By Kristen Whitmore, VP, Consumer Intelligence & Analytics, Lotame
We all have areas where we wish our actions better aligned with our values. This extends to what we buy, whether it’s big-ticket items or small everyday purchases. These little ethical disconnects can be very frustrating to marketers who want to be able to target high-intent consumers, not simply those who desire to be perceived as one.
While an individual might want to buy an electric vehicle, and maybe even leave a trail of online signals suggesting that they will, if their local infrastructure isn’t set up to accommodate electric cars, they’re much more likely to stick with a petrol-powered vehicle. And while most people probably know that the bamboo toothbrush ordered online is the more sustainable option, when one realises it was left at home while on a business trip, a plastic-handled one grabbed from the local shop does the job just fine.
We see these tensions expressed in aggregate with consumer surveys, where responses often reflect aspirations more than real world behaviours. This is known as social desirability bias, and even in anonymous surveys, it skews results towards what people think the most socially acceptable answer is.
It’s a well-known phenomenon that can be corrected with some smart data wrangling. When it comes to predicting consumer behaviour, declared data(such as surveys) need to be combined with a portfolio of observed sources to keep audience profiles up to date, with data gathered from the past 30 days being particularly valuable.
Putting declared and behavioural data side by side allows us to see the gap between aspirational versus observed behaviour, which itself is a valuable signal. Knowing that an audience might feel like they are falling short of their ideals can be leveraged in campaign creative. It can also reveal purchasing factors that ultimately trump aspiration, such as pricing or convenience.
First-party data can tell the truth, but nothing new
First-party data should be treated as the foundation from which more robust audience intelligence can be built. In isolation, it can tell marketers about their current customers and programmes, such as loyalty cards, and can result in incredibly rich first-party profiles. But, without other data sources, all marketers can achieve is retention. Audience, mindshare, and sales will shrink unless there is also a prospecting strategy.
Connecting these first-party audience profiles to the wider data ecosystem illuminates cross-retailer purchasing, offline spend, wider category exploration, and the media they engage with; the list goes on. This all begins with establishing a strong identity foundation that can connect first-party identifiers to other data sources (such as publisher audiences and retail shopping data) to third-party marketplaces.
This detailed map of consumer movements allows marketers to identify high-growth audience segments by seeing which are underrepresented within their first-party profiles. By targeting these prospective customers, marketers can ensure they achieve incremental growth rather than simply selling to the same audience again.
Put analytics before activation to avoid audience-shrinking assumptions
The more we rely on assumptions, the more we will struggle with unpredictability. Consumer habits change fast, and marketers need continuous audience analysis to keep up. We see so many situations where marketers build creative decisions around a singular assumed persona, while the engagement data tells a different story. This activation-first strategy puts the cart before the horse, resulting in campaigns directed toward phantom audiences that will never transact at the expected levels.
A brand might feature teenagers prominently in a campaign, only to discover that households with toddlers are driving the bulk of interaction. Or, a product that was historically perceived as the value option might migrate into the premium space. Audience growth depends on being able to identify when assumed demographics clash with who’s actually in market for a particular product or service.
Marketers can keep up by testing new targeting attributes, comparing personas against each other, and tracking how audiences change over time. Audience data isn’t etched in stone; it’s part of a dynamic feedback loop that must be continually validated and challenged.
The good news is, we are seeing more marketers move away from an activation-first to an analytics-first strategy. Historically, data platforms were used to build audiences and push them into DSPs for activation. Now, marketers are pausing to ask themselves what can be learned about an audience before activating it.
Using data this way paints a richer picture of current and prospective audiences. For example, if we are targeting affluent leisure travellers, we shouldn’t just think about which channels can reach them, but what differentiates them from other segments. They may attend live events, over-index for health and fitness activities, or be engaged with fine arts. Such insights shape media planning and creative strategy before a single impression is bought.
We’re also seeing broad demographic labels being challenged by more nuanced behavioural attributes. Targeting ‘tech enthusiasts’ isn’t enough anymore. Marketers want to understand things like purchase intent, category migration, brand affinities, and spending patterns. Demographics are still layered in, but they’re a starting point rather than an end point.
Access to data and analytics capabilities has improved dramatically. The challenge now is differentiation. Everyone has tools and scale, today’s advantage comes from combining multiple validated data sources, understanding where that data originates, and using it to tell a more authentic story to your audience than your competitors can. By layering multiple signals in a portfolio of insights, marketers see which sit closest to the truth.
