Bad Personalization Is Worse Than No Personalization

Data presents an opportunity for enterprises to utilize personalization. Do it right and you can win. Do it wrong and you turn off those customers.

Meta S. Brown, Business Analytics Consultant

January 16, 2015

3 Min Read

Month after month, my electric bill scolds me. “You used 80% more electricity than your average neighbor,” it tells me one day, and “You used 100% more,” on another. It’s pointed, and it’s personal. I get the message: I’m wasteful, I’m bad.

Personalization, highly individual messaging, is hot stuff in marketing now. Marketers use personalization to make their messages relevant to the customer. When you add a product to an online shopping cart, and you get instant offers for related products, that’s personalization. Emails or texts alerting you about discounts for products you’ve viewed in the past are another form of personalization. Done right, personalization can boost conversion rates and customer satisfaction.

Good personalization motivates you to visit often, linger longer, and buy some extras. Maybe you’ve bought one of the items that Amazon recommended just for you, or ordered from a great new restaurant that you discovered in a GrubHub email that highlighted restaurants in your own neighborhood. If so, then you’ve experienced good personalization in action.

Personalization isn’t always on the mark. One online retailer offered me a size 0 dress today, while another hit me with ads for weight-loss products. For a while, I was bombarded by ads for a cougar dating site. None of these merchants got any money from me. Maybe their tactics work often enough to be profitable, maybe not.

As JC Grubbs, president of digital development and design firm DevMynd Software puts it, “Poorly executed personalization is vastly worse than letting users choose their own content.” He’s so right. Bad personalization is the reason that I have come to hate my electric bill.

The electric company thinks it knows me, and it thinks it knows my neighbors, but all it knows is the number of kilowatt hours we use each month. It doesn’t know that many of the apartments near mine are used as vacation homes and aren’t occupied most of the time. It doesn’t know how many people live in each household, who works at home, or who has kids. It happens that my household is a little bigger than the US average, and our energy consumption is a little lower than average, so those notes on the electric bill don’t do much except to annoy me. That’s bad personalization.

To do personalization well, you need to test your ideas and learn what works, and what doesn’t. You can use conventional techniques such as A/B testing to compare personalized offers to controls. Start small and choose low-risk projects at first. (JC Grubbs suggests testing personalization with a group of past customers who haven’t been active for a while, so you’ll have little to lose if your test flops, and a lot to gain if you succeed.) Once you’ve had some small successes, gradually work up to a larger scale, explore customer segments, and experiment with a variety of messages.

No personalization scheme succeeds 100% of the time. Personalized offers don’t always work as well as ordinary offers. Yet those who use personalization well are rewarded with increased revenue and happy, loyal customers. So, don’t just personalize! Test personalized offers and measure the response, so you’ll know what works.


About the Author(s)

Meta S. Brown

Business Analytics Consultant

Never Miss a Beat: Get a snapshot of the issues affecting the IT industry straight to your inbox.

You May Also Like

More Insights