Personalisation, as applied to online retailing, is no longer an optional strategy. The best retailers make significant investments into platforms that offer real-time, personalised and contextually relevant offers to their users. And evidently the bottom-line benefits are clear: Amazon says that 35% of revenues come from its recommendation engine. Whilst 75% of Netflix subscribers watch the shows that its algorithms suggest. Spotify’s Discovery Weekly playlist proved so popular that it can now be sponsored and has become a revenue stream.
Online personalisation is not limited to recommendation engines or optimisation of the product/price offering. In particular, many retailers are looking into how best to personalise the presentation layer itself. The idea is, much like a bricks-and-mortar retailer like a supermarket. So as to sell the best shopping cart, you have to think carefully about how each product is presented. Applied to online retailing the challenge becomes more complex. How do you present your products, your information, your brand, in the best possible way to each customer?
Several OTAs and metasearch platforms have been exploring different ways to personalise the presentation layer in order to help their customers find, and buy, what they want.
For example, Skyscanner is experimenting with a platform that tries to dynamically determine the best placement of a new UI widget onto a screen. The widget could be a new product, piece of information, anything. Better than this, the platform will seek out different optimal placements for different customer segments. And the best thing of all is that these segments are dynamically determined by their algorithm.
Another example is Expedia. Expedia combines machine learning with image recognition to decide which photos a user sees when searching for a hotel. This is done based on their previous interactions with the site and which convert the best.
Some airlines are also starting to explore ways to personalise the presentation layer. They usually focus on how to increase ancillary attachment and revenue, either by prioritising which ancillaries are shown to the user or adapting some of the ancillary messaging. There are a few vendors whose platforms help airlines do this. And typically, they can enable only limited personalisation use cases. This often happens due to the fact that they are integrated at the UI tier rather than at the level of the digital experience services themselves.
What is dynamic presentation?
Dynamic Presentation refers to a set of tools for real-time personalisation of the experience. It uses rules, experimentation and signals from other systems and can help airlines to personalise not only the way their products and services are sold but also the overall user experience.
There are three constituent parts to Dynamic Presentation – layout, messaging and styling. A key benefit of Dynamic Presentation is that multiple parts of the experience can be customised and optimised at the same time.
The personalisation can be based on: context – such as location, calendar events, trip-related timings; user data – behaviour and personas traits; and historical data – previous search and booking history.
Dynamic layout describes the ability to optimise the user interface by presenting relevant messages and products in different ways. Components can be positioned on the page through business rules or dynamically through real-time data-driven processes like machine learning.
Dynamic layout concept driven by persona
The signposting of messaging is becoming a science within a science, as the intelligence around conversion rates gathers momentum. Urgency messaging in the hotel space has caught the attention of UK regulators and OTAs will be forced to comply with the recommendations. This will result in a different type of pop-up rather than the end of them.
Rules engines can be applied to different message sets – such as recent activity (“last booked five minutes ago”) price tracking (“this price likely to increase in the next three days”) or inventory (“only three window seats left”). Each of these messages will have a different yield profile between personas and can be presented selectively.
Example of dynamic messaging from Expedia
Exit messages can help airlines address cart abandonment and can be customised according to the user’s recent search and shopping flow. Reminding passengers about to leave a site that price and availability are subject to change is entry-level. Sophisticated airlines with the right rules and integration platform can offer various messages, from holding the price to emailing a reminder.
Dynamic styling allows multiple style sheets to be run side by side at the same time, giving the airline greater control of the overall customer experience. That is proving a great way for airlines to personalise the look and feel of their loyalty scheme according to the persona of the member. Moreover, it enables large airline groups to display different brands to the same user.
Example of CSS-driven styling from Delta Group
Airlines which limit their personalisation efforts to what happens in their offer management system will fail to take advantage of the opportunities to influence customer purchase behaviour through effective personalisation of the way their offers and messaging are displayed.
They should look for inspiration from thought-leading OTAs and Metasearch brands to look for opportunities to experiment with and deliver more tailored and engaging user experiences for their customers and ultimately better shopping cart revenues.