Article: Sunday, 10 March 2013
When done right, personalisation online can be an effective way to enhance the buying or consuming experience, thus stimulating purchasing. However, it can also rely heavily on customers’ private details, a very sensitive and currently hot issue. So, how do online marketeers deal with this and other challenges?
Pioneered by online retailers Amazon and eBay, recommendations and other types of personalisation services are increasingly catching the attention of many online suppliers and marketeers of goods and services who see them as excellent tools to attract and retain customers, and of course increase sales.
Based on customer preferences, past purchases and recent browsing behaviour, personalised recommendations can var y from interesting books and films (Amazon) and selected news items (Google News), to attractive travel opportunities (Expedia).
Personalisation technology, which leverages such techniques as collaborative filtering, data mining and click-stream analysis, is mostly rules-based and triggered by customer behaviour and interaction (Prescriptive Personalisation), with insight into users coming from several sources, including customer profiles and preferences, recent activity and history, and current clicking behaviour. This type of personalisation can be either Explicit (where customers provide or are given a profile with specific requirements or choices), or Implicit (where customers’ behaviour is monitored and appropriately interpreted as they navigate a website, and tailored content presented accordingly).
Alternatively, suppliers can deploy a novel, smart type of personalisation (Adaptive Personalisation) where business logic is automatically created on the fly and no database or rules need to be set up beforehand (or subsequently maintained).
Incremental behaviour of website users is analysed to model a user or user type and the personalisation software deploys this knowledge to personalise content displayed automatically. In addition, business rules are continually evolving as user behaviour and patterns change. Significantly, Adaptive Personalisation allows a supplier to focus on content for a minority segment. This technology is in its infancy but given the benefits it offers its popularity is bound to soar in the future.
Personalisation is a win-win proposition that benefits customers and suppliers. Customers receive accurate and timely information on what they are seeking, but with less information overload. This means a more informed decision-making process and an improved buying or consuming experience. And for suppliers, this personalised experience can increase the probability of a purchase being made, as well as improve customer loyalty and retention rates.
Reflecting the increasing demand for such online services, MyBuys, a personalisation technology provider with some 300 e-retailer clients, reported in April 2011 an increase in sales of 3.4 per cent on the previous year.
Furthermore, its monthly E-Commerce Wellness Index indicated a 12 per cent rise in average order value for ‘personalised’ transactions compared with sales without personalisation. Rising interest is further underscored by a recent Bloomberg Businessweek study in which some 80 per cent of marketeers surveyed named personalised marketing as their top priority.
The key to a successful personali-sation service is knowing what data to collect, which critical variables to track and monitor, and what profiles to define in order to make sure the services are accurate and useful. In fact, the accuracy of their predicted recommendations (compared to the customer’s actual choice) is something suppliers are most keen on improving and are willing to spend a large amount of money to do so.
Customers themselves are the major source of information on which recommendations are based. The quality and accuracy of the recommendations will depend heavily on the quantity and quality of the personal information customers are willing to provide directly, or through secondary sources.
However, the biggest barrier to personalisation is privacy. Customers may be concerned that their personal information could be used for unwanted targeted advertising or sold to third parties. They may also be wary of price discrimination, unauthorised access to accounts, or government surveillance, or even afraid of the risk of harassment rather than about how their data will be used.
What is most surprising, however, are the results of the research we conducted that show customers consider the quality of recommendations more important than privacy issues. Provided personalisation quality is high enough, customers are willing to provide their private details, with subjective expectations that the supplier will offer personalised services based on their profiles, and trust that the supplier will not indiscriminately share their personal information. This is referred to as the personalisation-and-privacy paradox.
Crucially, this means that under certain circumstances the perceived personalisation quality could offset the impact of privacy concerns. That is why we strongly advise service providers to focus on improving the perceived quality of their personalisation services, while creating or enhancing trust by ensuring that customer privacy is maintained at all times – and the customer is aware of that. This can be done using privacy signs, seals and security features on their websites, especially for personalisation services involving higher transaction values.
A word of caution: if providing quality recommendations is not sufficient to convince customers to part with their personal details, suppliers may be tempted to offer financial incentives, such as vouchers or discounts. These should be used with the utmost discretion, and for good reason. Even though incentives could prove beneficial at the start, they may undermine the supplier’s perceived reputation in the long run.
Our research also reveals that some customers are willing to pay a premium to purchase from news and financial services suppliers if they are more likely to use personalisation, something businesses could leverage. In addition, suppliers who provide recommendations through greater access to different sources and better value-added services will have a greater chance of monetising their personalisation services.
On the generational side, suppliers should bear in mind the needs and wishes of the next crop of customers. Simple product recommendation may no longer be enough. Younger people are looking for personalisation based on, for instance, location, device type, or even social interaction of their peers. Services could range from personalised form and email, to personalised search. Furthermore, suppliers need to look to other sources for useful customer data, for example online clickstream, social media and social networks. Only then will they be able to combine the power of these sources together with viral marketing to offer a truly personalised experience.
Finally, there are times when personalisation may not be the answer. Playing devil’s advocate, suppliers need to ask themselves whether personalisation is for them. There are several criteria to test this, such as the size of the website, the perceived reputation of the supplier and whether well-established relationships already exist between supplier and customers. Importantly, suppliers need to assess if they indeed have enough content to personalise, and whether the benefits of personalisation outweigh the technical and business overheads, and the cost of managing content.
Looking into our crystal ball, we see finance and insurance industries offering consumers personalised advice on financial products they can choose and insurance packages that are relevant. We also see further developments in the area of personalised search (similar to features in Google Search), given that different users have different preferences and search histories.
In addition, several areas could be researched in the future. For example, it would be useful to investigate how peer ratings and reviews, supplier reputation, and privacy statements and setting could be used to increase customer confidence, supplier trustworthiness perception and the use of personalisation. Then there are technological dimensions, such as user-interface design, which needs exploring to see whether improved accessibility and presentation affects usage. Similarly, it would be useful to experiment with different personalisation styles, designs and formats to measure the impact on customer acceptance and usage. We know that recommendations displayed in a textual format can increase trustworthiness, but what effect would, say, a different presentation design or format have?
Science Communication and Media Officer
Corporate Communications & PR Manager