Designing Rewards for Growth at Carbon and Branch: Loyalty Tiers, Cashback and Referrals

In Nigeria, anyone who wants a digital account has plenty of choice: OPay, PalmPay, Kuda, FairMoney, Carbon, Branch and more. Opening an account takes minutes, so the harder job is getting people to use it: to keep money in the wallet, pay bills, use the card and come back for a loan.

I worked on that problem twice. At Carbon the answer was tiers: the more you used Carbon, the higher your tier and the more you could access. At Branch, where Carbon was one of the apps we competed with, the answer was money: cashback when you paid a bill, and a reward when a friend you invited repaid their first loan.

Both programs reward the same thing, using the app, but each had its own job. Carbon needed a fair way to decide who gets new benefits first. Branch needed one number to move: bill payments. And they pay out in very different ways, which shaped almost every design decision.

Carbon is a credit-led digital bank in Nigeria with more than four million customers. Carbon wanted to offer its customers more: better rates, faster support, buy now, pay later. Offering all of it to everyone at once would have cost too much, so the real question was who gets it first. Tiers answered it. The most committed, most active customers, the ones who keep money in their wallet, transact, save and repay on time, unlock benefits first, and everyone else can see what they’d get by using Carbon more.

There are six tiers. Moving up unlocks things like higher loan offers at lower interest, applying for a loan twice a day, better savings interest, faster customer support, Carbon Zero (Carbon’s buy now, pay later, which spreads a payment over four instalments at 0% interest) and a direct line to the contact centre.

Each tier got its own colour and badge. Here’s a customer climbing from Tier 1 to Tier 6:

Carbon’s six tiers, from Tier 1 to Tier 6. In the progress row, the tiers a customer has passed are ticked, their tier is highlighted in its colour with “You” under it, and the tiers ahead are locked.

1. 
2. 
3. 
4. 
5. 
6. 

1. 
2. 
3. 
4. 
5. 
6. 

Pause

What I wanted was a points counter that grew with every transfer and repayment, so customers could watch themselves get closer to the next tier. Research on reward programs supports this: people work harder for a reward as they get closer to it.

The data didn’t allow it. Carbon recorded every activity that counted, from wallet balances and transfers to card payments, savings, Carbon Zero, loans and repayments, but it couldn’t total points in real time. Instead, a query ran across every customer at the end of each quarter and moved them up or down a tier.

So I designed for what customers could rely on: the tier they’re on, what it gives them now, what the next tier adds and what gets them there. Progress shows as tiers, not points. Here’s how a customer moves between them:

Dropping a tier feels like losing something, so Carbon warned customers before it happened. Customers at risk of moving down got an email before the quarter closed, with ways to keep their tier, and customers who moved up got a welcome. Here are two of those emails:

Customers reach their tier from two places they already use: their profile picture on the home screen carries a tier badge, and the Account page shows a tier label beside their name.

Both open the tiers screen. It had to answer three questions in order: where am I, what do I get, and how do I get more. Here’s how the screen is ordered, from the top:

1234567

1. 1Where you stand. Your photo with your tier badge, and “You’re on TIER 3” in your tier’s colour. It’s the first thing customers look for, so it leads the screen.
2. 2How far you’ve come. Six steps in a row: tiers you’ve passed are ticked, yours is marked “You”, and the ones ahead are locked. It shows progress without a points total the data couldn’t keep up to date.
3. 3What tiers are. A small “What are tiers?” link for anyone new, which opens a short explainer. It’s kept quiet so it doesn’t compete with your status.
4. 4Now or next. Two tabs separate what you have today from how to get more, so neither list gets long.
5. 5Your benefits first. What you can use right now, each with an icon and one line on what it means for you.
6. 6What’s next. The next tier’s benefits sit right under yours, so every visit shows what you’d gain by moving up.
7. 7One next step. “Level up to the next tier” closes the list and opens the second tab.

The second tab turns the next tier into a plan:

1234

1. 1The goal. “Next tier: Tier 4”, with a lock, so the target is concrete.
2. 2The reward. What Tier 4 adds, repeated here so customers don’t have to switch tabs to remember why it’s worth it.
3. 3The steps. Three things that count, in plain words: keep ₦20,000 in your wallet, transact regularly and save.
4. 4The detail, if you want it. “Learn more” opens the full article on how each activity earns points.

Once the benefits for each tier were agreed, I designed all six versions of the screen, so each one shows the right benefits and the right next step. At Tier 6 there’s no next tier, so the second tab becomes “Maintain your elite tier status” and explains how to keep it.

I tested the screens with customers and refined them from their feedback before they shipped. Tiers are live in Carbon’s app on iOS and Android, and Carbon’s help centre walks customers through all six.

At Branch I took on rewards as the growth team’s product designer. The product manager and I shaped the flow and journey together; I designed the screens and the gift illustrations, working with the engineers in our growth pod. The goal matched Carbon’s, getting customers who already had a Branch account to use it, but we had one number to move: bill payments, which were lagging behind the rest of the app.

This time the reward was money. Rewards launched on Android in Nigeria, Kenya and India. Here’s how customers earn and spend it in Nigeria:

Tying the referral reward to a repaid loan means Branch only pays for friends who actually borrow and repay. Limiting cashback on bill payments to five a month keeps its cost predictable.

I designed for two kinds of customer, and each needed something different first:

A new customer opening the Rewards tab has nothing to spend yet, but they can start earning straight away. Here’s how the tab is ordered for them:

123456

1. 1Your invite code. The one thing a new customer can act on right away, so it leads, with Copy attached.
2. 2The deal. “Earn ₦2,000 in rewards when friends you invite repay their first loan. They earn ₦2,000 as well.” It sits right under the code, so the reason to share is next to the thing you share.
3. 3The other way to earn. “Earn 2% cashbacks on bill payments” is a row, not a card, because it’s the second option.
4. 4Joined through a friend? “Redeem invite code” is there for customers who came in with a code.
5. 5Questions. “Learn about rewards” opens the FAQ, last.
6. 6One main button. “Invite Friends”, pinned at the bottom of the screen.

Sharing and redeeming each take one screen:

The gift box illustrations came from the illustration workflow I built at Branch, so they match the rest of the app.

Once a customer has earned something, the order flips. The balance moves to the top, because it’s what they came to see:

123456

1. 1Your balance. What you’ve earned and not spent, in the largest type on the screen.
2. 2Spend it. “Spend Rewards” is attached to the balance and opens airtime and data, so earning leads straight to spending.
3. 3Keep earning. Invite and earn and bill cashback move down into cards: still easy to reach, but no longer the headline.
4. 4Redeem a code. Still one tap away for anyone who has a friend’s code.
5. 5Check it. “View rewards history” shows everything earned and spent.
6. 6Questions. “Learn about rewards” opens the FAQ, last.

A returning customer mostly wants to know one thing: did I get the cashback I expected? So every cashback appears as its own transaction, the transaction history has a Rewards filter, and a rewards history page shows the balance with everything earned and spent.

The FAQ answers a question that matters when money is involved: is the bill discounted? No. You pay the full amount, and the cashback lands in your rewards balance straight after.

I built every screen with Neem, Branch’s design system, so Rewards uses the same app bar, cards, list rows and buttons as the rest of the app. We tested the flow internally and with customers across several rounds before launch.

Both programs reward people for using the app, but they differ in what customers get and how soon they see it:

The biggest difference is timing. Cashback lands seconds after a bill, while a new tier can take a quarter. That’s why the Carbon screens lean on what each tier unlocks, and the Branch screens lean on the balance and its history.

Branch

Bill payments were the number we set out to move, and they went up after Rewards launched, as customers earned cashback and spent it on more bills. Other numbers moved with them, including loan applications. Rewards launched on Android in Nigeria, Kenya and India, and it’s still live in Nigeria today, in an app with more than 50 million downloads on Google Play.

Carbon

Tiers let Carbon offer new benefits to its most active customers without the cost of giving them to everyone. They’re live in Carbon’s app on iOS and Android, with a help centre guide to all six, and Carbon was still emailing customers about their tier in 2025. Carbon has more than four million customers, and its app has more than 5 million downloads on Google Play.

These programs taught me to:

Check what the data can do before designing.

At Carbon, the quarterly review shaped the whole experience, not just one screen.

Match the screen to how fast the reward arrives.

Fast rewards need a visible balance and a history. Slow rewards need a clear picture of what’s coming.

Explain once, then get out of the way.

New customers need the explainer; returning ones need their balance.

Show the next step at every level.

Even at the top tier, there’s something to do: keep it.