The CEO's guide to building a fintech business
For banks, non-banks and high-growth firms : start with a customer vision, two cofounders and the next ten hires.
On this page
A new fintech business built by an established company should start with a customer vision and the people who will own it, not with a technology programme.
Licensing, risk and capital still need early attention, but they serve the vision rather than define it. Whether the parent is a bank, an accounting platform or a high-growth firm, the early decisions that matter most are who the business is for, who will build it and how it will win its first customers. The platform can come later.
This guide follows that journey in order:
How an enterprise arrives at a vision
How it chooses the cofounders and the next ten people
Why marketing and sales should lead in the first months, and
How the other workstreams (proposition, economics, regulation, technology, governance and scale) organise themselves around that core team.
It uses one vision throughout as a case study, a drill-down into a single segment: micro businesses with fewer than ten employees. They are the vast majority of businesses. Around 95% of the UK's 5.7 million private sector businesses were micro businesses at the start of 2025, and more than nine in ten enterprises in the EU employ fewer than ten people.
Worldwide, the self-employed and micro enterprises together account for about 62% of employment across the 99 countries studied by the International Labour Organization in 2019.
Start with a vision that ignores what you already sell
The vision used in this article is simple to state.
A micro business should be able to run its entire back office in one app
Registering the business, opening a business account, taking and making payments, handling foreign exchange, invoicing, bookkeeping, payroll, business insurance, VAT and income tax returns, and eventually closing the business down are the processes many business owners under-estimate.
The pain behind it is well documented. According to research by American Express and Small Business Saturday UK, reported in trade press in July 2026 (we have not been able to locate the original release), a survey of 1,000 owners of UK micro, small and medium-sized businesses found they spend an average of 11 hours a week on administrative or finance-related tasks, which the researchers equate to about six working days a month, against 3.6 days a month on sales and business development. At 11 hours a week, that would amount to more than 60 eight-hour days over a 44-week working year; that is our calculation, not a finding of the survey.
The obligations are also growing. Since 6 April 2026, sole traders and landlords with qualifying income above £50,000 have had to use Making Tax Digital for Income Tax, and the threshold falls to £30,000 in April 2027 and £20,000 in April 2028.
What makes this a useful vision for an established company is that it has to be written with total abstraction from the company's current products. A bank asked to design for this customer tends to start from the business current account; an accounting platform starts from bookkeeping; a payroll provider starts from payroll.
The customer doesn't think in any of those categories. They think about getting their admin done.
That is why the vision should neither overlap nor underlap with what the enterprise already sells. Overlap means the new business mostly repackages existing products, which creates internal competition for the same customers and blurs ownership. Underlap is subtler: the customer's real pain often sits in the gaps between existing products (the tax return that depends on the bookkeeping that depends on the bank feed), and nobody in the parent owns those gaps. A good vision is defined by the customer's whole job, so that it naturally fills the gaps rather than duplicating the catalogue.
The market for this vision shows companies converging on it from three directions. Fintechs got there first. Tide, founded in 2015 and launched in 2017, reported two million members worldwide in May 2026, including 900,000 in the UK, and now offers business registration, accounting, payroll, insurance and Making Tax Digital tools alongside its account. Qonto, launched in France in 2017, reported 600,000 customers across eight European countries in July 2025, offers online company creation, invoicing and bookkeeping, and applied for a French banking licence in 2025. Banks followed: NatWest bought the accounting software company FreeAgent for £53 million in 2018 and includes it free with Mettle, its app-based account for sole traders and small companies, while Crédit Agricole built Blank, an all-in-one app for independent professionals, through its startup studio. Accounting platforms are moving towards banking too. Countingup, which combines a business account with bookkeeping and tax estimates, was founded by the former chief executive of Clear Books accounting software and counts Sage among its investors. Ageras, a Danish accounting and admin platform, bought the German freelancer account provider Kontist in 2022. In France, the freelancer accounting app Indy launched a free business account in January 2024, built on the banking-as-a-service provider Swan.
We are not aware of a provider that yet covers the whole list in one app, from registration to closure. That gap, rather than any individual product, is the opportunity the vision describes.
Crédit Agricole shows what happens when a group's own ventures overlap. Its startup studio built Blank, an all-in-one app for independent professionals. In October 2022 the group's regional banks launched Propulse by CA, a digital account for micro-entrepreneurs and the newly self-employed, built in partnership with Blank. In 2023 the studio unveiled Kolecto, an invoicing and cash-management platform for small businesses that has since become an approved e-invoicing platform. Three businesses, each defensible on its own, were serving closely related customers from different angles. Then regulation forced the issue. Under France's e-invoicing reform, every business must be able to receive electronic invoices from 1 September 2026, and small and micro-enterprises must issue them from 1 September 2027. In June 2026 Crédit Agricole launched Pro by CA, a single range for professionals with Kolecto's invoicing built into every tier. According to trade press, Blank and Propulse are being closed, with customers moved to Pro by CA by January 2027 at the latest, and the group has described the aim as "une gamme unique, plus simple et plus lisible" (a single, simpler, clearer range). Public sources describe this as a consolidation of offers rather than a legal merger of the companies. Our reading is that it is a lesson in both directions. Testing several routes to the same customer can be a sensible use of a studio, but the customer never wanted three apps, and the one anchored in a regulatory obligation (invoicing) became the platform for the rest.
Choose two cofounders before you choose anything else
Once the vision is clear, the most consequential decision is who will lead the business. We'd argue for two cofounders with complementary strengths: one who owns the customer, the proposition and the commercial engine, and one who owns the product and how it gets built. Both should have what the investor Chris Dixon called founder/market fit in 2011: a deep understanding of the market they are entering, and the kind of people who personify their product and company. In a venture built by an established company, we call this cofounder-market fit, because the fit has to hold for the pair, not just for one leader.
The founders of the businesses above illustrate what that pedigree looks like. George Bevis founded Tide after working in banking and fintech roles at Capital One, Barclaycard, Zopa, Worldpay and RBS, and after launching earlier ventures of his own. In an interview he described the problem in the customer's terms: "Banking and admin tend to soak up a huge amount of time for business owners and that's something they rightly loathe." Qonto's cofounders, Alexandre Prot and Steve Anavi, were school friends who had already built and sold a company, Smok.io, acquired in 2015. Their frustration with their own bank as founders led them to interview hundreds of entrepreneurs before launching. Tim Fouracre had run Clear Books, an accounting software company, before founding Countingup in 2017 to tackle the paperwork he saw self-employed people buried in. In each case the founders had lived the problem, understood the customer and had built a business before.
Established companies choose cofounders in different ways, and each route has trade-offs. Crédit Agricole's La Fabrique by CA works as a studio. It identifies ideas, tests their market potential, and then selects project leaders to run them, on the condition that the business fits the group's strategy. Its chief executives don't own equity; the group keeps full ownership and pays them a salary with performance bonuses. Blank came out of this model. Paul-Henri Blaiset, who had founded and run an online business before joining La Fabrique as an entrepreneur in residence in 2018, co-founded Blank and led it as chief executive until early 2022. His cofounder, Simon Parisot, an engineer who had helped set up La Fabrique, later became chief executive. In 2023, according to trade press at the time, Blank had more than 20,000 accounts and about 60 staff, and Crédit Agricole put €47 million into it to expand into Italy. Three years later, as described above, the group consolidated its offers for professionals into Pro by CA, a reminder that a studio's ventures also have to be managed as a portfolio.
NatWest took a different route, buying a proven team and product in FreeAgent and pairing it with a bank-built proposition in Mettle. Ageras bought Kontist, which brought an existing founding team and around 50,000 German customers. Our reading is that acquisition buys cofounder-market fit ready-made, at a price, but it also imports someone else's vision, which then has to be reconciled with the customer-first one. The studio route keeps the vision clean but depends heavily on the quality of the people it can attract without offering equity. Neither route is right in every case; the decision should be made deliberately rather than by default.
The next ten hires
The cofounders' first job is to hire the next ten people, and those hires set the culture and the speed of everything that follows. In our view most of them should face the customer: a growth marketer who can run paid acquisition and A/B tests, someone who owns content and community, a partnerships lead for accountants and other intermediaries who already advise micro businesses, and customer operations people who treat every support conversation as research. Because the business will handle other people's money, a compliance lead with money-laundering reporting experience belongs in the first ten, not the second fifty. The remainder can be a product designer and two or three engineers who are comfortable building on partners' infrastructure.
In the examples above, headcount grew quickly after that. Qonto had more than 60 employees by February 2018, about seven months after launch. Countingup planned to grow from 30 to 80 people after its 2021 funding round. Blank relied on La Fabrique's shared roles in finance, legal, marketing, technology, risk and HR[1], which let a small core team stay focused on customers. That is one of the genuine advantages an established parent can offer: shared services that remove work from the venture, rather than processes that add it.
Customers first, marketing and sales before technology
Our view is that in the first months, marketing and sales should take priority over every other workstream, including technology. The reasoning is straightforward. Building a first version of a fintech product has become far cheaper, because partners now provide accounts, cards and payments, and AI[2] has cut the cost of building software. What remains expensive and uncertain is finding out whether customers will come, which message brings them, and what it costs to acquire and keep them. Those are marketing questions, and they are answered by testing, not by building.
The examples suggest the founders above understood this, although they do not prove that marketing-first is always right. Qonto interviewed hundreds of entrepreneurs before launch and signed up more than 5,000 business customers within six months of its July 2017 launch. Blank's team spent months co-designing the service with its first customers before opening it to all independent workers. Tide gained more than 30,000 members in its first year after launching in 2017, yet George Bevis identified awareness, not technology, as the main obstacle, because many small businesses simply didn't know the alternatives existed.
A/B testing is the discipline that turns this into evidence. Before the product exists, a team can test which promise brings more sign-ups to a landing page ("all your admin in one app" against "never miss a tax deadline", for example), which price point converts, whether accountants or direct channels bring cheaper customers, and how many onboarding steps people will tolerate. Two cautions apply. Tests need enough traffic to mean anything, so early tests should compare big differences rather than small ones. And a test measures what people click, not what they keep using, so it has to be followed by retention data once real customers arrive.
Technology, in our view, is almost an afterthought at this stage. Indy launched its business account on Swan's infrastructure; Tide's first accounts were provided through PrePay Solutions, an e-money issuer; Countingup's account is provided through Prepay Technologies. Owning the platform can come later, when scale and economics justify it. In our view, building it first, before the customer and the message are proven, is the most expensive way to find out that nobody wanted it.
Regulation: know what an e-money licence does and doesn't allow
The regulatory route shapes the proposition, so it belongs in the early decisions, even if the first product launches on a partner's licence. Many fintech accounts for micro businesses are e-money products rather than bank deposits. In practice that covers a business account with its own sort code and account number or IBAN, a debit or prepaid card linked to it, domestic and international payments, and currency conversion for paying suppliers abroad. Countingup's account, for example, is e-money provided through Prepay Technologies, with customer funds held in a safeguarding account.
The limits matter. Under UK rules an e-money institution must safeguard customer funds, keeping them separate so they can be returned if the firm fails. It cannot pay interest on e-money under the Electronic Money Regulations 2011, and money held directly with an e-money institution is not covered by FSCS deposit protection if the e-money institution itself fails (the FSCS says it "can't protect the money you have with e-money institutions and payment providers"). The absence of direct FSCS deposit protection remains true even when the account comes with a sort code, a debit card and BIN sponsorship (the arrangement under which a card scheme member lets a fintech issue cards under its bank identification number). BIN sponsorship governs who issues the card, not where the balance sits: FSCS protection depends on whether the firm holding the funds is a licensed bank. Since 2023, PRA rules have given one partial safeguard: if the bank holding an e-money firm's safeguarded funds fails, eligible customers can be compensated for their share, but not if the e-money firm itself fails. A banking licence removes those limits but brings far heavier capital and supervisory requirements. The difference shows up in the market: Mettle is provided by National Westminster Bank, so eligible deposits are protected by the FSCS up to £120,000 and it can offer an interest-bearing savings pot, while Qonto, a licensed payment institution since 2018, applied for a French banking licence in 2025 as its ambitions grew. For most new ventures of this kind, we think the practical choice is to start with a partner or an e-money licence and decide on a banking licence once the customer case for deposits, interest or lending is proven.
The other workstreams sit around the core team
With the vision, the cofounders and the first ten people in place, the remaining workstreams become questions the core team owns, answered in roughly this order: the customer and the problem (tested continuously), the proposition and distribution (tested through marketing), the economics (what it costs to acquire, serve and keep a customer), regulation (as above), technology (partner first, own later), organisation (which of the parent's resources the venture may use and which of its processes it is exempt from), governance (a small board, evidence-based stage gates and stopping conditions agreed before launch) and scale.
Scale is where an established parent's backing becomes decisive. Blank's leadership has described having an industrial shareholder, rather than venture capital investors, as protection in a difficult funding market. Tide, which has raised from outside investors, now serves more members in India than in the UK. Qonto reports that it has been profitable since 2023. There is more than one way to reach scale, but each depends on the early work having produced customers who stay.
What Aulay thinks CEOs should do
Write the vision in the customer's words and check it against your current catalogue only to confirm it neither repackages what you sell nor leaves the gaps unowned. Choose the cofounders on evidence of cofounder-market fit, decide deliberately whether to build, run a studio or acquire, and agree their incentives before they start. Make the first ten hires mostly customer-facing. Spend the first months proving the message, price and channel through marketing and A/B testing, on partners' infrastructure, and treat your own technology platform as a later decision.
If you lead a bank, an accounting or payroll platform, a PE portfolio company or a high-growth firm in financial management and are weighing a new fintech business, contact Aulay at aulayco.com to test your vision, cofounder choice and first-hire plan before you commit capital.
References
1. House of Commons Library, "Business statistics", research briefing SN06152, 3 December 2025, based on the Department for Business and Trade's Business Population Estimates 2025. Supports: 5.7 million UK private sector businesses at 1 January 2025; about 95% micro businesses (0 to 9 employees).
https://commonslibrary.parliament.uk/research-briefings/sn06152/
2. Eurostat, "Structural business statistics overview", Statistics Explained, section "Size class analysis" (2023 data). Supports: more than nine out of ten EU enterprises are micro enterprises employing fewer than ten persons.
3. International Labour Organization, "Small Matters: Global evidence on the contribution to employment by the self-employed, micro-enterprises and SMEs", 2019, Table I.1, p. 29. Supports: self-employed (35.2%) and micro enterprises of 2 to 9 employees (26.4%) together about 62% of employment across 99 countries.
4. American Express and Small Business Saturday UK, SME Business Barometer research, reported by ChannelX, "SMEs spend twice the time on admin as growing their business", 20 July 2026. Secondary source; no primary release located. Supports: 1,000 UK SME owners surveyed; 11 hours a week on admin or finance tasks, about six working days a month; 3.6 days a month on sales and business development. The 60-day annual figure is Aulay's calculation (11 hours × 44 weeks ÷ 8 hours).
https://channelx.world/2026/07/smes-spend-twice-the-time-on-admin-as-growing-their-business/
5. HM Revenue & Customs, "Check if you're eligible for Making Tax Digital for Income Tax", GOV.UK guidance, last updated 26 March 2026, section on qualifying income. Supports: £50,000 threshold from 6 April 2026, £30,000 from 6 April 2027, £20,000 from 6 April 2028.
https://www.gov.uk/guidance/check-if-youre-eligible-for-making-tax-digital-for-income-tax
6. Tide, "Tide Crosses 2 Million Members Worldwide", press release via Business Wire, 26 May 2026. Supports: two million members, 900,000 in the UK, 1.1 million in India; products including business registration, accounting, payroll, insurance and Making Tax Digital tools.
7. Companies MadeSimple, "Startup Story with George Bevis, founder of Tide", interview. Secondary source. Supports: George Bevis's prior roles; quotation on banking and admin; 30,000+ members in the first year; awareness as the main challenge; early account provider PPS (PrePay Solutions).
https://www.companiesmadesimple.com/blogs/company-finance/interview-george-bevis-founder-tide
8. Tide, "1 in 10 UK Small and Mid Sized Businesses on Tide", press release via Business Wire, 19 September 2023, section "About Tide". Supports: founded in 2015 and launched in 2017.
9. TechCrunch, "French B2B fintech Qonto reaches 600,000 customers, files for banking license", 2 July 2025. Secondary source. Supports: 600,000 customers; eight countries; payment institution licence since 2018; French banking licence application; profitability since 2023; bookkeeping and invoicing products.
10. EU-Startups, "Making business banking great again: Interview with Alexandre Prot, Co-founder & CEO of Qonto", February 2018. Secondary source (interview). Supports: school friends; Smok.io acquired in 2015; hundreds of entrepreneur interviews; July 2017 launch; 5,000+ customers in six months; 60+ employees.
11. Qonto, "Fast-track company creation for SMEs", product page. Supports: online company creation and capital deposit.
12. Countingup, "About us", company website. Supports: founded 2017 by Tim Fouracre, former CEO of Clear Books; founding problem; investors including Sage and ING.
https://countingup.com/about-us/
13. TechCrunch, "Countingup closes £9.1M for its business current account with built-in accounting features", 4 March 2021. Secondary source. Supports: Series A with Sage participation; plan to grow from 30 to 80 employees.
https://techcrunch.com/2021/03/04/countingup-series-a/
14. Finder UK, "Countingup business account review". Secondary source. Supports: e-money account provided through Prepay Technologies, FCA-authorised; funds safeguarded at Barclays; no FSCS protection; pricing tiers.
https://www.finder.com/uk/business-banking/countingup-review
15. UKTN, "Accounting software firm FreeAgent acquired by RBS for £53m", 27 March 2018. Secondary source. Supports: acquisition price and date.
16. NatWest, "Mettle mobile account", product page, legal information. Supports: provided by National Westminster Bank plc trading as Mettle; FSCS protection up to £120,000; FreeAgent included; savings pot.
https://www.natwest.com/business/bank-accounts/mettle-mobile-account.html
17. Crédit Agricole, "Blank, l'app tout-en-un des professionnels indépendants, en conquête clientèle", group news, 10 February 2021. Supports: built by La Fabrique by CA with six regional banks; product scope; Paul-Henri Blaiset as CEO.
18. mind Fintech, "Paul-Henri Blaiset", people profile. Secondary source. Supports: prior founding of Pixopolitan.com; entrepreneur in residence at La Fabrique by CA (2018 to 2019); co-founder and CEO of Blank (2018 to January 2022).
https://www.mind.eu.com/fintech/data/personnes/paul-henri-blaiset/
19. Blank, "L'interview de Simon Parisot, CEO de Blank", company blog, updated 16 January 2025. Supports: engineering background; role in setting up La Fabrique; co-construction with first customers.
https://www.blank.app/blog/interview-simon-parisot
20. L'Usine Digitale, "Blank, la néobanque pour les pros soutenue par Crédit Agricole, lève 47 millions d'euros", 2023. Secondary source. Supports: €47 million from the Crédit Agricole group; 20,000+ accounts; about 60 employees; Italy expansion; Simon Parisot as CEO; quotation on industrial shareholder.
21. Crédit Agricole, "Le Crédit Agricole déploie Propulse by CA et LCL Essentiel Pro et se positionne comme un acteur majeur du marché des néo-pros", press release, 24 October 2022. Supports: Propulse by CA launched by regional banks with Blank; target micro-entrepreneurs.
22. mind Fintech, "La Fabrique by CA dévoile sa nouvelle start-up, Kolecto", 21 March 2023. Secondary source. Supports: Kolecto unveiled by La Fabrique by CA; invoicing and cash-management for small businesses.
https://www.mind.eu.com/fintech/article/la-fabrique-by-ca-devoile-sa-nouvelle-start-up-kolecto/
23. Crédit Agricole, "Le Crédit Agricole lance Pro by CA, une nouvelle gamme d'offres de banque au quotidien qui réinvente l'expérience des entrepreneurs", press release, 9 June 2026. Supports: Pro by CA launch; Kolecto e-invoicing embedded in all tiers; three pricing tiers; ambition of 150,000 new professional clients a year.
24. MoneyVox, "Le Crédit Agricole enterre Blank, sa banque en ligne pour les pro", 9 July 2026. Secondary source (trade press). Supports: Blank and Propulse to close; customers migrated to Pro by CA by January 2027; group quotation on a single, simpler range.
25. Direction générale des Finances publiques, "I want to understand electronic invoicing", impots.gouv.fr. Supports: all companies receive e-invoices from 1 September 2026; small and micro-enterprises issue from 1 September 2027.
https://www.impots.gouv.fr/internationalenbusiness/i-want-understand-electronic-invoicing
26. Maddyness, "Comment fonctionne La Fabrique by CA, le startup studio fintech du Crédit Agricole ?", 14 February 2024. Secondary source. Supports: selection of ideas and project leaders; strategy alignment rule; no equity for CEOs, salary plus bonuses; about fifteen shared roles.
27. Ageras, "Ageras Pushes Into Europe's Largest Market, Buying German Neo-Bank Kontist", press release via Business Wire, 27 July 2022. Supports: Ageras as accounting and admin platform; Kontist's ~50,000 customers; rationale of combining accounting with financial services.
28. mind Fintech, "Comment Indy concurrence les néobanques pour les pros", 20 May 2025, updated 17 April 2026. Secondary source. Supports: Indy founded 2016 as Georges.tech; free business account launched January 2024 on Swan's infrastructure.
https://www.mind.eu.com/fintech/article/comment-indy-concurrence-les-neobanques-pour-les-pros/
29. Chris Dixon, "Founder/market fit", cdixon.org, 19 June 2011. Supports: definition of founder/market fit.
https://cdixon.org/2011/06/19/foundermarket-fit/
30. The Electronic Money Regulations 2011 (SI 2011/99), regulation 45, "Prohibition of interest". Supports: e-money issuers must not award interest.
https://www.legislation.gov.uk/uksi/2011/99/regulation/45
31. Financial Conduct Authority, "Safeguarding requirements for payment institutions and electronic (e-money) institutions", last updated 7 May 2026. Supports: obligation to protect customer funds in the event of insolvency by segregation or insurance.
https://www.fca.org.uk/firms/emi-payment-institutions-safeguarding-requirements
32. Osborne Clarke, "UK's PRA extends deposit protection to e-money and authorised payment institutions' safeguarded funds", 2023. Secondary source (law firm commentary) on PRA policy statement PS2/23. Supports: FSCS protection for end customers' share of safeguarded funds only if the safeguarding bank fails, not if the e-money firm fails.
33. Financial Services Compensation Scheme, "Pots, pockets, piggy banks and vaults: keeping track of your money (or e-money), and its FSCS protection", 11 May 2023. Supports: FSCS cannot protect money held with e-money institutions and payment providers.
https://www.fscs.org.uk/news/protection/e-money-and-fscs-protection/
Frequently asked questions
- What is cofounder-market fit?
- Cofounder-market fit is our extension of the term founder/market fit, which investor Chris Dixon used in 2011 to describe founders with a deep understanding of their market who personify their product and company. For a fintech business built by an established company, it means the two cofounders together understand the target customer, have lived the problem and have built a business before.
- What is the difference between an e-money institution and a bank?
- In the UK an e-money institution can offer accounts, cards, payments and currency conversion, but it must safeguard customer funds, cannot pay interest on e-money, and its customers have no FSCS deposit protection if the institution itself fails (eligible safeguarded funds may be protected if the safeguarding bank fails). A bank can take deposits, pay interest and lend, and eligible deposits are protected by the FSCS, but it faces much heavier capital and supervisory requirements.
- Why should marketing come before technology when building a fintech business?
- Partners and AI have made a first version of a fintech product relatively cheap to build, while demand, messaging and customer acquisition costs remain uncertain. Testing those through marketing and A/B tests before investing in a proprietary platform reduces the risk of building something customers do not want.