Privacy Alert: How Far Personalization in Advertising Can Go

Over the past decade, the public has heard it a thousand times:

If the online product you use is free, then look in a mirror. You are the product. 

Since the birth of the internet, online advertising has exploded from minor distractions filling the margins of webpages to full blown campaigns embedded within videos, peppering search results, and spliced into news feeds.

In an age where businesses claim that data is the new oil, advertisements are the fuel that makes the digital world go round. And, with the ever increasing amount of data that becomes available every day, ads have the power to become more precise and efficient than ever before.

Big data and precision advertising come at a cost, however. The end user may not be paying platforms directly to browse their websites, but they are voluntarily giving up personal information with every click and keystroke, whether they’re paying attention or not. As advertising technologies get better, user privacy keeps getting smaller.

State of Personalization Today

Advertising is important. It’s how our favorite creators can get paid for what they do. It’s how many of the free services that power our lives and society can stay free. The world needs advertising, and the world needs it to be effective and relevant. But, how far can (or should) that effectiveness and relevance go?

Businesses benefit from some amount of personalization. Products are designed with target audiences in mind. It does neither the consumer nor the seller any good for ads to be shotgunned out into the world with no direction in mind. 

Think about a giant bulletin board posted at your favorite coffee shop, workplace, or school filled to the edges with overlapping sheets of paper all crowding in for a little bit of real estate. It’s hard to read, and only a fraction of what’s on the board is relevant to you, specifically. That’s advertising without personalization — a cluttered user experience with irrelevant content that businesses paid good money to display. Everyone wants at least some amount of personalization in their ads and online experience.

And that is where online advertising is today. Cookies, behavioral data, retargeting  are just a handful of the tools used to tailor ads to suit a specific user’s interests and web history. Google keeps an “advertising profile” on every one of its registered users that tracks their digital life. Facebook, on average, has 98 data points for each user that advertisers can use to tailor who they show content to.

Not all personalization has to feel so… personal, however. Many programmatic platforms use anonymized data to tailor advertisements. 

Sometimes, they simply exist to tell advertising platforms what webpages a certain IP address is arriving from without divulging more personal details. However, that is only sometimes. More often than not, advertisers may know your exact location (and your historical location data), previous search, click, and purchase history, demographic information, and personal interests based on what you view and like on social media.

The question is, is that level of personalization enough, too much, or… too little?

Arguments For and Against Personalization

How much personalization is too much? Of course, it depends on who you ask. Despite the media coverage on the aggressive targeting habits of social media giants and legislation such as the EU’s General Data Protection Regulation (GDPR), 90% of online consumers between 18 and 64 stated that they found personalization appealing in a 2017 survey. Meanwhile, many marketers want even more personalization than what is already offered.

Effectiveness of Personalization 

On the marketing side, more personalization means more data to further maximize the odds of a successful conversion. For example, research shows that 8 out of 10 marketers who used personalized ads see measurable increases in their ad performance, oftentimes getting boosts of over 10%.  

If a marketing campaign is working with relevant and recent data to personalize their ads, the odds of success are much higher compared to non-personalized ads or even ads that are personalizing based off of older data. The more recent and current the data, the more likely a viewer is to click on the ad.

From the marketing perspective, customization and personalization works, and it works well. It is only natural that they would want to see this trend continue to increase. 

What about the other side of the coin?

Consumer Attitudes 

We already saw the statistic that 90% of internet users prefer customized ads that are relevant to them, but there is more to the story. For starters, only 31% of consumers say they wish their shopping experience was more personalized than it is now. And, while slightly over half of surveyed users said that they are comfortable giving personal information to a company website, they only wanted to do it if they knew the data would be used responsibly and that it would directly benefit them in some way.

Based on the above numbers, it sounds like consumers are beginning to reach their peak satisfaction with the level of personalization they receive, but it all depends on trust. 

According to a 2019 data security survey conducted by RSA, the consumer’s chief concern regarding ad personalization is not about the ads themselves, but about identity and password theft, sharing of health and medical information, and the safety implications if their behavior and location data were stolen or leaked. They also demonstrated concern about the selling of their data to third parties without their knowledge.

This is a reasonable concern. As technology advances, so do the techniques of cyber criminals, stalkers, and others who would misuse data. This often occurs when user interests and email contact information are  used to craft a well-designed email phishing attack that allows hackers to steal passwords and banking information. 

In other examples, a user’s location history and habits were in the hands of a criminal, it could be used to facilitate a break-in during a time period where the criminal knows the user won’t be home. This happened in 2010 when 18 homes in a single U.S. neighborhood were robbed when thieves used social media location data to track resident whereabouts, stealing $100,000 worth of goods. 

While it was not leaked advertisement data specifically that enabled the break-ins (and major sites like Facebook and Twitter have taken steps to hide and scrub the location data they collect), location tracking has become much more prevalent over the past decade and there are many other avenues criminals can use to scrape information that a user provides a website, to include data used for advertising. 

The vulnerabilities of this event caused California Attorney General Kamala Harris to tell her constituents and USA Today: “Broadcasting your location can sometimes expose you and your family to risk of theft or physical harm.”

The bottom line: consumers do actually want to see personalized ads and experiences, but they want companies to be responsible with the data that is shared. If marketers want the benefits that come with data and personalization, they will have to earn it through trust.

Technologies Required to Implement Hyper-Targeting and Personalization 

In order to handle data responsibly, it helps to understand exactly how user information is collected today and into the future. Marketing tools are capable of accomplishing so much, but only if their power is well understood and managed properly.

Ad tech is changing and evolving every day, but here are a few tools coming onto the advertising scene that are quickly becoming cornerstones. 

Artificial Intelligence (AI) and Machine Learning (ML) Powered Programmatic

Normally, programmatic advertising is done by suppliers (those who owned web space for ads) describing what type of personas visit their site and how much the supplier wants to be paid for displaying an ad. Advertisers  then describe their desired target audience and how much they’re willing to pay for ad placement. These preferences are listed on real-time bidding platforms where best matches are found and ads get displayed every time a user visits a site.

Now this process has been given a science-fiction level update thanks to AI/ML. These tools can process magnitudes more data than traditional methods interpreting user behavior and interests while predicting what ads they are most likely to click on, where they are most likely to notice them, and in what format. AI/ML analytics and prediction engines are unlocking new levels of efficiency and optimization for programmatic advertising.

Data Management Platforms (DMP)

DMPs are a software that advertisers use to gain in-depth knowledge about the behavior of their target audience. DMPs scrub multiple sources across the internet such as web cookies, website and mobile app tags, and other resources to help describe the behavior of a specific audience. This information is then used in the programmatic platforms described above for marketers to tell ad exchanges who they want to display their ads to.

Goodbye Cookies, Hello CDPs

Anyone who has ever used the internet knows the pain of dealing with pop ups asking about cookies. In short, cookies are unique strings of text stored in your web browser that tell websites where a user has been… sort of like a crumb trail. Cookies have been around for a long time, and in 2022 they are being phased out.

In the future, Customer Data Platforms (CDP) will likely become a more transparent, responsible replacement. A CDP is website agnostic and collects non-personal (third-party) information about a user. Users can also voluntarily submit more personal (first-party) information to be used. Advertisers then turn to the CDP to view user data instead of relying on cookies.

This could be a good thing for user privacy and trust, as CDPs put more data control into the hands of the user as opposed to cookies that many sites currently require in order to work.

Future of Personalization: Threat to Consumer Privacy or Dream Targeting Tool?

As great as personalization is for both the consumer and the marketer, it can go too far. The last thing a brand wants is to scare their audience with how much knowledge they possess — no one trusts a mind reader. Brands should be striving to balance respect for privacy with the relevance and personalization that consumers benefit from. 

Additionally, those who hold the data have to ensure the consumer that their personal information is in safe hands. Otherwise, a company could be risking brand damage and legal trouble if the data is leaked. Strike that perfect balance, and consumers and marketers can step into the future digital age hand in hand with confidence.