently ended up on the wrong list.
Or maybe, the agency this insurance company has enlisted is employing a “machine gun” marketing tactic — fire enough bullets at once, one will hit the target and maybe, just maybe, the cost per acquisition (CPA) will still net a profit. Regardless of the cause, this sort of irresponsible usage of data just reflects poorly on their brand.
- Was me, but now I'm someone else.
That dress that follows me around on all the Web sites I visit? Well, my husband’s definitely not going to buy it for me. There’s a lot of debate about how to connect data on a single user across multiple devices, but one of the simplest things to forget is that families do share computers and accounts.
When building persona-level segments, look for distinct behavioural differences or other data sources that provide hints that your current audience is not as multi-faceted as you thought. You might actually be talking to multiple distinct individuals.
I’ve had many discussions with colleagues in the publishing business about their main challenges. One of their top responses is consistently related to data. We’re all trying to solve our data challenges to drive more revenue and better engage our audiences.
My advice: Follow the 80-20 rule and strive to simplify. Test, test, test and always have a non-segmented control against your audience segment to validate whether the data powering your micro-targeting is correct.
Be the customer you’re trying to acquire.
Have some flubs to add to my list? Please share!