In today’s markets, the quick identification of customer needs often translates to increased sales. This presents a challenge to commercial organisations; how do they react quickly enough, whilst making sure that the customer experience is personalised?
For a better understanding of these challenges, I examined recent academic research of Wind and Rangaswamy’s 2017 market analytics trends to identify the priorities in personalisation. So, what did I discover? Mainly that personalised recommendations are essential to business success: E-commerce marketers are adopting machine learning engines like Hybris Marketing to tailor recommendations to users’ behaviour, product popularity, and situational context. Yet, many companies still struggle to deliver relevant recommendations to their customers due to organisational complexity and an abundance of algorithm choices.
Everyone’s talking about personalised recommendations…
Findings from the 2001 paper prove that, among all possible advantages offered by E-commerce to retailers, the capacity to deliver a flexible and personalised customer experience is one of the most important. This year, a report published by Econsultancy on digital intelligence states that product recommendation and personalisation are still the priority that sits atop marketers’ list (Econsultancy).
Where’s the proof?
Amazon has reported that its well-designed recommendations have increased conversion rates by 900%. The level of personalisation used in Amazon’s recommendation strategy varies from basic geo-targeting, recommendations based on past purchase behaviour to advanced segmentation. It facilitates the division of a market of potential customers into groups that lets you isolate and analyse those subsets of data.
This enables analysts to examine and respond to the component trends in the business. To fully understand the buying patterns, the website registers customers’ interactions with products to discover underlying patterns. This collection of data insights is a top priority for businesses - almost three-quarters of organisations made it a strategic goal to collect these data insights (Gartner).
Machine learning for better recommendations
By choosing a recommendation strategy based on the signals received from each customer, retailers can apply a more targeted approach to their recommendations and realise the best return on investment. Advertisers and E-commerce businesses have the most to gain from machine-learning, because of the ease of measurement and instant feedback needed to train and improve machine-learning algorithms. A fantastic example of personalisation include recommendations that are shown dynamically on a web page depending on shopper’s interests at a given moment in time. An alternative method is the use of popularity-based algorithms - ‘Most Popular’, ‘Recently Viewed’ or ‘Also Viewed Items’.
These recommendation modules show the most popular items based on sales volumes, recent views, and complimentary products respectively. This ability to scale marketing and sales activities into a single pool of insightful data is done best by cloud platforms, i.e. SAP Analytics Hub, that are predicted to become the most profitable investment in the next five years:
“By 2020, 50% of cloud implementations will follow a multivendor postmodern ERP approach” (Gartner).
Despite this emerging awareness, widespread adoption of quality personalised insight and content are currently absent across the Internet.
Struggling to adopt a relevant recommendation technique?
There are two factors driving the lack of adoption of personalisation:
- Complexity of a product recommendation engine
- A belief that effective product personalisation method is already in place
The complex nature of a product recommendation engine may be the result of data and organisational silos. Since last year, however, there is a growing confidence with using and handling data, fuelled by the continued dominance of customer experience.
Retailers appear to be making great traction in individual product recommendations, while many industries, such as banking, are yet to consider effective personalisation. Banks are often cautious with technology due to security and risk management. This cautious approach could have an impact on their ability to recommend products in a responsive way. Often, recommendations represent ‘approved’ banking products, rather than results that are based on customer needs and behaviour.
Banks should be using Artificial Intelligence to identify what the customer wants at a specific moment in time to make a sale. A well-constructed web-page with integrated recommendation algorithms has the power to retain customer engagement, therefore making a sale more likely. So, how does it work? Through the integration of OData protocol, for querying and updating data within Hybris Marketing, the web page recalls the shopper’s browsing history as recorded on their previous visit and shows content that is relevant to their interest.
Therefore, when visitor selects a personalised product recommendation and makes a purchase, the proportion of site revenue from product recommendations grows significantly. And, as reported by JD Sports in 2014, it accounts for nearly one-fifth of all sales (Barilliance).
How can Hybris Marketing help?
Given that personalisation is one of the top marketing trends in 2017, SAP Hybris Marketing is an advisable solution to commercial businesses struggling to achieve their sales goals. Hybris marketing not only facilitates the identification of customer needs but also links customer transactions to build a full customer profile. It helps discover patterns in the data (both marketing and financial) and uses inbuilt algorithms to address two problems.
First, it is easy to use and delivers intelligent recommendations with a self-learning solution that reads customer interactions. Second, it challenges already existing personalisation by offering targeted campaigns to win back clients who abandon their shopping cart – and then converts them to a sale.