Thoughts, Insights and Inspiration

Personalisation: The Thin Line Between Useful & Invasive

Data has always been the most significant tool in the marketer’s and retailer’s arsenal, and in recent years the digital revolution has made it even more important. With personalisation becoming an increasingly prevalent, and much more sophisticated, force in the way marketing is undertaken and eCommerce sites are set up, it’s becoming even easier, and even more important, for retailers to use data smartly. But can brands go too far when personalising their website for their users, and can it have an adverse effect on how much customers trust them to respect their privacy?

The short answer is: yes, certainly. With hacking and the infiltration of personal data becoming a regular occurrence with significant global ramifications, questions of privacy are everyday news. Nowadays, internet users expect their data to be taken and harvested to make for a better, more personal browsing experience, but they also know they run the risk of that data falling into the wrong hands. Finding the right balance between data collection and data protection is critical for eCommerce websites then, but how can that be done?

How is Personalisation Viewed?

The significance of personalisation to marketers is clear. In its market insight report, The Pressures of Personalisation, The Drum spoke to 200 marketers and found that 80% of those questioned believed personalisation had a significant positive impact on their business, with a further 69% agreeing that it’s very important for the customer experience.  This second point is perhaps the most important one. Where personalisation is concerned, the end user should always be the top priority, and businesses are starting to realise that, with research suggesting that 89 per cent of marketing experts believing that by 2017 customer experience will be the primary point of differentiation with their competition.

Of course, that also means that it doesn’t really matter what marketing experts or the retailers themselves think: what counts is what the customer thinks, and encouragingly customers do want personalisation. In a recent study, The University of Texas found that personalisation helps customers deal with the pressures of information overload, and allows them to feel like they have a measure of control over what they’re seeing. “I’ve shopped on Amazon for jumpers,” their thinking might go, “so now Amazon is showing me jumpers and other warm winter clothing. This allows me to feel in control and cuts down my browsing time. I win on both counts.”

It’s not difficult to find statistics to back this up. Digital Trends found that 76 per cent of customers prefer doing business with retailers that use personalisation, while a further 86 per cent told Infosys that personalisation plays a role in their purchasing decisions. But personalisation is a broad church with a steep sliding scale. We may be pleased to see Amazon recommending other warm winter clothing for us, but what about if it starts digging deeper? What if it starts advising that customer about warming foods to eat, winter drinks to drink, tips for saving on heating, and ways to make sure the car isn’t freezing in the morning? Isn’t that kind of personalisation a little invasive?

Can websites be too personalised?

In the example we’re giving, it’s easy to see why a customer might find such recommendations a breach of privacy. That customer has visited Amazon to buy a jumper. Recommendations of other jumpers are fine as it’s within the sphere of the initial interest. Extending that sphere to include warm winter clothes is also fine, as it’s only a little outside of that initial interest. Food and drink recommendations are a jump outside, but hardly key personal information, so many would be happy. Extending to include financial tips though is dubious ground. Too personal and too disconnected to the initial search, many are likely to find such personalisation a breach of trust.

Personalisation is built on dispassionate, objective data, but it’s pitched to emotional, subjective human beings. It’s difficult to know where to draw the line because everyone has a different line. One person may be happy to see tips and hints on creating the perfect healthy meal, while others may be unsettled by the fact a machine knows that they’re uncomfortable with their weight or be outraged that the site is suggesting that they need to lose weight. A personalised site should engage with its visitors in the same way one person would engage with another. It can’t understand the personality of a person in the way one would be able to in a conversation, but it can understand context.

That means going deeper than simply identifying purchases or pages visited. Personalisation has to be smarter and more adaptive than that. To avoid encroaching on privacy, a website needs to understand why the person is visiting, and context and the actual content of the pages that are visited are important in that. If they’re buying jumpers, they’re probably not interested in cutting their energy bills. Where this becomes difficult is sensitivity, so it’s important to ensure copy is well-worded and sensitive. A healthy eating recommendation should be phrased as a positive (“Delicious healthy recipes”) rather than a negative (“Lose weight with these delicious healthy recipes”). Finding the balance is hard, but not impossible, and it’s significantly eased once the following question is answered.

What does the public want from personalisation?

As digital has evolved, customers have found themselves more willing to hand over personal data and be served personalised information. Years ago, we’d have been outraged to find Google learning our searching habits and personalising the results based on them, but now it’s commonplace and accepted. Why? Because the service is inherently useful. If we’ve searched for ‘Coffee Shops in Manchester’ regularly before, Google knows there’s a good chance we live in Manchester and like coffee, so it’ll adapt future results accordingly. Likewise Netflix knows that if we’ve watched Snow White, Cinderella, and Sleeping Beauty we like Disney films, animated films, and fairy tales, and serves other films along those lines as recommendations.

Usefulness is the key: Google is helping you cut down the time it takes to work out where to get a cup of coffee, and Netflix is helping you find your next favourite film. The customer gives the business something (their data) and the business gives them something back (ease of use), and that’s the key value exchange. It goes deeper too. Data has revealed that the more useful, the more people are likely to favour personalisation: for example, 44% of those polled found in-store location deals to be a significant use of personalisation, but 74% found the idea of a salesperson greeting you by name (not inherently useful) based on mobile information creepy.

“The value exchange defines what’s OK, but as time goes on people will have greater understanding of the value in their data,” Omaid Hiwaizi, global president of marketing at Blippar, told The Drum. “Over time people will become more adept at mashing their own data sets to extract value and improve their lives – helping in this gives brands a huge role ongoing.” Touching on another key point, Claire Aldous, Proximity London’s data strategy and innovation partner, added: “It’s also important to make a distinction between machine-driven data personalisation and emotive or needs-driven personalisation. The latter involves a human intervention. You need to understand why customers make the decisions they come to, not just identify the decisions themselves.”


It’s clear then that there are two strands to making personalisation work without encroaching on privacy or breaching trust. For brands and marketers, there’s a need to appreciate both the science and the art of personalisation: gathering and harvesting the data, and then appreciating the complexities of human emotion enough to make that insight sing. Without the strength in the data, the insights aren’t strong enough and the wrong people are being targeted with the wrong thing. Without understanding the emotion, the insights are squandered with insensitive or just plain annoying messaging.

For the end customer, meanwhile, it’s all about usefulness. Everybody wants their life to be made easier, everybody wants their journey through a laborious task made shorter and simpler. Finding the right value exchange for the customer is critical for brands, because when they do, customers are willing to give a surprising amount of personal data away. Personalisation is all about delivering the right thing at the right time, and by finding that “right thing” businesses can ensure their personalisation avoids going too far.

How do you think personalisation will impact marketing? Let us know in the comments below or via social media, and don’t forget to subscribe to the blog to receive new articles straight to your inbox.


From our Blog