Welcome to the social media shopping mall

By Karan Girotra, Ruslan Momot and Elena Belavina

http://knowledge.insead.edu: Social media and e-commerce have converged on China’s smartphones, thanks to one groundbreaking app. Since its release in 2011, WeChat has grown from a WhatsApp-style messaging client into a veritable one-stop shop for lifestyle management. 

Without exiting the app, WeChat’s loyal users – whose number is rapidly approaching one billion – can book a doctor’s appointment, hail a taxi and purchase products directly from brands. At least one in five users has linked WeChat to their debit or credit card, a potential cash bonanza of which Mark Zuckerberg can only dream.

It is only a matter of time before this functionality, or something like it, is the norm worldwide. Now is the time for brands to start preparing for a future in which social media will be not just a place to market, but also a major marketplace in itself. Our recent working paper, ‘The Use and Value of Social Network Information in Selective Selling’, describes how, unlike the current e-commerce paradigm, social media selling will thrive not on transparency and accessibility but on strategic exclusivity.

Selective selling

Social media’s most avid users include those who derive pleasure from the envy of their peers. Facebook, Instagram and others offer myriad ways to show off, from flaunting the latest designer handbag to advertising social concern through a publicly announced charitable donation.

That’s why luxury companies such as LVMH have seen a huge rise in social media use among purchasers of exclusive, limited-edition products, especially in the key growth markets of India and China. 

Naturally, LVMH has begun to use social media to identify and market to customers who most enjoy one-upping their friends. Theoretically, this select group of customers – if they could be identified – would jump at the chance to buy a product intended just for them. So far, however, LVMH’s efforts at what we call “selective selling” have been hindered by haphazard information gathering and the sheer volume of data available.

The seamless integration of e-commerce and social media represented by WeChat could hold the solution. A few companies are already starting to exploit the possibilities – for example, Tiffany & Co. uses its official WeChat account to refine customer segmentation. Hitting the bull’s-eye will depend on a brand’s ability to pinpoint which customers, out of all their social media connections, are the highest-value targets for selective selling. That’s where our research comes in.

Knowing your customers

In our study, we used mathematical models to replicate the typical behaviour of social media followers who like to show off. We also took into account the fact that firms differ in how much they know about their customers on social media. 

Some know nothing or next to nothing. But there are two types of information that are helpful: the degree of connectedness (i.e. how many friends each customer has) and the degree to which each customer partakes in consumerist one-upmanship (researchers call this ‘conspicuity’). Totally uninformed firms obviously cannot engage in any segmentation; firms who have one or both types of customer information can use all available data to curate their target set of customers.

Trying out the various combinations yielded a couple of interesting general principles. Firstly, the most popular customers were not the most high-value. Fully informed firms did best when they targeted customers with a high level of conspicuity and an intermediate number of followers: high enough for the purchase to touch off the desired envious effect, but not so high as to reduce the product’s novelty  – and, by extension, its value in the game of competitive consumption – by too much.

Secondly, not all customer information is created equal. We found that knowing how many friends each customer has is far more valuable than knowing their propensity to show off on social media. Both types of information are better than no information, but conspicuity information added no value for firms that already knew the extent of their customers’ social network connections. 

In the aggregate, conspicuity information’s value-add amounted to between 2% and 6% in profit gains, while connectedness information brought profit gains ranging from five to 30 percent.

What customers want

The difference in value between the two kinds of information highlights another general principle of social media information gathering. Information is always less valuable when it describes aspects of customer behaviour that directly align with company objectives. 

In this case, the customers’ own actions can compensate for the missing knowledge – the most prolific show-offs aren’t shy about showing you who they are. They would happily opt into a selective selling program, assuming the marketing campaign were strong in most other respects. In fact, the company can then design appropriate schemes that have customers self-select into the relevant segments.

Privacy concerns

Employing a discerning information-gathering process, rather than a data dump, helps firms avoid running afoul of privacy concerns, which are sure to continue increasing as e-commerce and social media become more intertwined. The advent of WeChat heralds the transformation of social networks into digital High Streets, where products can be found and flaunted in an unbroken clickstream. Going forward, brands will need to cultivate a neighbourly intimacy and trust with their customer base. That means e-commerce will become more reliant on tailored tactics like selective selling, and less on off-the-rack approaches.

(Karan Girotra is an Associate Professor of Technology and Operations Management at INSEAD, the Paul Dubrule Chaired Professor of Sustainable Development and co-author of The Risk Driven Business Model: Four Questions That Will Define Your Company. You can follow him on Twitter @Girotrak. Read more at http://knowledge.insead.edu/marketing/welcome-to-the-social-media-shopping-mall-4839#2ERgLL5gWgQbrJKB.99)