At Stellaris, we have historically carried a level of scepticism on the sector, i.e. companies selling SaaS to banks, wealth management firms, and insurance companies in India. This is primarily on two counts – firstly, lack of conviction in the appetite of banks to pay for software (not including IT services), which in turn could imply a lower ACV (annual contract value) per bank, and secondly, the limited universe of needle-moving logos, i.e. large banks.

Unlike in the US, in India, large banks are still far and few. When you combine the two factors, we were not convinced that a bank SaaS startup can reach a $100M ARR (annual recurring revenue) scale in a span of 8-10 years with headroom to grow more – a requirement for a fund of our scale. However, in light of bank SaaS companies like Perfios, M2P, and Lentra reaching meaningful scale, especially post COVID, we decided to study the sector and refresh our views on the space. Here is a summary of our findings.

Bank SaaS continues to be challenging for the following reasons:

1. Banks’ existing software stack is very sticky and hard to displace: Banks’ existing software is often custom-built, developed either by an IT services company or the bank’s in-house IT Team. Over time, these customisations, including custom integrations, have compounded. As a result, the replacement cycle involves significant change management on the bank’s part, creating a massive inertia to replace.

2. Universe of bank logos to sell to is limited, concentration risk an inevitability: Private banks are more open to adopting SaaS than PSU banks. Within private banks, it’s only a handful of accounts (<20) that really matter in ACV terms. This limits the target market for SaaS companies in India, unlike the US or China, where there are a larger number of sizable banks to sell to. A downstream effect of that is customer concentration risk, often reflected in current bank SaaS companies, where 2-3 accounts contribute to the majority of ARR.

3. No concept of MVP, need a near-final product at get-go: Principles of ‘The Lean Startup’ fail here. Any new product needs approval from several teams within a bank, including IT, information security, compliance, and legal. As a result, minimal viable products don’t get the necessary clearances. Additionally, the concept of fast iterations is impractical, as each iteration requires going through the same approval process with multiple teams.

4. Long sales and implementation cycles: Bank SaaS companies often take the ‘exclusivity’ approach with an anchor bank for the first 3-4 years. This is usually how long it takes to sell, implement, and create a successful case study, that can be leveraged to approach other banks. This also explains why the gap between seed and Series A funding rounds for bank SaaS companies can extend to 3-4 years.

5. Need a very specific founder archetype to sell to a bank: Founders need to speak the banks’ language, navigate various stakeholders inside the bank, appreciate and understand the bank’s buying cycle, and lastly adopt a quasi-service mindset. When dealing with Indian banks a stance like “we’re a product company, and won’t do customisations or custom integrations” doesn’t go well. Founders with prior experience building and selling to banks are more likely to succeed in this space. This includes those who have run IT services companies serving banks, senior ex-bankers with ideally one founder from the bank’s in-house IT team, or team’s from bank SaaS companies.

6. High need for customisation and custom-integrations: While this may be true for enterprises in general, it is even stark for banks. A bank’s existing stack is often custom-built by an IT service provider, with workflows that are highly tailored. Serving a bank inherently requires customisations rather than one-time efforts. This impacts scalability and margins, at least in the early days, but on the positive side, it also creates defensibility.

7. Multiple products (and PMFs) needed to reach $100M ARR: Given the limited number of bank logos, there is a ceiling to how much a product can scale in terms of ACV. As a result, bank SaaS companies need to develop a suite of products over time to increase their ACVs from banks. This adds complexity to the business, as it requires building and establishing PMF for multiple products.

While challenges are plenty and the $100M ARR question still remains, bank SaaS also have several attractions, as below:

1. Existing stack is old and carries poor NPS

There is high dissatisfaction, especially among users of software within banks, with the existing legacy tech stack. No one we spoke to disputed this issue, validating the presence of a problem that is widely recognized. The existing stack suffers from high upfront pricing, archaic architecture which makes it less agile and incompatible with new initiatives (e.g., credit on UPI).

2. Expanding tech spend in banks, increasing SaaS mix

There has been a material increase in tech spends from what used to be nearly 5% of OPEX a decade ago, to nearly 10% of OPEX today for several banks. Drivers include:

a. Regulatory tailwinds for tech modernisation: The regulator has made its intent clear that banks must upgrade their software. They’ve demonstrated the seriousness of this mandate by slapping recent penalties on Kotak and barring HDFC from issuing new credit cards (which led to no new issuances for the leading credit card issuer in the country for nearly 9 months).

b. Competitive race among banks, especially private banks: Competitive intensity drives a domino effect where if one modernises their software, particularly systems that directly impact revenues, others quickly follow. During my time as an operator at Ola, partnering with a leading bank on credit cards, I saw first hand how board-level pressure pushed the bank to enable instant card issuance and increase STP (straight-through processing journey without manual approvals) because their primary competitor already offered these capabilities. The “so-what” for a bank SaaS company is the relative ease of acquiring the second and third customers once a successful case study is established with the first bank.

c. Evolving consumer needs, fintechs raising the bar for banks: It’s no secret that bank apps and net banking are being unbundled by fintechs, one use case at a time. Payments have largely been unbundled by UPI apps, information viewing (bank balance, recent transactions etc) is beginning to shift with the advent of account aggregator, and there have been some strides in credit as well. It’s well known that CRED has created a large lending book by targeting bank’s key customers – metro affluents. This makes it clear that customers now place a higher premium on user experience, especially post-UPI, and banks are lagging behind fintechs in this regard.

3. High ACVs possible on the back of usage-based billing

This approach is evident in existing bank SaaS companies like Perfios, Lentra, M2P, which charge per file, per API, per transaction, respectively. This model allows these companies to scale their ACV as they grow transaction volumes with banks, avoiding a fixed revenue ceiling each year. However, on the downside, this approach can negatively impact cash flow cycles, as it requires navigating through billing, reconciliation, and receivables processes with banks instead of collecting cash upfront.

4. High NRR with relative ease of cross-selling new products and to new buyers within bank

Banks have a high bar for onboarding new software vendors; however, once proof of reliability is established, they are more open to buying additional products from the same vendor. This leads to higher NRR (net revenue retention) compounding for a bank SaaS company compared to, say, a US non-bank SaaS company. Many examples illustrate this, such as Perfios successfully cross-selling new products to the same buyer – credit underwriters like Bureau Connect, GST Connect, Karza, and various fraud offerings. Similarly, Credgenics sells products in the same category (collections) to different buyers within the bank, such as the head of business loans.

5. Network effects allowing first mover to get to dominant market share

This was a significant revelation for us. Examples can be seen in Perfios, Lentra, M2P, which are the dominant leaders in their flagship product categories: bank statement analyser (or credit intelligence more broadly), loan origination systems (LOS) for consumer durables (CD), and pre-paid stacks, respectively. The phenomenon of key opinion leaders (KOLs), or influencer banks, gives software companies a disproportionate advantage in securing follower banks once they establish a successful case study with a KOL bank. HDFC, ICICI, Axis, Kotak are considered KOL banks for private banks, while SBI, BOB, PNB hold that status for PSU banks.

6. Banks’ openness for multi-stacking can serve as a wedge to land the bank

The practice of multi-stacking or parallel stacking of LOS in banks is well known. I was surprised to see some mid-to-small sized banks are also willing to multi-stack their core banking systems, which is how M2P’s core banking product is landing customers. This approach lowers the barrier for banks to try out new software, as they don’t necessarily need to undergo a replacement cycle; instead, they can experiment while parallel stacking, and kick the decision of replacement down the road. This can help shorten sales and implementation cycles for bank SaaS companies.

7. Ability to command a price premium and high GMs (80%+)

We learned that in the first few years, gross margins (GMs) are lower, ranging from 60-65%, due to the high service component involved. However, over time, these margins can expand to 80-85% as the service component reduces. Additionally, existing SaaS companies can command price premiums by virtue of being sticky. For instance, Perfios’ flagship product, bank statement analyser, commands a price premium despite facing credible competition from companies like Finbox, Lentra and others that offer more competitive pricing.

At an overall level, we came out much more positive than going in. Bank software for India has clearly come a long way. 

Regarding the segments within bank SaaS that interest us, we don’t have a very sharp view. However, it will be important to pick a segment that addresses a pressing need for banks – a ‘painkiller’ segment – to shorten sales and implementation cycles. Additionally, there may be merit in selecting a segment with limited credible competition, allowing for the potential to scale quickly, even if one isn’t the first mover in that segment.

Based on our cursory conversations with bankers, segments such as fraud intelligence – particularly mule account fraud – compliance tech, co-lending infrastructure, and specialised LOS for business loans might be interesting segments to consider. In addition, AI copilots for bank’s front-end teams (e.g., sales, RM, collections, support teams) could present an exciting opportunity.

We carry a preference for the below as we look at investment opportunities in bank SaaS, however, these are mere preferences.

1. Contracted with at least one KOL bank

2. Contracted on a usage-based billing model

3. No credible competition, materially ahead, in the segment at the time

4. First or flagship product has the potential to scale to at least $30-35M ARR over 8-10 years

We look forward to more founders building for banks, and would love to chat. Do write to us at mayank@stellarisvp.com or shreyan@stellarisvp.com.

Originally published in The Financial Express