Mind the Gap: Avoid These 5 Pitfalls of Personalization

“It takes twenty years to build a reputation and five minutes to ruin it. If you think about that, you’ll do things differently.” Warren Buffet

TL;DR Summary

• There is such a thing as too much personalization, especially when a customer’s data and privacy are ignored or misused. So, it’s important to understand how your customer data is collected, stored, permissioned and shared. Penalties for violating privacy rules can result in hefty financial fines. More importantly, it can have a detrimental impact on your customer's trust and brand perception.  
• Some of the common personalization traps to avoid are: overreliance on data; lack of transparency; lack of empathy; lack of diversity and equity; and violating your customer's data privacy. 
• The key to striking the right level of personalization is to be mindful of the gaps and err on the side of common sense. Apply the golden rule of personalization: respect your customer’s privacy as you would your own. Customers are already becoming cynical of how their data is being used. So, the risk of alienating customers by going too far with personalization just isn’t worth it.

It was a cool and foggy spring day in Washington D.C. where Mark Zuckerberg was set to testify in front of the U.S. Senate Judiciary Committee on April 10, 2018. The founder and CEO of Facebook, since renamed to Meta, was summoned to Capitol Hill to answer inquiries on the company’s alleged role in the 2016 U.S. Presidential Election. It was primarily in response to a New York Times and The Observer article who uncovered weeks earlier that an analytics company by the name of Cambridge Analytica had illegally harvested the private data of over 80 million Facebook users without their consent. The personal data was then used by the Trump campaign to influence millions of votes in the U.S. presidential race. It worked. Donald J. Trump would go on to win the election and become the 45th president of the United States in what was widely viewed as the most shocking U.S. election in modern history. Simultaneously, it would also be one of the largest and most significant data privacy breaches the world had ever seen up to that point. 

What happened was that Cambridge Analytica built a model that predicted the personalities of any user on Facebook by using their profile data 一 connections, likes, shares, and comments. In terms of methodology, all of this was actually pretty typical as many marketing agencies, brands, and even the Barack Obama campaign in the 2012 election were already using social media data to do similar targeting to users (although not at this scale). However, where Cambridge Analytica went offside was when they illegally harvested the data of 80 million Facebook users who then would become the subject of psychological profiling and the target of carefully crafted digital campaign ads designed to sway voter sentiment towards Donald Trump during the Presidential campaign. Gary Coby, director of advertising at the Republican National Committee, explained how it worked in a 2016 Wired interview:

“On any given day, the campaign was running 40,000 to 50,000 variants of its ads, testing how they performed in different formats, with subtitles and without, and static versus video, among other small differences. It was A/B testing on steroids. The more variations the team was able to produce, the higher the likelihood that its ads would actually be served to Facebook users."

The Cost of Pitfalls

Mark Zuckerberg largely denied any explicit wrongdoing of Facebook during the 2016 election but eventually settled with the U.S. regulators to pay a record $5 billion in fines for not keeping their user’s personal information safe. 

It’s amazing how quickly and profoundly a company’s reputation 一 and bottom-line 一 can be shattered by such a breach in consumer trust. In the aftermath of the scandal, Facebook’s stock price fell more than 24 percent which equated to an erosion of $134 billion in the company’s market value. Furthermore, a public opinion survey revealed that trust in Facebook had nosedived 66 percent 一 an all-time low and dramatic fall from its peak just a year earlier. The scandal also launched numerous investigations from other governments including the United Kingdom suspecting similar interference in the 2016 Brexit referendum. The U.K.’s data protection watchdog subsequently fined Facebook ￡500,000 for breaching their data privacy laws, which was the maximum fine allowed under their data protection act. It also sparked public outcry including a global online movement to delete Facebook (#DeleteFacebook trended on Twitter throughout the hearings).

“It’s not good. I think it’s a clear signal that this is a major trust issue for people, and I understand that. And whether people delete their app over it or just don’t feel good about using Facebook, that’s a big issue that I think we have a responsibility to rectify,” Mark Zuckerberg

For their part, Cambridge Analytica suffered a blow too big to recover from. Amid months of intense media scrutiny and a rapidly dwindling client base, they filed for bankruptcy and eventually shut down its operations in 2018. As it turned out, that scandal was a watershed moment that changed how millions of people thought about data privacy 一 what they’re sharing online, what they’re consenting to; and how data breaches may impact their lives. 

Responsible Personalization

Today, however, for the most part, customers are still willing to share their data with trusted companies in exchange for better experiences. According to an Accenture Interactive survey, 91 percent of consumers said they’d prefer to shop with brands that know their preferences and offer personalized recommendations; Three-fourths said they wanted brands to deliver a curated experience; and only 27 percent complained about companies being too invasive. But as we saw with Cambridge Analytica and Facebook, once that trust is violated, the reputation of that company can be severely damaged, even irreparable as was the case of Cambridge Analytica. To be sure, Cambridge Analytica won’t be the last. Cases of data breaches have been on the rise. In Europe, for instance, the number of fines for data privacy violations increased 120 percent between 2018 and 2021. The most common violations for the fines were tracking consumer data when the companies shouldn’t have, failing to protect consumer data, and generally flouting the mandate laid out by the regulation. 

As technology makes it more and more possible to gather personal data and create detailed personalized experiences for each individual, it becomes more incumbent on companies to use consumer data in a responsible way.

The laws are still playing catch-up with the speed of technology but virtually every country has enacted some type of data privacy law to regulate how information is collected, used, and controlled. In addition to navigating and complying with the various data privacy laws, the data landscape itself is changing. For instance, Google announced that they are stopping the use of third-party cookies 一 the ability to track users from one website to another 一 on their Chrome browser by the end of 2024. And Apple announced changes that will make identifiers for advertisers significantly less valuable. Also, more and more websites are increasingly asking users for explicit consent to track cookies when they arrive on their site to meet consumer privacy demands. Despite this, however, consumers still want personalization, so marketers need to ensure that the data they use is clean and accurate while balancing the fine line between delighting consumers and being creepy as the line is becoming ever blurrier. 

Minding the Gap

It's an important reminder of the pitfalls of too much personalization 一 from navigating the increasingly complex data privacy landscape to avoiding the common personalization traps. Afterall, there is such a thing as over-personalization and it's important to understand the potential blowbacks of going too far. How much personalization is too much? What’s the right balance between accuracy and being creepy? While personalization can provide amazing experiences that are tailored to each individual customer, leading to increased engagement and loyalty, there is a dark side. In the quest to provide personalized experiences, brands can fall into a number of pitfalls that can damage their relationships with customers and undermine their marketing efforts. Here are some common pitfalls of personalization and what brands can do to avoid them.

Pitfall 1: Overreliance on Data

One of the biggest pitfalls of personalization is an overreliance on data. Brands that rely too heavily on data risk losing sight of the human element of marketing. They may also miss important insights that are not captured in their data.

One example of a brand that fell into this pitfall is Amazon. The company faced backlash when they sent emails to women informing them that purchases were made from their baby registry. Only these women didn’t have a baby registry with Amazon nor were even expecting a baby. This caused a huge backlash against Amazon as it struck a sensitive nerve of many women who’ve been trying to get pregnant 一 or have had difficulties with pregnancy. It triggered emotions of sadness, pain, confusion, fear and anger. As a Buzzfeed news reporter said in a tweet who received the email: “I do not have a baby registry on Amazon or anywhere else. In fact, I have spent the better part of the last two years trying to get pregnant. So, getting this email was…unwelcome.” The news trended on Twitter for days with many decrying Amazon for their insensitive targeting practices to sell more stuff.

To avoid this pitfall, brands should supplement their data-driven insights with qualitative research and empathy-building exercises. This can help them better understand their customers' motivations, pain points, and values. By combining these insights with data, brands can create more nuanced and effective personalization strategies.

Pitfall 2: Lack of Transparency

Another common pitfall of personalization is a lack of transparency. Customers want to know how brands are using their data and what they can expect in terms of personalization. Brands that are not transparent about their data practices risk eroding customer trust and damaging their brand reputation.

An example of this was in 2012, when Target made headlines for an unfortunate incident where the company mistakenly revealed the pregnancy of a teenage girl to her father. Target's predictive analytics algorithms had identified certain shopping patterns of the girl, indicating she might be pregnant. As a result, Target sent pregnancy-related coupons to her home where her father found them and was shocked to learn of his daughter's pregnancy.

This triggered a huge backlash against Target and made big enough news that it spurred a conversation on privacy laws. While Target didn’t break any of the existing laws, it illustrated how predictive analytics and personalization has the power to deeply affect people, and when gone too far, can make people feel that their privacy was invaded.  

To avoid this pitfall, brands should be clear about their data collection and use practices, as well as how they personalize their customer experiences. They should also give customers control over their data and personalization settings.

Pitfall 3: Lack of Empathy

Personalization can feel manipulative if it lacks empathy. Brands that prioritize personalization over customer needs and preferences can come across as pushy and invasive. Customers may feel like they are being stalked or manipulated.

An example of a brand that fell into this pitfall is Spotify. In 2017, the music streaming service faced criticism when it launched a marketing campaign that used customer data to create humorous billboards. While some customers found the billboards amusing, others found them invasive and creepy. Spotify responded by acknowledging the feedback and adjusting its marketing strategy to be more empathetic to its customers.

To avoid this pitfall, brands should prioritize empathy and customer needs over personalization. They should use personalization as a way to help customers achieve their goals, rather than as a way to push products or services.

Pitfall 4: Lack of Diversity and Inclusion

Another pitfall of personalization is the potential for creating a lack of diversity and inclusion. Personalization algorithms can learn from past behaviors, preferences, and demographics to make predictions about what a particular user might be interested in. However, if the data used to train these algorithms is not diverse, the outcomes can be biased and discriminatory.

For example, a study found that Google's advertising algorithm showed higher-paying job ads to men more often than to women. This is because the algorithm had learned from past data that men were more likely to click on these ads than women. As a result, the algorithm perpetuated gender-based pay discrimination, even though this was not the intent of the advertising campaign.

To avoid this pitfall, brands should ensure that their data sets are diverse and inclusive, and that the algorithms used to analyze them are free from bias. Companies should also regularly audit their algorithms to identify any instances of bias or discrimination and take steps to address them.

Pitfall 5: Data Privacy

Personalization requires collecting and using customer data, which can lead to concerns around data privacy and security. If brands do not properly protect customer data, they risk losing customer trust and damaging their reputation.

To avoid falling into this pitfall, brands must prioritize data privacy and take steps to protect customer information. They should have clear data privacy policies and ensure that their data collection and usage practices are transparent and ethical. Brands must also obtain valid user consent for collecting and processing their data and provide customers with the ability to opt-out of data collection and processing.

Since the Senate hearings in 2018, Meta 一 who owns platforms Facebook, Instagram, and WhatsApp 一 continues to be the poster child for massive data privacy breaches. In Europe, Meta’s behavioral ad targeting practices were found to have violated several of GDPR’s data privacy laws on user consent. In 2022 alone, Facebook was hit with a €265 million penalty for a data-scraping breach, a €60 million penalty for cookie consent violations, and Instagram was fined €405 million for violating the privacy of children. In total, Meta is estimated to have paid about €750 million in data privacy fines in 2022 alone. And according to recently released financial statements, Meta has set aside a total of €3 billion for data protection fines earmarked for 2022 and 2023, suggesting that much more hurt to the bottom line is expected. In the U.S., on December 23, 2022, Meta agreed to settle a record $725 million class action lawsuit filed by Facebook users for the Cambridge Analytica data leak. It was the largest recovery ever achieved in a data privacy class action and the most Facebook has ever paid to resolve a private class action.

Putting It All Together

Personalization can be a powerful tool for brands to create meaningful and authentic connections with their customers. However, to truly reap the benefits of personalization, brands must avoid the common pitfalls that can undermine their efforts. By being transparent about their data practices, balancing personalization with other marketing strategies, investing in customer research and empathy-building exercises, and ensuring that their recommendation algorithms are accurate and relevant, brands can create personalized experiences that truly resonate with their customers. Ultimately, it's up to brands to strike the right balance between personalization and privacy, empathy and manipulation, and accuracy and relevance to create truly exceptional customer experiences that drive long-term loyalty and growth.

Louis Cho is a globally experienced Marketing, Data & Analytics, and Customer Experience Executive with 20+ years of experience in leveraging data, digital and technology to drive customer loyalty, engagement and growth.