Taking the mystery out of digital personalisation

For retailers and brands, the word ‘personalisation’ conjures up different ideas for different people;  some see it as a named greeting on a website, some as product recommendations, for others it can be a completely tailored digital experience – or a combination of all of the above.

The common thread that runs through these ideas, is the desire to use known data to provide a more relevant, engaging experience for consumers, because doing this is proven to improve numerous business metrics, from acquisition and conversion rates, to loyalty and customer lifetime value.

Let’s consider a recent study conducted by Boston Consulting Group that looks at the personalisation practices of over 50 popular brands. The results are astounding: the report estimates that companies excelling in personalisation today already experience a 6-10% increase in revenue as a result.

>See also: Let’s make it personal: the UK retail digital revolution

What’s more compelling is that the research goes on to suggest that over the next five years, across retail, financial services, and healthcare, “personalisation will push a revenue shift of some $800 billion (over £600 billion) to the 15% of companies that get it right.” In short, personalisation is not about growing the pie. It is about cutting yourself a larger slice.

As ever, the devil is in the detail when it comes to achieving great personalisation in practice – something that many brands and retailers will find daunting. But what else is holding businesses back from embracing personalisation and the benefit it carries? Let’s take the mystery out of personalisation…

Certainty is not guaranteed

In our multi-channel, multi-device digital age, ideally every interaction with a customer, consumer or client would be certain. Not only would businesses have perfect knowledge about a consumer – where they live, what they have purchased in the past, what images are most compelling to them, etc. – they would know exactly what action is going to convert them into a buyer. Sadly, despite improvements in data architecture, this is still not the case.

 Take a new visitor to a website as an example. From one click you can already tell where they came from, what page they visited and determine whether they have purchased something from the site before in the past. This still leaves a lot of uncertainty. So, learn from this when a second new visitor arrives. And the third and then the fourth. Each time tweaking the landing page using learnings from previous new visitors. By the 1000th person you know a lot more and you can make bolder conclusions about what you’ve learned. However, few businesses want to wait  for all 1000 new interactions to occur before they act on each piece of insight that is gained.

>See also: In-store personalisation: is it creepy or cool?

Making good decisions in the face of uncertainty is the foundation for differentiating your brand – because certainty is rarely guaranteed. The good news is that the right personalisation architecture can handle your constantly changing world without needing to drain uncertainty out of our interactions.

Segmentation:  A good starting point, but inherently limited

 Most retailers start by looking at ways to improve optimisation, where brands are looking for ways to create an uniform experience (that is, an experience that’s the same for everyone). To improve conversion, business’s try one or more alternatives. Then they can see how well the experience works for that same universe of everyone, and pick the one that performs best.

The problem with this approach is that it assumes everybody looks exactly like the average person in that universe. When you create an improvement on the collective experience, it means that in order to make it better for some people you had to make it worse for others. Quickly you hit a point of diminishing returns.

The natural thing to do from there is to say, “If I’m hurting Peter to improve Paul’s experience, then I should think about working on these different parts of the population separately.” This is called segmentation. This might start with just one or two, but could quickly move to tens or hundreds of segments.

Retailers could end up doing the exact same thing over again, improving the average. Then they might say, “Here’s the midpoint of each of these populations. This becomes my universe.”  This offers an improvement over the model of treating everyone the same, but it too has limitations.

>See also: Bringing meaning to personalisation in the context of B2B selling

Each segment you work on is as resource intensive as what you were doing for the whole audience. In fact, it’s more than that, because you have to discover the segments and prioritise them and build infrastructures for giving different things to different people. Your potential benefit is only the incremental improvement to the average for each of these segments.

That’s where segmentation struggles. The work scales linearly. If you have three segments, you’re doing 3x the work, but you can’t get 3x the return because you are working with smaller and smaller groups of individuals.

The 1-to-1 personalisation myth

 Many approach personalisation with a flawed segmentation framework, interpreting it as creating a “segment of one.” Imagine offering ice cream to a stadium of fans. In the flawed “segment of one” model, a business would have to create a customised flavour of ice cream for each person. One person would have vanilla, the next chocolate chip, a third strawberry. This is clearly not scalable when considering personalising experiences across every website interaction, mobile inquiry or email communication.

Believe it or not, true 1-to-1 personalisation is something different. 1-to-1 personalisation is about individual decisions, not about building unique, one-off experiences. Instead of trying to create a new flavour of ice cream for every person you meet, you need to decide which of all the available options you have on hand to offer a specific person.

>See also: Technology will revolutionise the retail experience

In the ice cream analogy that might mean having a dozen flavours to choose from and knowing which is going to appeal most to each person in the audience. The magic comes from building a flexible model that allows us to make a completely unique decision for everyone, even when what we know about them varies greatly.

This is a significant mind shift from thinking of personalisation as a segment of one. And it’s absolutely imperative retailers break the old thinking if they want to operationalise personalisation at scale.

Celebrate the diversity of your customers

Personalisation doesn’t have to be complicated or complex. Many retailers have already successfully implemented segment based personalisation, with dramatic improvements in business results.

As retailers such as Club Monaco and Jack Wills move towards 1-to-1 personalisation, driven by machine learning, the impact of the discipline in the way brands and retailers do business, will only increase. So, now is the time to see your customers, not as segments, but as individuals, interact with them uniquely and reap the rewards that this brings.


Sourced by Mike Harris, SVP & GM EMEA, Monetate


The UK’s largest conference for tech leadership, TechLeaders Summit, returns on 14 September with 40+ top execs signed up to speak about the challenges and opportunities surrounding the most disruptive innovations facing the enterprise today. Secure your place at this prestigious summit by registering here

Avatar photo

Nick Ismail

Nick Ismail is a former editor for Information Age (from 2018 to 2022) before moving on to become Global Head of Brand Journalism at HCLTech. He has a particular interest in smart technologies, AI and...