Insights
Fourteen fintech UX design agencies grouped by product, judged on whether their published work shows what happens after a customer commits. Choose a fintech design agency by the kind of product you run, and by whether its published work shows what happens after a customer commits: a payment that is pending, failed or r
Choose a fintech design agency by the kind of product you run, and by whether its published work shows what happens after a customer commits: a payment that is pending, failed or reversed, a loan waiting for a document, a payout that is not yet final. Fourteen firms show that kind of work. We sort them by the situation each suits, A to Z inside each situation, and rank none above another:
- Consumer banking, wallets and payments: Blink UX, Clay, Lollypop, Method, Phenomenon Studio and UXDA.
- Lending and credit: Eleken, Feelpixel and NetBramha.
- Insurtech and operations-heavy finance: Excited, Tcules, UX studio and Work & Co.
- Crypto and tokenised assets: Groto.
Disclosure. This page is ours, and we are one of the fourteen firms on it, in the insurtech and operations-heavy finance group.
How we chose: five checks you can run on any fintech case
We started from the sixteen firms Google, ChatGPT and Gemini named as the best fintech UX design agencies, added ourselves, and listed every firm with at least one fintech product case we could open and read. Work a firm walks you through under a non-disclosure agreement counts as much; we could check only published work, so ask each firm for the rest. Then we applied the same five checks to every firm, ourselves included.
| Check | What to look for | Why it matters |
|---|---|---|
| 1. A state model, not only screens | Pending, failed, blocked, provisional or closed, and what each status allows | A spinner is not a transaction model |
| 2. Roles and authority | Applicant, operator, approver and admin, and what each may do | A hand-off between roles is where an action can pass unchecked |
| 3. Exceptions and recovery | Disputes, holds, failed payments and rework, not only the happy path | Recovery is where trust is won or lost |
| 4. Where the work stops | Design only, design plus front end, or the full build, and who owned the code | Someone has to close the gap between the design and what ships |
| 5. Stated boundaries | What the firm says it did not own or did not measure | An honest boundary counts in a firm's favour |
Regulators treat some of these states as obligations. Under US Regulation E, a bank that needs more than 10 business days to investigate a reported transfer error must generally credit the account provisionally and tell the customer within two business days, so "provisionally credited" is a state the product has to show.
The checks do not produce a rank, and we compared no prices or review ratings. Every outcome figure below is the firm's own, and none of the pages we opened says how it was measured. The AI answers also named MetaLab, Superside and Baking UX; we could not find a readable fintech product case for any of them, which is not a judgement of their quality. The design method behind the checks is in our guide to fintech UX: designing for trust, compliance and speed.
The shortlist at a glance
| Firm | Best for | Case we opened | States, roles or exceptions shown | Builds as well as designs? | Figures and limits |
|---|---|---|---|---|---|
| Blink UX | Banks and credit unions | Gesa Credit Union: account opening and loan applications | Journey maps and screens; few states described | Not stated in this case | Blink reports abandonment fell from 89% to 39% |
| Clay | Banks and wallets redesigning brand and product together | STC Bank: consumer app, merchant portal and merchant app | Structure and dashboards; states not described | Lists front-end and back-end development | No figures |
| Lollypop | Payments and credit apps for mass-market customers | IME Pay; Bonify | Navigation by user type; offline transaction and error states; consent steps explained | Describes itself as research, design and development; the IME Pay case lists UX and UI design | No figures |
| Method | A new digital bank, from strategy to production software | Jenius Bank | Strategy, research set-up and build; states not described | Yes: proof of concept and production-ready software | No figures |
| Phenomenon Studio | Payments and money transfer | KlickEx: cross-border transfers | Fee and rate calculator with final amounts; a proposed status timeline | Yes: web development in Next.js and TypeScript | Phenomenon reports conversion gains in two flows |
| UXDA | Banks and large financial institutions | ITTI: core banking back office | Separate architecture for four employee roles | Design | UXDA reports learning time cut from weeks to hours |
| Eleken | Early-stage fintech teams with their own developers | PayUp; Alture Funds | Payout terms per invoice; sign-up that explains sensitive data | Says it does not develop software; offers product design engineering | Alture Funds case says not every proposal was built |
| Feelpixel | Lenders and insurers in India | Lendingkart; TATA AIG | Paid, pending and next-due instalments; approvals with roles and audit trail | Homepage lists design services; the cases do not say who built | No figures for the redesigns |
| NetBramha | Consumer credit in India | TransUnion CIBIL; ONE Muthoot | Dispute flows with status tracking | Neither case says clearly who wrote the code | NetBramha reports figures for the CIBIL work |
| Excited | Insurance agency software | InsuredMine | Product scope; states not described | Lists "iterative design/development"; the case does not say who built | No figures |
| Tcules (publisher) | Insurance, lending and trading software that moves money | Chestnut; Phoenix | Payouts marked Needs Recalculation or Blocked; publishing blocked on a mismatch; estimates that state what they leave out | Built part of the Chestnut front end and reviewed built features against the designs; did not own the Phoenix code | No usage or business results claimed |
| UX studio | Insurance, compliance and B2B finance software | Targens; Oversee.ai | Partner onboarding, risk set-up and approvals; anomaly review | Built Oversee.ai with the client's developers | Its Targens testimonial says compliance was new to the studio |
| Work & Co | Direct-to-consumer insurance | Hippo: enrolment and claims | Claims flow with a reimbursement timeline | Strategy, design and prototyping, not build | Work & Co reports twice the industry-average conversion and a 60-second quote |
| Groto | Tokenised assets and crypto | STBL.io | Buy, sell, convert and redeem; admin vault approvals | Design, with development-ready documentation | No figures |
Consumer banking, wallets and payments
For when customers move their own money through your app, and balances, fees and transfers must stay clear at scale.
Blink UX
Best for: banks and credit unions improving account opening and lending applications.
Blink, part of Mphasis, describes itself as a product design consulting firm. Its Gesa Credit Union case covers six months of workshops, competitive analysis, member profiles and journey maps of account opening and loan applications, then a redesigned banking platform and a design system. Blink reports that new-member application abandonment fell from 89% to 39%, and that Gesa rose to 23rd in a 2023 credit union digital experience ranking.
- Check before you hire: ask to see the application states the journey maps found (incomplete, pending review, declined) and how each was resolved.
Clay
Best for: larger banks and wallets redesigning brand, app and merchant tools together.
Clay's STC Bank case covers a Saudi digital bank that grew out of a wallet: a dashboard rebuilt around primary actions, one structure for banking, payments and lifestyle services, and a merchant portal and app that bring analytics, transactions and account tools into one view. Its services page lists front-end and back-end development alongside branding and product design.
- Check before you hire: ask how the merchant tools handle a failed or disputed transaction; the case says little about states.
Lollypop
Best for: payments, wallet and credit apps for mass-market customers, including people new to digital finance.
Lollypop, a Terralogic company, works across research, design and development. Its IME Pay case covers three payments apps in Nepal, for customers, merchants and agents, serving many unbanked and underbanked people: navigation and the home page change by user type, and transaction updates and errors are designed for when the app is offline. Its Bonify case rewrote a German credit-score platform's consent and legal steps to say why each matters.
- Check before you hire: ask to see the offline and error states in detail, and who built the IME Pay apps.
Method
Best for: a bank launching a new digital business that wants strategy, design and engineering from one firm.
Method, a GlobalLogic company, joined Jenius Bank, the US digital bank set up by SMBC MANUBANK, at the incubation stage. It shaped product and experience strategy, set up the bank's research tools and research repository, and helped its technology leaders design and build the proof of concept, the production-ready software and a storefront website.
- Check before you hire: the case describes scope rather than states; ask how an account application, a declined payment or a dispute is handled.
Phenomenon Studio
Best for: payments and money-transfer products where design and build should come from one firm.
Phenomenon's KlickEx case covers cross-border payments for Pacific Island communities: a UX audit, a redesigned transfer flow, verification across several jurisdictions, and a calculator that shows rates, fees and the final amount in both currencies before sending. For transaction history it proposed a status timeline instead of the detailed technical view the client preferred; the case does not say which shipped. It lists web development in Next.js and TypeScript, and Phenomenon reports gains of 35.3% in the "Add Money" flow and 30.7% in "Money Transfer".
- Check before you hire: ask how those gains were measured, and what a customer sees when a transfer fails or is held for review.
UXDA
Best for: banks and large financial institutions with many products and internal users.
UXDA works only on financial services. Its ITTI case moved a core banking back office to a cloud product designed around bank staff: interviews with 21 users, a journey map from a two-week workshop, and four employee personas. Because each role had different security access, UXDA designed one overall architecture plus one for each persona. It reports that learning time fell from several weeks to a few hours.
- Check before you hire: its published work centres on large institutions, so an early-stage fintech should ask how it scopes a smaller engagement.
Lending and credit
For when the difficulty sits in offers, repayments, documents and disputes, and the customer has to understand an obligation that lasts.
Eleken
Best for: early-stage fintech teams with their own developers who want design capacity quickly.
Eleken's PayUp case covers an invoice-factoring app that pays service workers early: a flow for changing payout terms on one invoice, a sign-up in small steps that explains why sensitive details are needed, and a design system annotated for developers. Its Alture Funds case, an investment platform whose web and iOS MVP was due in two months, says plainly that the backend could not change and that not every proposed improvement made the build.
- Check before you hire: its fintech page says it does not develop software, while it separately offers product design engineering, so agree who owns implementation.
Feelpixel
Best for: lenders and insurers in India digitising internal and customer workflows.
Feelpixel's Lendingkart case covers loan tracking for small-business borrowers: paid, pending and next-due instalments, active and closed loans, a document store, and support tickets whose status the borrower can follow. Its TATA AIG case moved commission rules for insurance producers out of spreadsheets into a tool with rule creation, approval workflows, role-based access, approver notifications and audit trails.
- Check before you hire: ask what happens when an approver rejects a rule or an instalment fails, and who built the products.
NetBramha
Best for: consumer credit and lending products serving large audiences in India.
NetBramha's TransUnion CIBIL case covers the credit bureau's mobile and web experience, including dispute management with status tracking, forms and a resolution journey. Its ONE Muthoot case merged a gold-loan company's separate portals for loans, cards and insurance into one app, and names what made it hard: each system had its own data model and edge cases. NetBramha reports outcome figures for the CIBIL work.
- Check before you hire: neither case says clearly who wrote the production code, so confirm who will implement, and ask to see the dispute states.
Insurtech and operations-heavy finance
For when operators, agents or reviewers inside the system carry the risk and one action reaches further than it looks, or when an insurance product must stay clear from quote to claim.
Excited
Best for: insurance agency software that combines a back-office tool, a customer app and a marketing site.
Excited's InsuredMine case covers a platform that brings insurance agencies' sales, customer messages and analytics into one product: email and text campaigns in one automation, a white-label app in which policyholders manage policies, get quotes, contact agents and report incidents, a design system and a new marketing site.
- Check before you hire: the case describes scope rather than states; ask how an incident report or a failed policy change moves between policyholder and agent.
Tcules (publisher of this page)
Best for: insurance, lending and trading software that moves money, where every figure must say whether it is final.
For about thirteen months we worked inside Chestnut, a producer performance platform that carriers use to calculate and pay what their agents earn: each payout says whether it needs recalculation or is blocked, and a manual override needs a reason and leaves an audit trail. We also coded the bonus designs and part of the communications front end, and checked what was built against the designs as each feature landed. The case reports no usage or business results. Our fintech and insurtech page shows how we approach these products.
- Check before you hire: our public fintech cases are few and include no consumer banking app, so ask to see work under NDA and to speak to a client.
UX studio
Best for: insurance, compliance and B2B finance software with dense data and several roles.
UX studio's Targens case covers a year designing compliance software for checking business partners: onboarding, risk evaluation and risk model set-up, analytics and approvals, with workflow diagrams and biweekly meetings with an external development team. A testimonial on the page says compliance was new to the studio at the start. Its Oversee.ai case covers a platform for insurers monitoring the AI models in underwriting and claims, built from scratch with the client's developers. It also shows Duck Creek insurance software work.
- Check before you hire: ask who from these engagements would join your project, and how much of the build the studio would take on.
Work & Co
Best for: direct-to-consumer insurance where the quote, the purchase and the claim all have to feel simple.
Work & Co's Hippo case covers a home insurer whose founder set out to remake home insurance in 2015, and in which Work & Co took an equity stake. The work put enrolment first, with a form that prefills public data about the home to produce a quote, clear coverage descriptions and transparent pricing; it also included a claims flow and dashboard with a reimbursement timeline, app concepts and a rebrand. Work & Co reports twice the industry-average sales conversion, a 60-second quote and a Net Promoter Score of 75, and lists strategy, design and prototyping rather than build.
- Check before you hire: the work dates from Hippo's early years, so ask for recent financial work, what a customer sees when a claim is disputed or part-paid, and how the equity stake shaped the engagement.
Crypto, trading and investing
For when orders, positions and custody must make sense to newcomers and experienced traders alike. Firms in other groups have trading work too: Eleken (Alture Funds), Lollypop (PocketBits), UX studio (Zignaly) and our own work on Phoenix, a decentralised exchange, where we prototyped two trading models to the same depth so that testing could choose between them.
Groto
Best for: new tokenised-asset and crypto products built from scratch.
Groto's STBL.io case covers a platform for gold- and silver-backed tokens that previously ran offline: customers buy, sell, convert, redeem and trade tokens, admins manage vaults and approvals, and super-admins oversee system controls. The work ran from discovery workshops and personas through information architecture, high-fidelity design and Maze testing.
- Check before you hire: ask what a customer sees when a redemption or vault approval is pending or refused, and who built the product.
Questions to ask any fintech design agency
Ask us these too. A firm's answers to them tell you more than its portfolio will.
- [ ] Show us one state model. Walk us through every status a payment, application or claim can be in, who can move it, and what the customer sees at each.
- [ ] Who approves what? Which roles prepare, approve and operate, and where are limits or dual control enforced?
- [ ] What happens when it fails? Show a failed payment, a dispute or a hold, and the way back.
- [ ] What did you not own? Compliance, underwriting, the codebase. Who supplied missing policy inputs, and what did you do while they were missing?
- [ ] Where does your work stop, and who checks what ships against the design?
- [ ] Which figures were measured, and how? Separate what the design changed from the client's growth.
- [ ] Can you walk us through work under NDA? Apply the same five checks to it.
- [ ] Who will do the work? Meet the people, and ask whether they worked on the cases you are shown.
If two firms remain, give both the same failed payment or disputed claim from your own product, and compare the questions they ask before they design anything.