Every company is becoming a financial services company.
Retailers, marketplaces and platforms are already selling financial products under someone else's licence — and the split between relationship, product and balance sheet is deciding who captures the value.
Amazon is not primarily a retailer any more. Alongside logistics and cloud computing, it now brokers business loans and processes payments at checkout — all through the relationship it already has with customers and sellers. It is not alone. Across the UK and Western Europe, retailers, marketplaces and platforms are becoming financial services companies one embedded product at a time, and the institutions that used to own that territory are discovering that a banking licence buys less influence than it once did.
The pattern
This is not a niche experiment, and it is not confined to consumer credit. Embedded finance generated an estimated €20 billion to €30 billion in revenue across Europe in 2023 — roughly 3% of total banking revenues — and grew three times faster over the previous decade than loans distributed directly by banks, according to McKinsey's 2024 analysis of the European market. McKinsey's central case has the market surpassing €100 billion by 2030, accounting for 10 to 15% of banking revenue pools. Buy-now-pay-later is the most visible slice of this shift: across seven European markets, its share of e-commerce sales rose from roughly 2% to 10% between 2016 and 2023. None of this required the retailer, marketplace or platform to become a bank. It required them to find a bank, or a licensed lender, willing to sit behind the product — which, as the next section shows, is exactly what several of Europe's largest retailers have now done, in strikingly different ways.
The three layers
Aulay's shorthand for what is actually happening is three layers, and almost every embedded-finance product can be pulled apart along them. The customer relationship is who the customer trusts enough to buy from, sign up with, or open an app for. The embedded financial product is what actually gets offered inside that relationship — a loan, a card, a buy-now-pay-later instalment, an insurance policy. The balance sheet is who is licensed to carry the underlying credit or insurance risk, and who absorbs the loss if it goes wrong. A bank, historically, held all three at once. Most of what is being built now deliberately separates them — and the separation is the business model.
Why the split makes economic sense
The economics explain why this keeps happening rather than staying a one-off deal. Distribution — reaching the customer, building the interface, using existing trust — scales cheaply once it exists. Carrying credit or insurance risk does not: it requires regulatory capital, underwriting infrastructure, and a licence that takes years and real money to obtain and maintain. Splitting the two lets a retailer or marketplace capture the cheap, scalable part of the value chain and rent the expensive, regulated part from someone who already has it. It also lets a bank or licensed lender sell its most expensive asset — balance-sheet capacity — into relationships it could never have built, or afforded to build, itself. Both sides get something they couldn't get alone; the question the deal actually settles is how the margin between them gets split, and who keeps the option to change partners later.
Three companies, three different splits
Tesco. Tesco is the clearest UK illustration, and a recent one. When Barclays completed its £0.6 billion purchase of Tesco's banking business in November 2024, Tesco kept the brand, the Clubcard integration and the customer relationship on credit cards, personal loans and deposits, under an exclusive marketing arrangement running an initial ten years. Barclays took on the credit and balance-sheet risk for that book. Tesco held onto a narrower slice for itself too — insurance and money services, including gift cards, travel money and ATMs — though those are being rebranded away from the Tesco Bank name over the following two years, a reminder that these deals rarely split perfectly along one line. The core signal is the same regardless: the relationship stayed exactly where it was; the risk moved to whoever was better placed, and better licensed, to carry it.
Amazon. Amazon runs the same logic on the lending side of its marketplace. Since June 2023, sellers on Amazon's UK marketplace have been able to apply for financing of £500 to £2 million directly through Seller Central — no fixed repayment schedule, no late fees, repayment tied instead to a percentage of future sales, so it eases off automatically in a quiet month. Amazon owns the seller relationship and surfaces the offer inside an interface sellers already use every day; the underwriting and the credit risk sit with YouLend, the licensed third-party lender behind the product. Amazon never touches a banking licence to do any of it, and doesn't need to — sellers apply through the same dashboard they use to manage inventory and advertising, which is precisely the point.
IKEA. IKEA shows the split can run the other way. Ingka Group, IKEA's parent, first took a 49% stake in its long-standing financial-services partner Ikano Bank in 2021, then completed the purchase of the remaining 51% in October 2024 to take full ownership. Ikano Bank funds IKEA-branded credit cards, financing and savings products across several European markets, including the UK. Rather than hand that balance sheet to a partner bank the way Tesco did, Ingka concluded it was worth owning outright — a bet that the margin from running the lending itself was worth more than the fee income from renting it out to someone else. Two large European retailers, a similar starting position on the relationship, opposite conclusions about who should hold the risk behind it.
What determines who wins the relationship
Not every company that embeds a financial product ends up owning the relationship for the long term, and the three cases above suggest what separates the ones that do. Frequency matters: Tesco's Clubcard already generates a weekly, sometimes daily, reason to interact before any financial product enters the picture, a different starting point from a bank a customer sees a handful of times a year. So does data: Amazon can price a seller's financing risk using real-time sales data a traditional lender would need weeks to assemble, part of why it can turn a financing decision around in days rather than a typical underwriting cycle. And so does whether the underlying trust was built for a reason that had nothing to do with money in the first place — nobody chose IKEA because of Ikano Bank; the bank exists because IKEA already had a relationship worth financing. A bank trying to compete on relationship alone, without one of these three advantages, is not competing on the same terms.
The uncomfortable question for banks
Put those three cases together and the uncomfortable part for a bank executive is not that fintechs are picking off market share. It is that Barclays now carries meaningful UK consumer credit risk on a book where the customer doesn't particularly think of Barclays at all — they think of Tesco. Owning the balance sheet bought Barclays scale and a fee stream, not the relationship, and the relationship is where the next product gets sold, the next cross-sell happens, and the next competitor gets kept out. Whether owning the balance sheet is worth it on its own terms depends on the economics of the specific book: Tesco decided it wasn't worth running itself; Ingka decided the opposite about Ikano Bank, in the same year. Neither answer is universally right. What both cases confirm is that the customer relationship and the balance sheet are no longer bundled by default — and a bank whose strategy still assumes they are is planning around a structure that increasingly doesn't hold.
The IKEA-style bet isn't confined to Europe, either. On 8 September 2026, Chime — a US financial-technology company — announced an agreement to acquire Stride Bank, its banking partner for the past seven years, for $590 million, about 1.5 times tangible book value. Chime's own stated logic echoes Ingka's almost exactly: own the charter outright, cut the fees paid to a partner bank, and capture the economics rather than rent them. Stride Bank has no UK or European presence, so the deal itself carries no direct relevance to this market — but the reasoning behind it does, and it's a reminder that the “buy the balance sheet” side of this trade is being placed by more than one company, in more than one market, in the same week this article went live.
Aulay's point of view
Before reacting to embedded finance as a threat, a financial institution should map its own products against these three layers and be honest about which one it still actually owns, product by product. A bank that has quietly become a balance-sheet-only supplier behind someone else's brand is not necessarily in a bad position — the Tesco deal is a scale business for Barclays, not a mistake — but it is a different business than the one most banking strategy still assumes, and it needs a different plan for where growth comes from next. If your organisation is trying to work out which of these three layers it actually competes on, and what that means for where to invest next, that's a conversation worth having with Aulay.
References
1. McKinsey & Company, “Embedded finance: How banks and customer platforms are converging,” published 15 July 2024. https://www.mckinsey.com/industries/financial-services/our-insights/embedded-finance-how-banks-and-customer-platforms-are-converging
2. Barclays, “Completion of the acquisition of Tesco's retail banking business and commencement of long-term strategic partnership,” press release, 1 November 2024.
3. About Amazon UK, “Amazon launches flexible financing for small and medium-sized businesses,” 20 June 2023.
4. Ikano Group newsroom, “Strengthening the IKEA financial services offer: Ingka Investments to acquire full ownership of Ikano Bank,” 17 October 2024.
5. Chime Financial, Inc., investor relations, “Chime Announces Agreement to Acquire Stride Bank,” press release, 8 September 2026.
Frequently asked questions
- What is embedded finance?
- Embedded finance is a financial product — a loan, a card, an instalment plan, an insurance policy — offered inside a non-financial company's existing product or checkout, rather than through a standalone bank or insurer. McKinsey estimates it generated €20 billion to €30 billion in revenue across Europe in 2023.
- Who is actually liable if an embedded loan or credit product defaults?
- It depends on who holds the balance sheet, which is not always the company the customer is dealing with. In Amazon's UK seller-financing product, for example, the credit risk sits with YouLend, the licensed lender behind it, not with Amazon.
- Should a bank sell off its balance sheet and focus on distribution, or do the opposite?
- Neither, universally. Tesco concluded its balance sheet wasn't worth running itself and sold it to Barclays; IKEA's parent concluded the opposite and bought full ownership of Ikano Bank in the same year. The right answer depends on the specific book's economics, not a general rule.
- What should a bank do differently because of this?
- Map existing and planned products against the three layers — customer relationship, embedded product, balance sheet — and be explicit about which ones the bank actually owns for each. Losing the balance sheet on a product line on purpose, as Tesco did, is a defensible strategic choice; losing the customer relationship by accident while still carrying the balance-sheet risk is not.