Originally posted on The Drum.
The biggest barrier to personalization isn’t AI, CDPs, cookie deprecation or even data quality, argues Zachery Van Doren, SVP product strategy & ecosystem at Acxiom. It’s the assumption that marketers should always optimize for hyper-personalization.
Winning brands have stopped chasing one-to-one personalization in every interaction and shifted to something more practical: hybrid personalization, which recognizes that your ability to understand who someone is varies with signal quality. Sometimes you can recognize someone at the individual level. Still, you may also need to consider personalization at the household, device or pseudonymous ID level for optimal results, because the value of each interaction is not equal.
A past customer who authenticated and is requesting information for a new product is more valuable than an unauthenticated customer who hasn’t visited in months. And that customer who hasn’t engaged in months is more valuable than an unauthenticated visitor who has never interacted with your brand. A personalization strategy that only functions if you know precisely who the individual is and can target them with one-to-one messaging is not worth the investment. Winning brands have evolved to scale personalization based on the quality of the signals and context available.
Sometimes that means 1:1 personalization, but other times it’s more subtle and uses contextual browsing signals or household intelligence to better predict what will resonate without negatively impacting the customer experience or brand trust if the personalization treatment is not a precise fit.
Why the old playbook is broken
For years, 1:1 website personalization and remarketing rested on a simple assumption: third-party cookies provided reliable, persistent identity that was imperfect but good enough for 1:1 hyper-personalization at scale.
That playbook is no longer sufficient by itself. Roughly 40 percent of web traffic now arrives without recognizable third-party cookie signals. For many brands, this includes their best customers.
The new playbook adds another dimension that starts with addressability. Digital marketers need to consider how to respond when they can make an educated guess about who someone is but do not know for sure. The focus becomes driving KPIs such as conversion rate, customer lifetime value and ROI when accuracy declines without creating additional risks.
From hyper to hybrid
Digital marketing teams are responding by building a personalization tool kit that allows them to adjust the level of personalization based on the quality of the digital signal. They’re accepting that deterministic recognition isn’t always possible and leveraging contextual and probabilistic methods to safely increase lift.
Instead of treating every interaction as a one-to-one opportunity, they’re building a signal confidence hierarchy. So sometimes you know who someone is and you can run your full hyper-personalization playbook. Other times, you have less person-level intelligence to work with but can make educated guesses about the visitor’s household that help you understand the visitor’s life stage, financial situation or brand preferences.
For instance, if you are an American car maker, building a signal confidence hierarchy can help you determine which vehicle lines and images to display. It can also help improve marketing performance by only retargeting visitors who likely can afford your vehicles or are loyal to domestic brands.
Like a coach calling a series of plays, you adjust your approach to the game in front of you, while being mindful of the potential risk if you call a play that’s off the mark.
This approach requires a mindset shift. The question isn’t whether marketers can personalize. The question is how aggressively they should personalize based on the quality of signal available. This adds impacts to customer experience, trust and reputation to the consideration set. When you own the journey, you can adjust the message frequency, sequencing and personalization level to ensure high-value triggered messages strike the right balance between 1:1 personalization and hybrid approaches based on signal quality and do not get buried by lower-value brand campaigns.
Your untapped signal advantage
Most marketers are still focused on individual, not household-level signals. According to eMarketer research, 65% of marketers use email and hashed email as primary digital identifiers and 52% use first-party authenticated identifiers like first-party cookies. Half use device or mobile IDs. But only 35% use household identifiers.
Most organizations have access to these less deterministic signals that could meaningfully improve personalization without requiring first-party authentication.
That said, they’re also rightfully concerned about privacy and compliance. Does their privacy policy permit them to remarket to pseudonymous visitors across all channels? If they send a retargeting email to the wrong person, are they going to invite media scrutiny or consumer complaints? Feeling confident in these answers requires building the operational muscle to match signal confidence to execution strategy.
The winning move
The winning move is acknowledging that not all interactions are of equal value. Personalization may be 1:1, but it may also be to a household, device or pseudonymous ID. This mindset goes against the industry’s promise of 1:1 marketing at scale, but it’s backed up by top analyst firms that believe the focus on hyper-personalization sets up the wrong expectation for clients and, when taken too far, actually leads to people feeling uncomfortable, resulting in a negative ROI. According to Forrester: “Consumers don’t care about hyperpersonalization…Many organizations feel compelled to achieve this mythical ideal state. But consumers only care whether an interaction with a brand is relevant to them and if that interaction yields value for them – regardless of whether that interaction arrived via one-to-many, one-to-few, or one-to-one.”
The new personalization hierarchy
Finding ways to deliver better, more personalized customer experiences is top of mind for marketers. As you explore new ways to improve the brand experience, do not just ask: ‘Can we improve personalization?’ for your use case.
Ask: Can we recognize our customers? How confident are we in the signal? Do we have permission to engage them, and in which channels? What level of personalization is appropriate? What business outcome are we trying to influence? What obstacles are preventing us from achieving it?
Hybrid personalization isn’t a compromise. It’s recognition that not every interaction deserves the same investment, and that’s exactly what makes modern personalization work.