More Personalization Run Amuck

By on September 18th, 2012

Recently, I posted about how marketers wanting to use data to engage customers and improve ROI need to understand more than just what their customers do. They have to understand why they do it.  I received the perfect example in my mailbox the other day.

The mailing came from my power company with some helpful observations about my power usage compared to those of my neighbors within a half-mile radius. There were three lines on the bar chart:

Efficient Neighbors

All Neighbors

YOU

Big, bold letters — YOU.

I am, apparently, the worst of the worst when it comes to energy. My energy usage, I have now discovered, is 19% higher than everybody else’s. In case I didn’t get the message, there was a box next to the bar chart that said: “How You’re Doing: You used more than average!”

Gee. . . thanks.

This is a great example of how marketers should to be careful about using data out of context.  They may have the numbers, but what do they mean?

In my case, I work from home nearly 100% of the time. Therefore, I am running the heat, the electric, and the like throughout the day. We rarely eat out, so I’m also cooking at home rather than spreading my energy footprint across fast-food chains and local restaurants. After school and in the evenings, we tend to do family activities rather than going out of the home to high-energy-use entertainment. We are also only one of two families in that radius with a pool. In other words, just because we are a high-energy-use home doesn’t mean we’re a high-energy-use family.

All of these factors make the energy comparison largely irrelevant, and the way the copy was written, actually insulting. The company would have done well to consider the impact of the wording and the comparison on the wide variety of its customer base and how, for certain segments, the copy might have been worded differently in order to achieve the same result (lower energy use) without the sting. For example, it might have eliminated the “How you’re doing” line and reworded “You used more than average” to something like, “Are you in a high-use home?”

The moral of the disaster is, when producing data-driven documents for your customers, encourage them to think about not just what their customers are doing, but why they are doing it. If there are key variables they don’t know, they might want to think through the copy carefully and adjust wording to achieve the marketing goals while minimizing the opportunity for gaffs.  Otherwise, their personalization can have opposite the intended effect!

Be Sociable, Share!

    6 Responses to “More Personalization Run Amuck”

    1. Noel Ward Says:

      It’s less personalization run amuck than it is the way stats like this are compiled. The power company should be able to tell you work at home by the times of day when you’re using energy. They probably know you have a pool. They could certainly tailor their message differently, though, but that would require too much effort on their part! They’d rather just have the computer system generate a generic letter to show you how terrible you are.

      But don’t take it personally! I did one of those carbon footprint tests, figuring we’d be pretty good. But no.

      Although we have cars with 4-cylinder engines, heat partially with wood, don’t commute to work, have a decently insulated house, etc., we come up looking terrible because I fly 50,000+ miles a year and one daughter flies about 20K. All these kinds of measurement metrics are skewed to make a point, so you can’t trust any of them, IMHO.

    2. Heidi Tolliver-Walker Says:

      No fear — I don’t take it personally at all. We actually got a great laugh out of it, and it sure makes for great blog post material! But for all the reasons you mention, I would consider this “amuck” — running off using data to “personalize” materials without real thought to the consequences. My point here is that marketers have to look beyond raw numbers and understand that they are creating documents that communicate, and sometimes, raw data doesn’t give us everything we need.

    3. Vic Nathan Barkin Says:

      As they say, the more things change… The problem here is that the data used did not have the necessary demograghic differentiators to make it an effective campaign for everyone (specifically you, Richard). That said, the intended message did reach a majority of the audience in a positive manner. You, it seems are a casualty of the possible cost-benefit of the data collected and/or available. Yes, they could have used more sophisticated algorhythms to discover your specific SOHO situation, but they didn’t.

      Just for fun, here are a few more simple blunders from a presentation I made on VDP a dozen years ago:

      Due to a clerical error a British bank sent a letter to a customer’s dead wife that began “Dear Mrs. Deceased.”

      American Family Publishers mailed a sweepstakes notice addressed to “God” at a church in Florida. American Family wrote in the letter sent to Bushnell Assembly of God, “God, we’ve been searching for you.” The letter went on to say that God was a finalist for the multi-million dollar top prize.

      A Florida travel agent sent an offer for a dream Caribbean vacation to a former U.S. president – unfortunately to one that has been dead for over 100 years. The mailer addressed to “Dear Rutherford” was sent to the Rutherford B. Hayes Presidential Center in Fremont, Ohio. The offer was declined. “Rutherford’s a little lethargic these days,” a presidential center spokesperson said.

    4. Heidi Tolliver-Walker Says:

      Funny stuff! More than lack of demographic differentiation, I think my point is that data is just data, and when creating personalized campaigns, marketers need to anticipate the outliers, as well.

      If wording can be tweaked to achieve the marketers’ goal and decrease the possibilities of gaffs, that’s part of good marketing. The effort should at least be made—and in this case, it clearly wasn’t.

      Not even just outlier, but just the impact on customers in the higher energy use group. The text blocks for high-energy-use customers should have been tweaked to be less condemning. As long as the company was swapping out data, it could easily have swapped out that text block, too. It just didn’t.

    5. Robert Says:

      I have a friend that had his entire acre landscaped with water intensive landscaping and his water bill came to him with the same kind of rating as shown here, except a employee crossed out “You” and wrote in big bold letters “WATER PIG”. Didn’t go well for the person in billing as that dept only has 2 people in it,:) .

    6. Heidi Tolliver-Waler Says:

      As they say, YOU CAN’T MAKE THIS STUFF UP! I’m glad I wasn’t drinking coffee when I read this because I would have spit it all over my computers. So funny! (But not for the employee, clearly . . . )