Case Study: Nubank and the Economics of a Brand People Recommend

By Yamm Labs · Updated September 2026 · 11 min read

The growth review at a mid-sized NBFC in Gurugram starts the way it always does. The performance agency puts up a slide of cost per disbursed loan, the number has crept up again, and someone asks for a bigger budget on the same three apps. Then a board member holds up a phone showing a purple card from Brazil and asks a different question: how did a lender that barely advertised get most of its customers to bring in the next ones? That question deserves a careful Nubank case study, because the economics behind it are in public filings, and most of what gets repeated about it is not.

What was Brazilian banking like when Nubank started?

Nubank describes its journey as beginning in 2013 in Brazil. Its first product, launched in 2014, was the Nu Credit Card, a purple Mastercard-branded credit card, and the company says it was a pioneer in offering credit cards without any annual fee. Those details come from the prospectus Nu Holdings filed with the US Securities and Exchange Commission before its 2021 listing, which is the most useful primary source on how the company thinks about growth.

The same prospectus describes a concentrated market. It states that the five largest banks in each of Brazil, Mexico and Colombia control between 70% and 85% of all loans, deposits and overall banking revenue. It also includes a founder’s letter describing four months of queues, call-centre calls and repeat branch visits with more and more documents, all to open an account that would charge hundreds of reais a year in fees.

That is the situation in one line: a small number of large incumbents, high fees, and a customer experience people remembered for the wrong reasons. Nubank is regulated in Brazil by the Central Bank of Brazil, and it began with a credit card rather than a full bank account.

Why could a new card issuer not simply buy its way to scale?

A challenger in a concentrated market faces a cost problem before it faces a product problem. The incumbents own the branch networks, the salary accounts and the default mental slot for “bank”. Buying attention against them through paid media is expensive, and every rupee or real spent on acquisition has to be earned back from a customer who may leave after the first offer.

There is also a trust problem. A new financial brand asks people to hand over identity documents and, eventually, their salary. Advertising can create awareness, but it rarely creates the kind of trust that makes someone move their primary account. A recommendation from a friend who already uses the product does that work faster.

So the complication for Nubank was a narrow one. It needed millions of customers in a market where the big banks controlled most of the revenue, and it could not rely on outspending them. The strategic answer was to make the customer the acquisition channel, which only works if customers have something specific and positive to say.

SIGNALNewsletter by Yamm Labs

The next piece, straight to your inbox.

New writing on BFSI brands, B2B and influencer marketing, and AI search.

Email *

Subscribe

Subscription form is not available at the moment

One click to unsubscribe. Privacy policy

What did Nubank decide to do instead of outspending the big banks?

Read against its own filings, Nubank’s decisions fall into three groups. None of them is a campaign. All of them are design and service choices that give customers a reason to talk.

• Remove the fee people resented. The first card had no annual fee, in a market where the prospectus describes bank accounts charging hundreds of reais a year. A fee that disappears is easy to repeat to a friend in one sentence.
• Remove the queue. The founder’s letter frames the problem as months of branch visits and paperwork. In the same prospectus, the company estimated that it had saved customers over 113 million hours of waiting time in branches or on hold with call centres, and over US$4.8 billion in banking fees. Those are the company’s own estimates, not audited results, but they show what it chose to measure.
• Make service the positioning. The prospectus reports a Net Promoter Score of 90 in Brazil and 94 in Mexico, and says the company achieved the fewest customer complaints when compared with incumbents and disruptors on the latest publicly available information. Whatever one thinks of NPS as a metric, a company that puts it in a listing document is telling investors that service is the growth engine.
• Make the product visible. The card was purple. Every time a customer paid, the card itself started the conversation.

The common thread is that each decision produces a sentence a customer can say out loud: no annual fee, no branch, someone actually answered, and the purple card. Referral-led growth is what happens when those sentences exist. It cannot be bolted on with a referral bonus alone.

Why does owning one colour matter so much for a financial brand?

Most Brazilian and Indian banks sit in the same narrow band of blues, reds and oranges. A new brand in that band has to work much harder to be remembered, because customers cannot tell it apart at a glance. Nubank took a colour no incumbent owned and put it on the one object customers carry and show to others.

The colour stuck. Pentagram, which worked on the Nubank identity, notes on its case page that the cards are nicknamed “roxinho”, and that the refreshed identity kept the signature purple while expanding the wider palette with colours drawn from art, fashion and culture. It also describes a simplified mark, “Nu”, which means naked in Portuguese.

Nubank’s own note on its new logo, published on 17 May 2021, says the curves of “nu” were made smoother and less square, the purple shade was adjusted, and the letters were given more colour weight to make reading more accessible. At that point the company said the logo was part of the daily lives of more than 38 million people.

The lesson is not “use purple”. It is that a colour becomes an asset only when it is used on the same thing, the same way, for years, and only changed with care. Nubank’s refresh adjusted the shade and softened the mark. It did not walk away from the colour customers had already named.

Nubank case study: what results has the company publicly reported?

Only a few outcomes are stated in primary sources, and it is worth keeping them separate from the many figures repeated in blogs.

The latest figures come from the second-quarter 2026 results release Nu Holdings filed with the SEC. That release reports customer numbers and activity but does not restate the organic acquisition share, the acquisition cost or the NPS. So the fair reading is this: the referral economics are documented for the period up to 2021, and the scale since then is documented, but whether the same referral share still holds today is not publicly reported in the sources used here.

What the numbers do show is the shape of the model. A company that acquires most customers without paid media can spend its budget on product, service and credit, and a low acquisition cost gives it room to be patient with customers who start small.

What would an Indian NBFC have to change before copying Nubank’s referral model?

This is where most “learn from Nubank” advice stops, and where the real design work starts. An Indian NBFC runs its brand under RBI rules, and those rules decide what the share moment can say and in what order the customer sees things.

Take the most common copy of the Nubank idea: a “zero processing fee” personal loan, with a referral card the customer can forward on WhatsApp. Three regulator-driven design decisions follow.

• The fee claim cannot be the whole story. RBI’s key facts statement notification of April 2024 covers NBFCs and applies to retail and MSME term loans sanctioned on or after 1 October 2024. It requires a key facts statement with a computation of the annual percentage rate, which includes all charges the lender levies. So the design decision is to carry the all-in rate language on the same screen the “zero fee” headline leads to, not three taps later.
• The screen sequence is fixed before the design starts. Under the same notification, fees not mentioned in the key facts statement cannot be charged later without explicit consent, the statement must be in a language the borrower understands, and it must stay valid for at least three working days for loans of seven days or more. In practice the journey goes referral link, eligibility, key facts statement in the customer’s language, then agreement. The referral reward screen cannot jump ahead of the key facts statement.
• The lender’s name goes on the share card. RBI’s digital lending rules expect a borrower who applies through an app or partner to be able to see which regulated lender is making the loan offer. So the forwardable card needs a lockup rule: the NBFC’s registered name sits inside the card artwork, not only in the link preview, because the preview is often cropped when the card is forwarded.

None of these rules stops referral-led growth. They change where the words sit. Our BFSI compliance-creative cheat sheet collects the placement decisions like these that we check before any creative goes to legal review.

What can a mid-sized Indian NBFC copy from Nubank today?

A mid-sized NBFC does not have Nubank’s scale, and does not need it to use the same logic. Five moves transfer well.

• Pick one pain customers already complain about. Nubank picked the annual fee and the branch queue. For an Indian NBFC it might be disbursal time, foreclosure charges or a support line nobody answers. Fix it before running any referral scheme, because the scheme only amplifies what customers already feel.
• Write the sentence you want customers to say. If a happy borrower cannot describe the difference in one plain sentence, there is nothing to refer. Test the sentence with existing customers before designing the campaign.
• Own one visual signal and keep it. In a category of blues and reds, a distinct colour on the card, the app icon and the disbursal message becomes the thing people recognise when a friend forwards it. Our history of bank rebranding in India shows how often Indian lenders spend a colour’s equity by changing it before customers have learnt it.
• Treat service as the positioning, and measure it. Nubank put NPS and complaint rankings in a listing document. An NBFC can track complaint volumes and resolution time by product and publish the parts it is proud of.
• Design the share moment as a regulated asset. The referral card, the landing screen and the reward message are advertising. Build them once, with the lender name, rate language and screen order fixed, and let the product team reuse them.

What should an Indian NBFC not copy from Nubank?

Three things do not transfer. First, the scale numbers. A company with 139 million customers across three countries has network effects a regional NBFC will not have, so the referral share it reports is a direction, not a target.

Second, the product entry point. Nubank started with a credit card in Brazil. An Indian NBFC’s entry product is more likely a personal, gold, vehicle or MSME loan, each with its own disclosure rules and customer anxieties. The service promise must match that product, not a card.

Third, the rebrand timing. Nubank adjusted its mark after years of use and tens of millions of customers had learnt it. A young NBFC that keeps changing its colour and logo while trying to build word of mouth is resetting its own memory structures every year. Settle the identity early, then protect it.

A referral programme cannot create something to say; it can only carry what the product already makes people say.

People also ask

How much of Nubank’s growth came from referrals?

Nubank’s 2021 US prospectus says it acquired approximately 80% to 90% of its customers organically on average per year since inception, through word of mouth or unpaid referrals from existing customers. Its second-quarter 2026 results release does not restate that share, so a current figure is not publicly reported in those sources.

Why is Nubank’s card purple?

Nubank’s prospectus describes its first product as a purple Mastercard-branded credit card launched in Brazil in 2014, and Pentagram’s case page on the Nubank identity notes that the cards are nicknamed “roxinho”. Detailed reasons for the original colour choice are not set out in the primary sources used here, but the effect was a card that stood apart from incumbent bank colours.

Can an Indian NBFC advertise a zero-fee loan in a referral campaign?

It can promote a fee waiver, but RBI’s key facts statement rules require the borrower to see an annual percentage rate that includes all charges before signing. Design the journey so the all-in rate and the key facts statement appear before the agreement, and never let the reward screen skip them.

Does Yamm Labs work with NBFCs outside Gurugram?

Yamm Labs works from Gurugram and serves BFSI clients across India, and in the Gulf and other markets remotely. The brands we have worked with are listed on our clients page.