01The story in three questions
One recommendation. Two sides of the same economy.
Imagine someone recommending a service they genuinely use. A friend decides whether to join. The business gains a potential customer. The person who made the introduction might receive a reward. That exchange raises three much bigger questions: does the customer stay, can the model scale and how much value reaches the individual?
01 / CUSTOMER QUALITY
≥0%
Higher customer value in one German bank study, compared with matched non-referred customers.[1]
02 / COMMERCIAL REACH
>0%
Of Revolut's new retail customers in 2025 attributed to word of mouth and referrals combined.[6]
03 / PERSONAL INCOME
0%
Of current or recent U.S. gig-platform workers say their earnings are essential or important for basic needs. This is broader side-income context, not referral-specific.[11]
The argument in one minute
First, test whether a recommendation produces a valuable customer. Then ask what happens when recommendations become a serious acquisition channel. Finally, follow the reward back to the person who shared it.
The figures above describe different populations and outcomes. The chapters that follow explain what each one really means, rather than combining them into a misleading global average.
02The empirical evidence
The first test: does a referral produce a better customer?
A signup is only the beginning. The economic question is what happens afterwards. In one influential study, researchers followed roughly 10,000 customers of a German bank for almost three years. Referred customers had at least 16% higher value than comparable non-referred customers. That is an important clue about customer quality, but one bank cannot tell us what every referral program will achieve.[1]
The referred-customer value advantage
Normalized comparison in the German bank study
More invitations are not the goal. More valuable customers are.
A measurable growth engine
How an invitation is presented can change what people do. A randomized Facebook experiment studied adoption across 9,687 users and 1.4 million friends.[3] Airbnb later reported that redesigning its referral experience increased signups and bookings generated by referrals per day by more than 300%.[5] The two findings measure different things, but both show that program design matters.
The hidden economic trade-off
Incentives are not free. Banking and telecom field evidence found that larger rewards attracted more people but reduced the profitability of new customers.[2] On freemium platforms, referral activity can also displace paid upgrades.[4] Both effects matter when deciding what a successful program actually earns.
The metric that matters
Incremental customer lifetime value − incremental rewards, operating costs, fraud and cannibalization
Only the value that would not have appeared otherwise can justify the program's true cost.
So the first finding is about quality, not volume. The next question is what happens when recommendations move from experiments into the growth strategy of a business serving millions.
03Commercial importance
When a recommendation becomes part of a growth strategy.
A bank study tells us something about customer quality. A global financial platform can tell us how prominent recommendations may become at scale. Revolut offers one such example. The affiliate economy offers another, related form of performance-based distribution. The distinction matters: a friend recommending an app and a publisher earning commission are not the same acquisition channel.
Revolut · 2025
63%+
Of new retail customers reportedly attributed to word of mouth and referrals combined.
Revolut reported adding 16 million retail customers in 2025. Its acquisition percentage covers organic word of mouth as well as referrals, so it cannot be read as a paid referral-program conversion rate.[6]
impact.com · multinational survey
74%
Of surveyed brands reported that affiliates contributed 11–30% of their company revenue.
The 2025 impact.com survey included 818 marketers across eight markets. This is affiliate marketing evidence rather than a global count of friends referring friends.[7]
Organic WOM
A voluntary, trusted recommendation; often invisible to attribution.
Customer referrals
Trackable invitation or code; a reward may go to either or both people.
Affiliate & creator
Performance-based distribution by publishers and creators; distinct from friends inviting friends.
How does this look outside a research paper?
See real referral links and rewards being shared in the ReferTree community.
04Worldwide perspective
The referral economy is not developing at the same pace everywhere.
The global picture becomes much clearer once we separate two kinds of progress. In the United States and United Kingdom, industry records show how much brands are investing in affiliate and partner marketing. In India, Brazil and African markets, official statistics reveal how quickly the digital systems used to transact and receive rewards are expanding. Together, they point towards a wider opportunity, but they do not measure the same activity.
United Kingdom · affiliate spend
+7.3%
Year-on-year increase in 2025 based on network and platform data.[22]
Developing economies · digital payments
+7 pp
Adults making digital merchant payments, from 35% in 2021 to 42% in 2024.[15]
Measured channel growth
Brands are putting more money into performance partnerships.
The strongest cross-year evidence comes from national affiliate industry measurements. These figures describe affiliate and partner marketing, not ordinary customer-to-customer referral bonuses.
United States
USD billionsPMA reports 49.8% growth across three years using eight affiliate networks and more than 50 publishers.[21]
United Kingdom
GBP billionsAPMA reports 7.3% annual growth in 2025 and 357 million affiliate-attributed transactions.[22]
The infrastructure story
Elsewhere, the rails for participation are reaching more people.
A referral reward is easier to receive and use when digital payments are routine. That makes financial access an important part of the opportunity, even though payment adoption alone says nothing about how many referrals have taken place. Three regions illustrate why this matters.
India
Sep 202624.07bn
Monthly UPI transactions
The national payments operator reported 24.07 billion transactions in September 2026. Everyday digital transfers now occur at immense scale. This measures payment activity, not referral payouts.[23]
Brazil
Sep 20267.8bn+
Monthly Pix transactions
Brazil's central bank recorded more than 7.8 billion Pix transfers in September 2026, with over 150 million individual users that month. These are payment-system figures, not referral volumes.[24]
Sub-Saharan Africa
2021 to 202449% → 58%
Adults with a financial account
World Bank Findex shows a nine-percentage-point rise in account ownership. The widening ability to transact is relevant to potential reward distribution, not a measure of referral adoption.[15]
Nigeria
2023 to 2024+15.5%
Electronic payment transactions
NIBSS reported that electronic transactions rose from approximately 9.7 billion in 2023 to 11.2 billion in 2024. Different payment channels are included, and the increase does not identify referral transfers.[25]
Worldwide financial access
The foundation is getting broader.
Across low- and middle-income economies, the share of adults making a digital merchant payment rose from 35% in 2021 to 42% in 2024. The nationally representative Findex study interviewed about 148,000 adults across 141 economies. The result helps establish the size of the opportunity without mislabeling digital payments as referrals.[15]
2021
35%
2024
42%
Where affiliate spending is measurable, it is growing. Where digital finance is expanding, the conditions for wider participation are improving. Those are different findings, and both matter.
Our interpretation: Mature digital markets offer evidence of increasing commercial investment in measurable partnerships. Fast-growing payment ecosystems could make it easier for more people to earn or redeem small rewards, especially where digital account access is still expanding. That second link is a hypothesis about future opportunity, not an established increase in referral income.
Affiliate spending, consumer referrals and underlying payment infrastructure are distinct metrics. Country differences in reporting years, denominators, payment systems and currencies prevent a valid global referral-growth ranking. Google Trends can supplement this analysis with search-interest context, but normalized search scores should not substitute for transaction evidence.[14]
The infrastructure may be global, but the transaction is personal. Which brings us back to the person sharing the link and the value that actually reaches them.
05The human side
What does the person making the recommendation actually earn?
Follow the same referral back to the person who shared it. The reward might be cash, a discount or credit for another purchase. The company sees an acquisition cost. The individual sees a financial benefit, although not always one they can spend freely. This is where our knowledge becomes strikingly thin: there is no representative worldwide estimate of income derived specifically from referral rewards.
The same reward is a cost in a company's accounts and a potential benefit in someone's budget.
Why supplementary income matters
71%
Income that helps cover essentials
In a Pew Research Center survey, 71% of lower-income U.S. gig-platform workers said their gig earnings were at least important for basic needs. That illustrates the potential significance of additional income, but does not measure referral rewards.[11]
The unanswered question
What proportion of a person's income comes from referrals?
Most statistics cannot isolate it. Labor research measures digital-platform work overall.[9][10] Creator surveys mix affiliate commissions with sponsorships, ads and other earnings.[12][13] Neither tells us how much an ordinary person earns from sharing referral codes.
The distinction is practical: a €20 discount may reduce spending, but it is not €20 of cash income.
Until that earnings data exists, the defensible approach is to show the arithmetic without pretending it describes a typical user. The interactive example below illustrates how rewards, qualifying referrals and other income fit together.
An illustrative calculation
What could a referral reward contribute?
Change the assumptions to see a mathematical example, not a prediction of what users earn.
€100
Hypothetical monthly gross rewards
4.8%
Of combined hypothetical gross income
Assumes all referrals qualify and are paid. Excludes taxes, expenses, refunds, limits, expired offers and zero conversions. No representative worldwide referral-income distribution is available.
There is value in making referrals easier to manage.
Bring your existing links and codes together in a shareable profile, whether you are just starting or already recommending services.
06The economic backdrop
Why this exchange looks different when money is tight.
The household and the company respond to the same economic climate in different ways. A modest bonus can become more attractive when everyday costs rise. For a business, paying for a verified referral may look less risky than paying upfront for attention. But the logic cuts both ways: consumers under pressure may be more interested in a reward and less able to complete the purchase that unlocks it.
A modest-growth backdrop, not a referral forecast
2026 world GDP projections from three institutions provide economic context for this two-sided market.
For the person sharing
A €20 reward might feel more worthwhile under cost-of-living pressure. Yet it only becomes an actual benefit if the recipient meets the offer's conditions, which may themselves become harder to satisfy.
For the company acquiring
Paying only after a qualifying action can help align spending with outcomes. But weak demand, lower retention or overly generous rewards can erase that advantage.
These are economically motivated hypotheses, not statistically estimated effects of inflation, unemployment or interest rates on referral demand. A global causal panel has not yet been established.
07Looking ahead
The next decade depends on whether trust survives scale.
Economic pressure may draw more people towards rewards, but that will not sustain the channel if recommendations stop feeling authentic. AI could reduce the effort of finding relevant offers, while creator commerce could expand distribution. The opposite future is possible too: more spam, larger incentives and less attention. The interactive scenarios below explore those conditions, not a predicted market size.
Illustrative strategic-importance index
Three futures, not one forecast.
2026 = 100 · through 2036
≈148
Index in 2036
A 4% annual increase in relative strategic importance as digital adoption and performance-based partnerships expand. Values are scenario assumptions, not measured industry forecasts or a market-size CAGR.
Figure 4. Original scenario model presented in the underlying study. The index starts at 100 in 2026 and is not a prediction of market size, earnings or a realized revenue growth rate.
AI-enabled discovery
Less friction for matching offers and audiences, alongside new verification challenges.
Creator convergence
Content, social influence and measurable affiliate results become more interconnected.[7]
Trust as the constraint
Disclosure, genuine recommendations and fraud prevention become differentiators.
08What comes next
What would make this market genuinely worth trusting?
The studies point to a conclusion that is easy to miss when we focus on headline numbers. More links do not automatically mean more growth, and bigger bonuses do not guarantee better outcomes. Good referral systems have to work for the company, the person sharing and the recipient. These are the design principles that follow from that three-sided reality.
- 01
Measure incremental value
Run proper treatment/control experiments. Account for retention, reward cost, reversals and cannibalization.[2]
- 02
Reward meaningful outcomes
Payment after qualified activation or retained revenue can protect unit economics better than paying for invites.[5]
- 03
Design for both people
What motivates a referrer may deter a recipient; a large 2026 experiment shows incentive design has asymmetric effects.[20]
- 04
Make incentives transparent
Show who benefits from a recommendation, eligibility rules, actual payout status and any restrictions.[9]
- 05
Preserve the social context
Private, relevant recommendations deserve priority over indiscriminate broadcasting and synthetic endorsements.[5]
The opportunity is real, but only partly measured. We know referrals can influence customer value and acquisition. We know rewards can benefit individuals. Understanding both sides with the same rigor is the work that comes next.
09Transparency & reproducibility
How we kept a compelling story from outrunning the evidence.
This article brings together experiments, customer cohorts, company statements, industry surveys and economic projections. They answer different questions and often measure different populations. We therefore keep their results distinct rather than manufacturing a single worldwide referral-growth figure from incomparable percentages.
Strongest evidence
Peer-reviewed field experiments and customer cohorts on incentives, diffusion, customer value and profitability.
Industry indicators
Company disclosures and industry surveys are useful examples, but often self-reported, selective or non-representative.
Cross-market comparability
PMA and APMA report affiliate investment, while NPCI, Brazil's central bank, NIBSS and Findex measure digital financial infrastructure. None alone establishes national referral growth.
Search interest versus adoption
Google Trends scores are relative search interest, not volumes of referral transactions. Without program conversion and payout data, no global rankings are published.
Income measurement gap
Surveys rarely isolate referral bonuses from creator sponsorships, affiliate revenue and other platform income.
Scenario disclosure
2036 values reflect explicit assumptions, not a fitted econometric forecast or a third-party market-size estimate.
Recommended next dataset
The next step is to connect verified referral activations and rewards with the indicators already available. Monthly partner marketing statistics, World Bank Findex and official real-time payment data illuminate different parts of the system, while a new representative household survey could measure referral-specific gross and net income. Proper cross-country research would then control for digital access, inflation, wages and local program incentives.
Separate customer refer-a-friend payments from affiliate commissions, creator sponsorships and non-cash discounts. Use country/time fixed effects and controlled program experiments; report uncertainty and contradictory findings.
Research bibliography
Sources & further reading
Select a citation to open the original publication. Sources are labeled by evidence type so that field experiments, official forecasts, commercial surveys and archived search snapshots are not conflated.
- [1]
Referral Programs and Customer Value
Schmitt, Skiera & Van den Bulte · Journal of Marketing · 2011
Peer-reviewed / customer cohortOriginal source - [2]
Referral Reward Size and New Customer Profitability
Wolters, Schulze & Gedenk · Marketing Science · 2020
Field experimentOriginal source - [3]
Creating Social Contagion Through Viral Product Design
Aral & Walker · Management Science · 2012
Randomized experimentOriginal source - [4]
Social Referral Programs for Freemium Platforms
Belo et al. · Management Science · 2022
Field experimentOriginal source - [5]
Making Referrals Work for Airbnb
Airbnb Engineering · 2014
First-party engineering caseOriginal source - [6]
Revolut 2025 Full-Year Results
Revolut · company announcement · 2026
Company-reported resultsOriginal source - [7]
The State of Affiliate Marketing
impact.com · 818 marketers, 284 publishers, 421 creators in eight countries · 2025
Industry surveyOriginal source - [8]
U.S. Affiliate Marketing Channel Study
Performance Marketing Association, summary via Awin · 2025
Industry measurementOriginal source - [9]
Realizing Decent Work in the Platform Economy
International Labour Organization · 2024
International labor synthesisOriginal source - [10]
Handbook on Measuring Digital Platform Employment and Work
OECD, ILO & European Union · 2023
Official statistical methodologyOriginal source - [11]
How Gig Platform Workers View Their Jobs
Pew Research Center · United States · 2021
SurveyOriginal source - [12]
- [13]
- [14]
Introducing the Google Trends API (alpha)
Google Search Central · 2025
Official documentationOriginal source - [15]
Global Findex Database 2025: Financial Inclusion and Digital Payments
World Bank · representative surveys of about 148,000 adults in 141 economies · 2025 (2024 data)
Official international household surveyOriginal source - [16]
Expansion of the Gig and Platform Economy in India
International Labour Organization · 2026
Country report / digital workOriginal source - [17]
World Economic Outlook, October 2026
International Monetary Fund · 2026
Macroeconomic forecastOriginal source - [18]
- [19]
OECD Economic Outlook, Interim Report September 2026
OECD · 2026
Macroeconomic forecastOriginal source - [20]
Can Reward Uncertainty Encourage Social Referrals?
Zhang et al. · Management Science · 2026
Field experimentOriginal source - [21]
PMA Performance Marketing Industry Study 2025
Performance Marketing Association · data from eight U.S. affiliate networks and 50+ publishers · 2025 (2021–2024)
Industry transaction and spend measurementOriginal source - [22]
State of the Affiliate Nation 2026
Affiliate & Partner Marketing Association · UK platform and network submissions · 2026 (2025 data)
Industry transaction and spend measurementOriginal source - [23]
UPI Product Statistics
National Payments Corporation of India (NPCI) · September 2026
Official monthly payment-system statisticsOriginal source - [24]
Pix em números
Banco Central do Brasil · September 2026
Official monthly payment-system statisticsOriginal source - [25]
Nigeria's Payment Revolution
Nigeria Inter-Bank Settlement System (NIBSS) · 2026 (2023–2024 data)
Official payment-system operator statisticsOriginal source
Built around better referrals
Your referrals deserve a better home.
Organize, share and discover referrals in one place. ReferTree helps turn scattered links and codes into something useful for you and the people you recommend them to.
Create your free ReferTreeResearch synthesis published 25 October 2026 · ReferTree Research · Global perspective, with clearly labeled evidence limitations.