In 2015, ecommerce reached such a level of ubiquity and popularity it is hardly surprising that wherever you look online there is some form of advertisement or suggested purchase. Whether you are just searching the web or browsing social media, everywhere you turn there is a banner ad, marketing email or suggested post promoting a product or service.
What is evident from these suggestions is that they are getting more and more personalised in ways that for some consumers might, at first, raise a few questions. For example, how did my favourite online retailer know that I shared an article about running the other day? Is it just a coincidence that they are now sending me offers for running clothes and new trainers?
For the more inexperienced consumer they will just put it down to coincidence. However, this is an example of how new personalisation technologies are being used to target consumers for more effective marketing and convert more sales.
In the past, personalisation solutions primarily hinged on suggested items based on a customer’s purchase history, transaction records, “wish list”, channel activity and clickstream data, promotional emails opened, loyalty card records and even customer CRM interactions.
These quasi-personalisation techniques are no longer effectual. Personalisation solutions need to be kicked into hyper-drive with campaigns that look at dimensions such as customers’ activity, interests, opinions as well as attitudes, values and behaviours. In this way companies will be able to create truly customer-specific, persuasive campaigns that produce winning results.
Attribute and event sequence analysis
There are two computational techniques currently used to achieve hyper-personalisation. Both help to bring the customer to the centre of communications based on rich and reliable profiles.
The first is referred to as ‘attribute analysis’. This breaks down customers into a map that describes every attribute of that person. For instance, whether they are male or female; how old they are; where they live; whether they have children; what their likes and interests are and so on.
This kind of analysis has become much easier with the advent of social listening technologies deployed on popular social media platforms such as Facebook and Twitter. As these have become more advanced and able to understand people better, retailers have been able to create better campaigns and tailored suggestions. For example, if someone says they are ‘eco-friendly’, retailers can translate this into suggesting energy efficient products to them.
The second computation technique is 'event sequence analysis'. This technique monitors the sequence of events a customer goes through during the buying process. So, for example this could be: the method used to login; search terms; adding items to a wish list or basket; interaction with the retailer’s social media; and deciding to buy.
When the information from these two analysis techniques is combined retailers can offer a hyper-personalised experience where customers are offered relevant suggestions at appropriate times during their interactions with the retailer. Every event during the buying process offers an opportunity for hyper-personalisation based on what is known about the individual. Interactions can be made by email, a chat bot, an immediate offer on the web page or displaying items of interest.
For example, if a customer recently bought running shoes, instead of being offered to buy more shoes at the checkout phase, through an analysis of their interests and habits, they might be sent promotional offers for running clubs, or other running equipment. Through analysis of a customer’s social media presence the retailer might know that this particular consumer runs on a Saturday morning, therefore can make sensible suggestions on a Wednesday or Thursday – allowing time for purchase and delivery. Rather than being bombarded with a huge number of ill-timed suggestions, potential customers are offered useful suggestions at the time they want and need them.
In the past, personalisation solutions primarily hinged on suggested items based on a customer’s purchase history, transaction records, “wish list”, channel activity and clickstream data, promotional emails opened, loyalty card records and even customer CRM interactions.These quasi-personalisation techniques are no longer effectual.
For retailers the ultimate goal is to make the right offer at the right time thereby making the shopping process easier and faster which in turn creates more sales.
Ensuring privacy, striking a balance
While all this personalisation sounds great for retailers and consumers there is clearly a need to strike a balance between providing customers with an outstandingly personalised experience and respecting their privacy. The level of customer intimacy that comes with hyper-personalisation can result in customer delight, but also customer discomfort. It very much depends on how sensitive the customer is, the type of data being used and also the source for that data. If customers prefer not to have their data shared or to opt out of these types of services then this should be offered. Indeed, retailers need to make sure customers are fully aware of how they are being monitored and what kind of information is being collected on them so they can make an informed decision with regards to their personal data.
In an age of heightened expectations, what has previously passed for personalisation is no longer good enough. The increasingly busy and impatient customer will regard anything they deem to be ill timed or not relevant as a nuisance, resulting in a damaged relationship with the retailer. Hyper-personalisation is the way forward, not just for retailers’ bottom lines, but also for customers’ best experiences.
Srinath Sridhar is lead consultant for analytics consulting group, Wipro Ltd