Customers acquired through referrals made more subsequent referrals than customers acquired through other channels in several company data sets, research in the Journal of Marketing Research found.
Rachel Gershon and Zhenling Jiang studied the downstream effects of referral programs across businesses in finance, software and retail. An American Marketing Association account of the study says referred customers made 31% to 57% more referrals in the data sets examined.
Those comparisons do not mean a referral program will produce the same increase for every firm. People who arrive through friends can differ from other customers in ways that also affect their future behavior.
Counting the next introduction
A typical referral program encourages an existing customer to introduce someone else, sometimes with a reward for one or both people. A business may count the new customer’s purchases when evaluating the program.
The researchers argue that the new customer’s later introductions can also matter. Their analysis estimated that omitting those downstream referrals understated the value of a referral by 20% to 36% in the settings studied.
Their paper, published in the Journal of Marketing Research, combines company data with experiments. Controlled tests examined whether having been referred made a later recommendation feel more socially appropriate.
A reminder tested in the field
In a field experiment involving more than 10 million referred customers, a message that reminded people they had joined through a referral produced more referrals than a generic invitation, the marketing association reports. Its account describes an increase of more than 20% in that test.
The finding is specific to the participating business, message and audience. The research does not establish that repeated prompts are welcome, that all referred customers remain profitable, or that a referral causes a sale that would not otherwise have occurred.
Measure the full cost and customer response
A company can follow referrals through to qualifying purchases and account for rewards, discounts, fulfillment, fraud prevention and customer support. It should also compare similar customers or use a randomized test when estimating what the program added.
Our earlier coverage of advertising experiments and additional sales explains why an attributed transaction is not automatically an incremental transaction. Clear terms and appropriate disclosures also matter when customers receive incentives to recommend a product.