How Much Ad Personalization is Too Much?

This is an edited extract from Data-Driven Personalization.

If you have friends like mine, they will enjoy sharing examples of marketing messages and advertising they’ve received with you, on the assumption that you—as a marketer—will have an opinion or be able to commiserate. For instance, my good friend Liz recently texted me about two video ads she was shown on YouTube. The first was for a hair serum that she was definitely going to buy—they truly understood her needs as a curly-haired woman! The ad was just one of several she had seen recently that made her feel like the YouTube algorithm was incredibly accurate. She’d bought several products based on ads on the platform, and she was starting to feel like it knew too much about her. But several days later, she texted me about a new ad she was shown: one for an American chain restaurant with a heavy focus on burgers and ribs, which Liz—as a health-focused vegetarian— would not be going to anytime soon. Although she was previously concerned about YouTube knowing too much about her, she now felt like they’d let her down with this latest ad.

What Liz’s texts demonstrated to me is the duality of customers’ expectations around personalization.

On one hand, customers want marketing messages to speak to their specific needs and to be relevant to their specific lives. On the other, there’s also discomfort, as customers’ concerns around privacy continue to grow. In the last several years we’ve seen the roll out of the General Data Protection Regulation (GDPR) by the European Union, as well as moves by Apple and other companies to limit tracking on devices and the phasing out of third-party cookies on web browsers like Google Chrome.

As marketers, especially ones who are looking to capitalize on the oppor­tunities that data and personalization bring to better serve our customers and therefore make more sales, we have to consider both the pros and cons of personalization for our customers. Not only is this important from an ethical point of view, but it’s important in building strong relationships with our customers that lead to loyalty.

Understanding the discomfort with personalization

Have you ever browsed online using incognito mode or cleared your browser cookies after you looked something up because you didn’t want it to be in your search history or you didn’t want to be targeted by brands based on something you bought for a friend? Or perhaps you’re trying to avoid being tracked when you’re browsing airplane tickets to avoid dynamic pricing?

You know that brands are there lurking, waiting to target you—and there’s something a little disquieting about them making assumptions based on your behaviours. Of course, customers do understand that data is being leveraged to offer more targeted, more personal experiences—like those YouTube ads that Liz experienced—still, they experience some discomfort in being followed around the internet, even if there is an upside. In fact, more than a third of consumers find it creepy when they get ads on social media sites for items they’ve browsed on a brand website, according to a 2018 Accenture Interactive survey. Yet, retargeting ads such as these are some of the most basic ways that marketers aim to reengage prospective customers and keep them interested.

For more than a decade, the idea of personalized content online has been inextricably tied with debates about privacy and ethics. Back in 2014, Kate Crawford, a research professor at USC Annenberg and leading scholar in the implications of AI, described “surveillance anxiety” as a part of the reality of living with big data. She described this anxiety as a fear that the data trail we leave is too revealing and also may misrepresent us. At the same time, Sara M Watson, a technology critic and fellow at the Berkman Center for Internet and Society at Harvard University, wrote about how targeted advertising felt creepy even when the ads weren’t relevant, simply because of what they seemed to identify as possibly relevant to a person. She outlined how our digital footprint is a doppelgänger that allows for both self-observation as well as self-criticism.

Provide feedback mechanisms for personalization

As we think about how our brands should navigate the balance between personalization and privacy, understanding why our customers feel uncomfortable allows us to address their discomfort and build trust. Customers want to be known, but they want to be known accurately. My friend Liz was happy to be known as someone with curly hair, but she was disappointed by the thought that YouTube thought of her as the kind of person who would go to a burgers and ribs restaurant.

Giving customers the power to give feedback based on our assumptions, not only allows them to mitigate their discomfort by providing a response mechanism to our assumptions, it also improves our data so we can provide better recommendations in the future. For instance, you may have seen the menu in the corner of ads from Google and Meta that allows you to hide ads and share feedback. In addition, both Google and Meta also do occa­sional surveys to improve their ad experience for users. Similarly, platforms ranging from Netflix to Zendesk (a provider of customer service and knowl­edge base tools) provide feedback mechanisms on every recommendation with their thumbs-up/thumbs-down buttons. Each of these touchpoints allows customers to improve the data that the brands have, providing a better experience for the customers and improving future conversion rates for the brand.

For brands that deal with sensitive information such as health-related topics, going a step further to allow customers to filter out certain keywords, categories, or topics, can also go a long way to building trust. For example, with all of the emotions that can come with a holiday like Mother’s Day (including the loss of a family member, infertility and estrangement), brands including Levi’s, DoorDash, Ancestry and Kay Jewelers have allowed their customers to opt out of marketing messages related to Mother’s Day. These brands presented options via email, as well as in customer accounts, to allow them to proactively avoid receiving messages that would upset them, thereby avoiding turning off existing customers or having prospects unsubscribe entirely from all marketing messages during this period.

Being transparent and proactive in communicating with your audiences about how you collect data and how you use data is a powerful way to build trust.