Using online referrals? Shoppers look for a fair share

What works best when you’re constructing an online referral program to attract new customers? One thing to watch is fairness, says Professor of Information Systems Yili (Kevin) Hong. According to him, equity is of particular importance when constructing online referral systems that give customers a chance to make referrals to folks they don’t know well.

The largest cab company in the world operates in more than 60 countries, serves people in some 300 cities, and owns a small fleet that is measured in mere dozens. It’s Uber.

Likewise, Airbnb now is valued at some $30 billion and has more than a million lodging options. It rivals the world’s largest hotel chains. Like Uber, it began as a software platform, not an asset owner, and it hasn’t even celebrated a 10-year anniversary yet. Here’s another similarity between these two large money makers: Both leveraged referral marketing to spur their extraordinary growth.


Passing it on

Referrals certainly pre-date online marketing. Who hasn’t shared the name of some restaurant they like or taken a car to that honest mechanic a neighbor uses? Professor of Information Systems Yili (Kevin) Hong’s research, however, looks at the world of social commerce, not referrals tossed over the back fence or shared with a dinner companion. 

“Your friend can affect your decision to buy something,” he says, adding that before he began his research some six years ago, online shopping was still “a pretty isolated event for individuals. You just sat in front of your computer, went to or, and made your purchase.”

Now, with systems that let you see what others like you have bought, share purchases via social networks, and more, there are plenty of ways to get a little friendly advice before you click the “buy” button. It’s called social commerce, and Wikipedia defines it as a subset of e-commerce that involves social media, online interaction, or user contributions that support the buying and selling process.

Hong’s research examined the simple referral part of social commerce, where one person offers a referral and another can accept or reject it. Such references are powerful, as Uber and Airbnb can attest.

A 2016 Nielsen study of U.S. consumers found that 67 percent of survey respondents were more likely to buy a product after a friend or family member shared information about it via email or social media.  What’s more, 88 percent said they’d like some incentive — such as money, loyalty points, gifts, or discounts — for sharing product or service information with email pals or social media contacts. For 77 percent, the reward of choice was money. Only 18 percent preferred gifts.

Unlike the referrals one makes in person to friends and family, we can conceivably send online references to distant acquaintances or even people who are practically strangers. That’s because many referral systems allow users to select recipients from Gmail, Yahoo, as well as other email and social media platforms.

Hong points to a recent survey that showed the average Facebook user has 145 Facebook friends, but only 28 of them are considered close friends, real friends. The internet, he notes, encourages acquaintances, and these can quickly get swept into an online referral system.

Consequently, Hong looked at reference results between people and their family or close friends, as well as those references that get sent to little-known acquaintances. He also looked at the interplay of incentives within those two types of relationships — friends versus acquaintances. As it turns out, both social distance and offer fairness make a difference in referral success.


Game on

People have multiple reasons for sharing product or service referrals. There may be an intrinsic motivation, something done because it feels good, like helping a friend fill a need or find a great deal. And, there are extrinsic motivations or external rewards, such as money.

To design his study and interpret results, Hong likened the referrals to ultimatum game theory, which is a type of economic experiment. In it, one player, called the proposer, is given a sum of money and allowed to share it with another player, who is known as the responder. The proponent must make a take-it-or-leave-it offer, and the responder gets to accept or reject the deal. Then, the game ends. There is no round two.

Imagine you’re the proposer in the game and you’re playing with $10. “You can keep $9 for yourself and offer $1 to your friend, or you can keep $7 and offer up $3, or you can divide it equally. You decide,” Hong explains.

“One prediction that is coming out of this game is that economically or rationally, as long as the money is divided in a way that the responder gets something, he has to accept it,” Hong continues. “Even if you only offer the other person one cent out of the $10, the other person will take it because that’s the optimum solution for him. If he doesn’t accept the offer, he gets nothing.” Presumably, getting nothing is not in the responder’s best interest. Something is better than nothing, right?


A friend in greed

As it turns out, though, that’s not how things work. “People’s behavior deviates from their best interest. They depart from the optimal decision,” Hong says. Studies using this oft-researched game indicate that responders probably won’t go below a 20- or 30-percent cut of the cash. “They won’t take it because of the idea of fairness,” Hong says, and that’s a standard interpretation. Many researchers see the ultimatum game as proof that people spurn injustice. They consider it an affront.

Hong considered this idea of fairness in designing six lab experiments and one real-shopper field study. In the lab experiments, we told study subjects there were varying degrees of friendship among the recipients of their referrals. There were also multiple ways the referral bonus was split up between proposer and responder.

Also, the scholars used two scales for determining how well referral proposers and responders were acquainted. One is a social distance scale that defines categories of relationships, such as friends, close relatives, co-workers, or acquaintances. The other scale used defines social distance by how often people interact, which is the strength of the bond between people.

All of the experiments examined what impact social distance and fairness have on people’s willingness to make or accept a referral. Results were consistent throughout the experiments, and they showed:

  • People are more likely to send referrals to those they know well. Folks are more apt to accept such referrals, as well.
  • Fairness isn’t much of an issue between close friends and family. Proposers may still send these folks a reference with an unfair incentive split, and responders may accept them.
  • Among those who don’t have strong social ties, fairness does impact referral success. If the referral incentive doesn’t seem fair, acquaintances are less likely to send a reference or accept one they receive.

Since fairness isn’t much of an issue between close friends and family, Hong says the wording of referral offers could have a significant impact. He’s currently conducting research to see if making reference language somehow altruistic — i.e. “do your friend a favor” or “share the love” — will increase positive responses when referrals occur among people with close ties.

Ultimately, Hong says, companies should think about who might be receiving the referrals their customers make. Is it the customers’ inner circles only or their entire online communities? Those distinctions matter in the design of referral systems and associated incentives.