How do you build a marketplace bank?
This essay first appeared in two parts on Fintech Futures.
It's 2021 and you're in Amsterdam for The Future Of Banking, a mega-conference packed with every Head of Innovation from companies across the globe.
Lights dim, futuristic music plays and a grey-haired man wearing a suit with sneakers steps onto stage.
"In five years, retail banking will be unrecognisable" he proclaims.
"Customers will pick and mix products from ecosystems curated by banks. We will transition to a truly customer-centric model of finance. The bank of the future will be a marketplace!"
You sigh softly.
Another keynote with the same tired theme! If only someone would actually venture how this model might work, to go beyond the catch-phrases and into the messy work of defining, prototyping and developing...
I am fascinated by what banking may look like in the future. I've read the consultancy white papers that outline grand and abstract visions. I've worked with industry practitioners, who grumble about implementation difficulties. And I've spoken with fellow founders, who sit uneasily between these two worlds. As yet, I haven't come across a comprehensive and compelling analysis of marketplace banking, so I'm writing one.
Over three parts I will answer these questions:
- What is the best model for marketplace banking? How might it actually look in practice?
- Who is in the best position to effect this evolution: legacy banks, challenger banks or third-party aggregators?
- How can fintech products win in the marketplace? Along what lines will the terms of competition be drawn?
- How can the marketplace bank make money?
These are my views, formed over the past two years, and I'd welcome criticism and debate.
What is the best model for marketplace banking?
The best model for marketplace banking is a small number of contextual, deep integrations into third parties. It is not an extensive menu of options, the kind that platform businesses like Amazon and Google operate in other industries.
Monzo's savings marketplace is a great example of marketplace banking for financial services. Once a user has selected 'Earn interest on your money' from the main menu of the app, it takes only ten clicks (yes, I counted!) to open up an OakNorth savings product. During the process, the user never leaves the Monzo app and is not required to re-enter any identifying information that Monzo already has.
There are only five options on offer at the time of writing, all of whom have competitive (though not market-leading) rates. The selection is not overwhelming, there is an impression that Monzo has done the sorting for you already.
Starling's Marketplace is designed more like a menu of options. In order to find a third-party product, the user needs to navigate to the Marketplace from the main menu, rather than being directed in the course of trying to solve a contextual problem (e.g. 'I need to earn interest on my money')
Further, products across multiple categories, from credit to insurance to investments, are displayed one after another in a scrolling table. Finally, in order to access a third-party product, the user is redirected to their browser where they need to re-input identifying information. There’s a relatively shallow level of integration between Starling and the third-party product and Starling.
The difference between Monzo and Starling in this instance is a question of degree. Neither are seeking to imitate an Amazon-style search ("97 results for 'investing'!), but Monzo has taken integration a step further by making third-party products a natural part of their user experience. They could improve this even further by serving contextual recommendations for saving or investing to users who consistently have surplus money each month.
The contextual, deep integration approach to marketplace banking works well for three main reasons:
- First, financial product decisions hinge on high trust. It is far more personal to open an investment account than it is to buy a pair of noise-cancelling headphones. An extensive menu of options (like what you would find in Amazon or Google Shopping) works incredibly well for consumer products, where the consequences of an incorrect purchasing decision are less severe and generally reversible. But this same approach is paralysing when making financial decisions. Even price comparison sites like Money Supermarket –an earlier manifestation of marketplace banking – establish trust by asking clarifying questions before presenting options to users.
- Second, most financial products are low frequency purchases. A user might purchase an insurance product once a year, so would have relatively low familiarity with the available options and the differentiating factors between them. An implicit recommendation by their bank – made by including a product in a small list of preferred providers – has the potential to play an important role in decision-making.
- Third, consumers simply don't want to scroll through a large menu of services. It's time consuming and cognitively difficult. Consumers constantly rely on rules of thumb for their financial decisions, so introducing too much choice will hinder rather than help the process. Banks occupy a position of authority for their customers, and recommending third-party products is one area where they can exercise that authority.I find choice overwhelming when trying to buy towels on Amazon, let alone a life insurance policy!
Who is best positioned to provide marketplace banking?
The main groups of competitors looking to provide marketplace banking are incumbent banks, challenger banks and account aggregator products. I believe that challenger banks are the best positioned to win this race and, in effect, become the focal point of consumers' financial lives.
Challenger banks typically facilitate their customers' day-to-day spend. In 2018 there was industry scrutiny on being the bank that consumers pay their salary into. I believe this focus is misplaced. In order to understand a customer’s behaviour, a bank needs information about their everyday transactional spending (not just the size of their pay packet). It is easily the highest frequency interaction that customers have with banks, and yields a wealth of data about habits that can be used to improve service.
To take a simple example, I pay my salary into an HSBC account but conduct all my routine spending using a Monzo card. HSBC knows my salary, but Monzo knows that I went to Slovenia two months ago. Spending behaviour provides a much fuller view of the financial products and services that I could be interested in.
Incumbent banks lack the technical agility to use spending data effectively. Customer data is often trapped in silos, generally segmented by product. It is a non-trivial challenge to match a customer’s mortgage history with their credit card spending within the same bank, let alone extracting value from this information so to make specific recommendations back to customers. This problem is not uniformly distributed – some incumbent banks like DBS in Singapore and Commonwealth Bank in Australia are rapidly evolving – but it is common.
Account aggregators like Yolt and Money Dashboard struggle to extract great insights from the data provided via Open Banking. There are a few constraints to how banks provide payment information:
- There is no unique identifier attached to transactions so it's difficult to consistently identify payments;
- Merchant details are only sometimes available, and often obscured in transaction description fields; and
- There is no standard categorisation of transactions.
Part of the reason so many middleware providers (such as Yapily, TrueLayer, OpenWrks and Plaid) have sprung up is to contend with these technical challenges.
There is also a limit to the number of times each day that third-party applications such as account aggregators can request refreshed transaction data. They're effectively always a few hours behind what has actually happened in their customers' transactional account, rather than enjoying a real-time view. So long as this constraint exists there is real power in being the provider of payment infrastructure to customers.
How can fintech products win in the marketplace?
If marketplace banking is the future, then any fintech product that isn't a core payment account should be designed to be "plugged in" through the marketplace. How do other financial products differentiate themselves in this scenario?
Michael Porter's generic strategies for competition are a useful framework to answer this question. The three strategies that he identifies for standing out in a market are cost leadership, differentiation, and focus.
Cost leadership is providing the lowest cost option for a product. In financial services this works well for commodity products – those that can be easily compared between providers and have simple on-boarding processes like savings accounts. Marcus by Goldman Sachs' Online Savings Account is a fantastic example of this strategy. They led with a market-leading rate and attracted impressive initial interest.
Differentiation is creating a product with attributes that are uniquely valuable to customers. This works best for product categories with a strong customer service component. An easy example is travel insurance, where even if the price is higher, a consumer may choose one product over another because the claims management process is friendlier. Pluto Travel, for instance, has attracted phenomenal Trustpilot reviews and built their proposition around a great customer experience rather than leading on price.
Alex Rainey, CEO and co-founder of Pluto, says: We quickly realised in travel insurance that in order to win on price, you need to cut some pretty scary corners and massively reduce the quality of cover you’re providing, this isn’t something we were ever willing to even explore…so we decided that one of the things Pluto would be known for was rockstar customer support. Monzo especially and others have proven so well that 24/7 human support through your smartphone was a must and people loved them for it.
Focus is offering a specialised service to a defined subset of consumers. For highly popular customer objectives like "I want to earn interest on my money", this will not be relevant. It is more important for niche objectives. For example Hammock, a new current account, markets itself exclusively for landlords. It offers automatic reminders to tenants, tax statements split by property, and estimates of property taxes. Landlords can (and have previously) done all of these things separately, but combining them into one proposition makes for an attractive proposition to their target market.
Manoj Varsani, CEO and founder of Hammock, says: “We want to focus on providing the best value to our users. Addressing a specific user group, rather than trying to support all consumers at once, allows us to identify clear priorities and to develop a product tailored-made for the property sector (landlords, letting agents and tenants).”
Can the marketplace bank make money?
A marketplace bank earns income by referring its customers to third-party products, either as a one-off transaction fee (e.g. travel insurance) or as a share of the recurring income that the third-party earns (e.g. savings and investments). It is effectively an affiliate business model. Whether this income is sufficient to justify the model is so far unclear.
Banks and aggregators may be better off building their own products in the long term. Ricky Knox, CEO of Tandem Bank, has stated previously that marketplace banking is not an important revenue stream in the bank’s strategy. He says “the marketplace is a core generator of customer benefit and not really the core place that we go to make money.” It’s telling that Revolut, which earns the most revenue per customer of any challenger bank, has so far eschewed the marketplace model.
There is also the existential risk that big tech, specifically Apple and Google, will take a more active role in recommendations to customers, effectively cutting out the aspiring marketplace providers. John Heaton-Armstrong argues that their ownership of the mobile devices and operating systems enable them to have a closer, more effective relationship with end customers. Further, these firms have “revenue streams beyond simply providing accounts, and as such could run these at an operating loss leveraged against profit margins generated elsewhere.” They might leverage their position to squeeze the profit from marketplace banking entirely.
Despite this, the lower cost base of aggregators and challenger banks might still allow them to operate sustainably on a marketplace model. A business model based on marketplace revenue may not be as rich as one based on net interest margin, but lower costs might offset this difference. It also avoids the risk of being disrupted by peer-to-peer lending. If loan aggregators like Habito and Trussle gain significant market share, the traditional business model of banking may be even riskier to pursue.
So how to proceed from theory to practice?
Aspiring marketplace providers should ask themselves the following questions:
- What customer objectives (or jobs-to-be-done) are related to my core value to customers?
- What are the main products serving this objective? Of these, which are consistent with my product’s values and experience?
- How can I help my customers access these products in the most frictionless way possible?
Other fintech firms should in turn be thinking about questions such as: what is my competition strategy? On what basis do I stand out against my competitors?And which aggregators and banks share a similar set of brand values to approach for partnership?
The early versions of marketplace banking will not match the grand words uttered from conference stages, but it will be the start of an important and irreversible shift!