Are we building a relationship, or just shouting at each other?
Planners commonly talk about advertising in terms of creating a relationship with a new or existing customer. Media folks use more military-sounding words like ‘targeting’ and ‘hitting’ people… At 55, we like to think about the role of data in a more nuanced way.
It seems that everyone is talking today about how data helps us target people at exactly the time and place of our choosing. But to make the most of the programmatic tools at our disposal, we have to ask whether, once an advertiser has carefully tracked and retargeted its valuable prospects, the right choices are being made about the content of the messages served in the ad?
What are the options?
The basic rules of intent divide each consumer’s expectations when viewing an internet page into four areas: they either want to
- buy the product on the page
- use the product
- watch someone else use the product
- learn more about the product
So, for example, if a person types “Jaguar XJR” into the Google search box, you might assume that their interest falls into one of these areas. If they add other words like ‘how much’, or ‘road handling’, or ‘crash-test dummies’ or ‘gas consumption’, it might help us put that person into one of our four ‘areas’. Namely, buying, using, watching or learning.
When a customer is not a consumer…
But consider this: what if half of the people in the market are not buying for themselves, but researching on behalf of someone else? Would you send a different message to this type of ‘proxy’ consumer, as we call them?
In a recent analysis for a skincare brand, our research revealed that 50% of the consumers searching for one type of product were not enquiring on their own behalf but for someone else – perhaps a friend, son, daughter, parent. It provided an interesting creative challenge to our client: what if we are not explaining the product benefits to you as a consumer? What if we could identify the behaviours of proxies, and talk to them in a way that helped them share their insight? Or share their opportunity to obtain a sample? Not for themselves, but for someone dear to them…
The Science of Persuasion
At 55 we have, in the jargon, brought decision-tree analysis into the equation, to integrate channel attribution into the messaging architecture.
What does this mean in normal English? It means we are analysing complex behaviour patterns before we decide which message to send.
We use qualitative measures to look deeply into the intentions behind people’s online actions:
- How can we encourage responses to ads?
When we provide more affirmative messages, what actions do people take?
What is the direct effect of different ads on the path of the customer? (e.g. is there statistically significant evidence of ads changing the type of behaviour from passive detachment to active engagement?)
- How can we classify the types of engagement or consideration generated by different messages?
Based on the last 5 actions taken by an online customer, which message is most likely to lead to a purchase?
For the ‘proxy’ customer in the skincare example above, we promoted ads that are immediately shareable so that people can share content and apps with their friends to encourage investigation.
And notice that we are using more open questions than you tend to hear in statistical analyses… The range of potential responses is huge, and requires creative minds to be at the centre of the process… This approach creates far richer insights for clients and their creative partners, and puts human behaviour at the heart of the attribution analysis.
What have we learned?
- Categorise your understanding of the user’s goals, and test the creative options against the range of hypotheses.
- Look for qualitative answers, not just yes/no responses.
- You’re on a journey – there are no absolute right answers, just better nudges
- Remember, this is advertising, not a military exercise…