Referral Program Research: Lessons for Service Businesses

Referral Program Research: What 10,000 Bank Customers Proved About Referred Leads

Referrals are the channel every service business says it relies on and almost none actually operates. The excuse is that referrals cannot be engineered. The evidence says otherwise. In 2011, the Journal of Marketing published a study by Schmitt, Skiera, and Van den Bulte that tracked roughly 10,000 customers of a German bank, comparing those acquired through a paid referral program against everyone else for almost three years. Referred customers were worth at least 16% more. This article unpacks what the study actually found, why the mechanism works, and how a service business turns peer-reviewed evidence into a working referral system.

By Joshua Agonya Pi'Rwot

Founder, Business Growth Accelerator

Section 1

The Study: A Referral Program Under a Microscope

Most referral advice rests on intuition. This study rests on data. Philipp Schmitt, Bernd Skiera, and Christophe Van den Bulte analyzed a leading German bank's referral program, which paid existing customers 25 euros for each new customer they brought in. The researchers tracked roughly 10,000 customers, referred and non-referred, for almost three years, controlling for demographics and acquisition timing. Published in the Journal of Marketing in 2011, the findings were unambiguous: referred customers had higher contribution margins, although that gap eroded over time; they had higher retention rates, and that gap persisted; and their customer lifetime value was at least 16% higher than comparable customers acquired through other channels. The value premium comfortably exceeded the modest reward the bank paid. This is the strongest public evidence that referred leads are not just cheaper, they are structurally better. For the step that usually comes next, see [Startup Success: How AI Automation Transformed Our Business](/blog/startup-success-how-ai-automation-transformed-our-business).

Section 2

Why Referred Customers Are Worth More

A follow-up study by Van den Bulte, Bayer, Skiera, and Schmitt in the Journal of Marketing Research explains the mechanism with two forces. Better matching: existing customers understand both your service and their friend's situation, so they refer people who genuinely fit, doing free qualification for you. Social enrichment: the referred customer's relationship with the referrer makes the new relationship stickier, which is why the retention advantage persists even after the margin advantage fades. McKinsey's separate word-of-mouth research adds scale: it estimates word of mouth is the primary factor behind 20 to 50 percent of purchasing decisions. For a service business, the implications are mechanical, not motivational, and the table below translates each documented finding into an operating decision.

Section 3

From Research to System: Engineering the Referral Ask

The bank in the study did not hope for referrals; it ran a program with an offer, a process, and accounting. That is the part most service businesses skip. They deliver good work, assume referrals will follow, and call the resulting trickle a channel. A system looks different. Define the moments of peak goodwill, results delivered, a milestone hit, an unprompted compliment, and trigger a specific ask at each. Tell clients exactly who you are looking for, because the matching mechanism only works when the referrer can recognize a fit. Decide your incentive deliberately; the research validates rewarding referrals, and in professional contexts reciprocity, recognition, or donated value can replace cash. Then track referral leads as their own source with their own close rate. Inside LeadOS, we build this as a quarterly rhythm, not a poster on the wall. A useful companion to this piece is [The Marketing-to-Sales Handoff: Where Service Businesses Lose Their Best Leads](/blog/marketing-to-sales-handoff).

Section 4

What the Research Does Not Say

Intellectual honesty cuts both ways, so note the limits. This was one bank, in one country, in retail banking; your effect size will differ, and the study itself shows the margin advantage decays, so referrals are not a permanent value subsidy. The research also cannot tell you the right reward for a consultancy or a clinic, only that a modest, clearly defined incentive produced customers whose extra value dwarfed its cost. And a referral program amplifies what exists: if your service disappoints, the same social mechanics spread that instead. The practical move is to borrow the direction, referred leads are better leads, run the experiment for a quarter, and read your own numbers. If you want a second set of eyes on designing that experiment, a strategy call with BGA is the low-stakes way to start. If you are turning this into practice, [Common Myths About AI Automation Debunked](/blog/common-myths-about-ai-automation-debunked) maps the adjacent system.

Direct answers for operators.

What did the Journal of Marketing referral study actually find?

Tracking roughly 10,000 customers of a German bank for almost three years, the researchers found referred customers had higher contribution margins initially, higher retention that persisted over time, and customer lifetime value at least 16% higher than comparable non-referred customers. The bank paid 25 euros per successful referral, far less than the documented extra value those customers delivered.

Should a professional services firm pay cash for referrals like the bank did?

Not necessarily. The research validates the principle that a defined incentive plus a defined process produces better customers; it does not mandate cash. Many professional firms use reciprocal referrals, public recognition, gifted services, or charitable donations instead, especially where regulations or client perceptions make payments awkward. What matters most is a specific ask, an easy mechanism, and consistent tracking of referral leads as a distinct source.

How do I start a referral system without feeling pushy?

Anchor the ask to moments of demonstrated value: a delivered result, a milestone, an unprompted compliment. Tell clients precisely who you help best so they can pattern-match, which the research shows is the engine of referral quality. Make the action one step, a forwardable note or intro link. Asked this way, a referral request reads as confidence in your work, not desperation for pipeline.