Banks, Co-Brands, and Credit Cards (#77)

Welcome to the 77th issue of Unit Economics. In this article, I discuss the credit card growth problem ailing the Indian banks, explain why banks have limited solutions, and suggest how the co-branding arrangements can help with role expansion for brand & fintech partners. Dive in!

I had written about the headwinds in 2024. Two years later, the credit card P&Ls continue to face the same pressures.

1. The delinquencies remain elevated, leading to higher provisioning and, consequently, increased credit costs. For example: SBI Cards’ write-offs up to Q3-FY26 are up 10% year-on-year, on the back of FY25, where the write-offs were 64% higher vs FY24. Credit costs, similarly, are up 2 percentage points from 6.5% in FY24 to 8.5% in FY26.

2. The increase in credit costs has reduced the margins across the board, and led banks to (a) cut down the card rewards and (b) make their risk policies more conservative.

3. This has impacted approval rates, particularly among new-to-credit-card cohorts. Leading to (a) lower issuance growth and (b) higher acquisition costs, with increased dependence on existing, carded customers for new card additions.

The increase in costs has coincided with sharp changes in customer behaviours and market economics, which are putting the two streams of income under equal stress.

1. Interest income, which forms >50% of the portfolio revenue, has been seeing a sustained drop in the % of revolvers in the portfolio. Across top banks, the drop is estimated at >10 percentage points, with that for SBI Cards at ~15% (from 38% to 23%). Banks are attempting to counter this by pushing up (a) conversion to EMIs, although these bear a lower APR, and (b) fee-based revenue, which fails to generate the same income volume, while also impacting the number of card applications.

2. Interchange income is also becoming less reliable with (a) increasing share of RuPay network cards in portfolios, lowering the blended take for issuers as low-value transactions migrate towards Card-on-UPI, and (b) regulatory pressures being faced by Visa and Mastercard globally.

The margin squeeze is making card portfolios more dependent on topline growth, which puts pressure on the business teams to reverse the slowdown in credit card issuances. Not only to increase the size of the pool, but also to reshape the portfolio mix towards a higher share of revolvers.

Banks cannot solve the growth problem by simply throwing more money at it. Especially when the traditional channels best known to banks are (a) getting costlier and crowded, and/or (b) lack the ability to provide sufficient scale. For instance:

1. An optimised performance marketing or offline tele-calling channel costs banks >₹2,000 per card (2X-3X for premium variants), and only becomes more expensive with larger targets.

2. Aggregators or marketplace platforms like Paisabaaar, Bankbazaar have become crowded with hundreds of affiliate partnerships, making them noisy and incapable of providing scale at the price that a T10 bank would command, and

3. Employer partnerships and salary account-linked cross-selling also lack the speed and volume that would excite.

Add to this that banks lack the appetite for creative ideas. The ‘experiments’ are not received well, and cost too many approvals. Instead, the growth has to come from channels they understand well. Leaving them with limited options that can offer both: high-scale and at predictable costs. Two channels meet the criteria:

1. Cross-selling to large, existing-to-bank (ETB) users, and

2. Co-branding partnerships

The channel provides the highest approval rate - due to targeted sell - and the lowest CAC, making it the first port of call for banks. But it’s also a channel which

1. Plateaus the quickest, since the eligible base gets depleted after 12-24 months. For the banks with long-running card programs, the adoption from the eligible base would have hit a mature number already, limiting the opportunity to an incremental 10-20% over BAU.

2. Further, the eligible pool is impacted significantly by the exclusion or limitations put towards multi-carding. This is often done since these users (a) introduce cost-expansion risk through ‘double-dipping’, i.e., by spreading their spends to maximise rewards on the multiple cards, and (b) cannibalise spends from existing cards, providing low or no expansion of the spending pool.

3. Of the increment driven via the ETB channel, often the highest is through the digital - particularly in-app - interfaces. However, the internal org design at banks pits the business units (BUs) against each other, and ends up dividing the returns into smaller, less-significant pies, limiting the effectiveness of the channel.

All combined, it’s a channel that banks continue to move aggressively on, but also one where the returns deplete fast, making it less exciting. Leaving banks with one last bet.

This channel commands the highest share of digital sourcing for a lot of banks already. But to understand it, it’s important to understand the two types of co-branding arrangements banks engage in:

1. Brand-led: where the objective of the brand is to drive higher on-platform spending through better average order values (AOVs) and spend frequency. The core offering of such brands is usually non-financial in nature. Some of the most well-known co-brand partnerships in India are brand-led → Amazon-ICICI, Flipkart-Axis, Swiggy-HDFC, IRCTC-SBI, BPCL-SBI, among others.

2. Fintech-led: where the objective of the fintech is to drive higher acquisition and/or generate revenue directly from the card program. Fintech companies often use the card proposition to acquire users off-platform and invest significantly in the card programs to get scale. Some of such partnerships include Scapia-Federal, super.money-Axis, PhonePe-SBI, OneCard, and TataNeu-HDFC (which is also a mix of Brand-led).

If stitched together well, co-branding partnerships are uniquely arranged to solve banks’ acquisition problems, offering them:

1. Access to a potentially large, new-to-bank (NTB) base,

2. Fixed costs of acquisition, with banks providing anywhere from ₹500 to ₹2,000 to the co-brand partners, while the co-brands take the hassle of acquiring cardholders.

3. Alternative data for underwriting, with users actively engaging with the co-brand platforms, which allows banks to build joint-risk models to approve thin-file or new-to-credit users.

4. Better funnel conversions than a performance marketing or marketplace channel, with more contextually embedded journeys (checkouts, app-home, etc.), and

5. Brand goodwill, as co-brands market the card to a wide base, with the bank’s branding made mandatory by the RBI for all marketing. This is particularly beneficial for the smaller banks such as RBL, Utkarsh SFB, and Federal Bank.

‘More co-branding partnerships’: Is that the solution then?

I argue that it can be, but it’s not today. Not until the industry fixes what’s ailing them: (a) the RBI’s guidelines on co-branding arrangements and (b) credit card economics for co-brands.

Today, the RBI guidelines explicitly limit the role of a co-branding partner to “marketing / distribution of the cards”, and go on to suggest that “post issuance of the card, the CBP shall not be involved in any of the processes or the controls relating to the co-branded card”, limiting a CBP’s role to a distribution agent.

This leads to a very odd arrangement where the lack of data with CBPs and the lack of customer ownership with banks – since CBPs own all interaction interfaces – makes it equally challenging for either to intervene post card-issuance.

Result? Untimely, unpersonalised, and shallow interventions, leading to poorer numbers across spends, retention, re-activations, and repayments – all of which impact the portfolio health of the program.

Compare that to a more mature credit card economy like the US, where the co-brand programs, such as Delta-Amex or Marriott-Chase, have deep integrations across the lifecycle of the customer, allowing

1. Transaction data to be used for (a) up-selling – tier upgrades, personalised pricing, milestone offers, (b) spend and retention interventions / experiments – spend campaigns, CRM nudges, and (c) cross-selling – recommendation system for flights, hotel stays, or taxis, and

2. CBP’s brand equity to be leveraged, allowing cards to be perceived as partner-brand cards: “Delta SkyMiles Card”, “Marriott Bonvoy card”, without forcing bank-first branding. This provides more freedom to marketing teams and makes it easier for them to build an emotional ownership in the minds of customers for the program.

The commercial arrangement between the banks and CBPs draws from the RBI’s definition of such partnerships. Where the incentives of the co-brand lean towards CAC – instead of LTV – optimisation. The partnerships invariably kick off with misaligned incentives, where the CBPs care more about building deeper distribution and not enough about the credit lifecycle.

This breaks the compounding effects on spending and retention that an LTV-focused P&L can drive. Also limiting the margin expansion of the co-brand programs, which is a self-goal for the credit card ecosystem that already struggles with:

1. Decreasing interchange rate,

2. Low revolver share, and

3. Competition with UPI for top-of-wallet spends

The regulations and the incentives from banks are both designed today to narrowly define the co-branding arrangements, with the ultimate goal of containing ecosystem risk. This, I believe, is behind the poor outcomes of a lot of such partnerships.

Instead, the focus should be towards ecosystem expansion through such arrangements, which would require (a) deeper integrations, allowing data sharing and access on both ends, (b) LTV-focused economics, where CBPs feel inclined to drive compounding post-onboarding behaviours, and (c) partner-first branding, allowing partners to not feel tied in their marketing campaigns.

If you want to discuss the topic, please reply to the article or feel free to reach out to me over email or LinkedIn. Until next time!

Share