'Creepy data' goes bump in the night
Halloween is a scary time for me and not because I fear hundreds of kids banging on my door and hollering for candy. I’m terrified that they won’t come, and I’ll have to eat all of the leftovers myself.
But, for marketers, there’s something even scarier than a stoop full of Elsas and Annas, Spidermen and Minions. It’s the dreaded “creepy data” problem.
Back in the 2008 presidential election, then-candidate Obama sent an email to campaign contributors and others who signed up for messages. The email said something like “Make sure you’re registered to vote.”
“This would be cool,” I thought. I clicked through, gave the campaign access my Facebook and then watched as the message came up: “Congratulations! You’re registered to vote!”
That is not creepy data.
The notice also had a share-with-friends button. As a loyal American, I encourage everyone to vote, then as well as today. So, I clicked on the “share” button, expecting to see a post reminding my friends to vote. Instead, I saw a stream showing all of my friends who weren’t registered to vote.
That is creepy data.
Third-party data is a vastly richer source of information that just the clickstream and preference data we collect on our customers. But it’s also a double-edged sword. Your customers don’t know how much data we do have on them.
Some big data houses have 300-plus data points on each person in their database: demographic, psychographic, behavior and more. Customers don’t expect that data to be out there for marketers to access and use.
Many companies buy third-party data for profiling, segmentation or modeling. These are proper uses of data, but it’s easy to overreach.
After all, you can be too smart in your data use. Remember the dad who found out his teenage daughter was pregnant because she started getting pregnancy-related mail from Target?
You can avoid this by first asking yourself, “Do my customers know I have this data, and would they want me to use it this way?”
What’s scary for a brand using third-party data is using too much data and using it the wrong way. The media blowback can be massive if customers take their grievances to social media, as Target found out.
Data misuse affects people lives and tarnishes your brand image. None of us want that.
These three steps can help you know if you have strayed into the creepy-data zone:
1.You’re describing a new marketing program, and you get a funny feeling in the pit of your stomach, or if you look back at a strategy, and it just feels wrong.
In my years of working with third-party data and hundreds of companies, I’ve learned that when someone gets that funny feeling, we usually find we did overreach. We either used too much data or forgot the customer doesn’t know we had that data.
Trust your instinct, and change course.
2. Never refer directly to data you get from a third-party source if your customers didn’t give it to you.
They don’t expect you to have this data. It applies to Facebook, mobile apps and the like. Have you ever looked at the app permissions on your phone?
As a marketer, you might, but the average user doesn’t look.
Forcing permission in your app’s tiny-type user agreement and telling the customer you’re using it are two different things. Use the “common person” test. Would this person expect you to have the data you’re using?
That’s why you should not refer to data unless your customers know for certain that you have it.
That’s where Target went wrong. The company was too blatant about letting its customers know how much data it had on them. A marketing piece – an email message or a direct-mail piece – is no place to brag about how smart your modeling is.
3. Develop a customer advisory group
Consult this group when planning marketing programs using complex data integrations or advanced segmentation. Discuss what you want to do, and ask their opinions. In other words, get a reality check on what regular people would find creepy or cool.
Naturally, you have to build your group carefully. Nondisclosure agreements and other security measures are key. You don’t need a large group, but the membership should constitute a good cross-section of your customer base.
Preview marketing materials and campaigns, brief them on big-concept ideas, and ask them whether they would expect you to know the data you present in the email.
I’ve worked with marketers who got ahead of themselves with data and made grand assumptions about how their customers will react. We all need a governor to tell us when we’ve gone too far too fast.
Creepy data doesn’t go away like Halloween. Data gets creepier the more often we use it recklessly. Always recognize your data can hurt real people’s lives. No fancy data integration is worth that.