explore

Fintech / E-commerce

Tinkoff Checkout

Turning checkout drop-off into ~30% conversion improvement.

Davit PedanyanLead Product Designer · SME · E-commerce · Fintech

Context

A familiar bank.
An unfamiliar checkout.

Tinkoff Checkout gave merchants a checkout they could integrate into their websites and platforms, either as the main flow or a Tinkoff-powered alternative.

It started as an experimental product with several hundred business users. The opportunity was to turn familiarity with the bank into confidence at the moment of payment.

~38M
Tinkoff customers at the time
The bank ecosystem, not Checkout users.
2–3×
Business-user growth during my time on the product
A shared product outcome, not attributable to design alone.
  1. Tinkoff ecosystem
  2. Merchant website
  3. Integrated checkout
  4. Customer payment

My contribution

I led the design.
We solved the problem together.

I owned and led

  • Design strategy and direction
  • Design exploration and final product design
  • Usability testing and experimentation
  • Multivariate testing across key funnel stages

I helped shape

  • Product strategy and roadmap direction
  • Quarterly planning and budget discussions
  • Metrics analysis with Power BI
  • Executive reviews and business priorities

We worked together on

  • User and merchant research
  • Qualitative and quantitative analysis
  • Problem definition and prioritization
  • Launch and iteration

Product managers · Researchers · UX writers · Business analysts · Engineering · Merchants

The problem

Analytics showed where customers left.
Not why.

We mapped abandonment across three transitions. The goal was to help more customers complete their purchase, not simply make the checkout look different.

01

Merchant websiteCheckout entry

Customers did not continue through the entry button.

Drop-off observed
02

CheckoutAuthentication / OTP

Customers entered, but did not progress through authentication.

Drop-off observed
03

AuthenticationPayment completion

Even authenticated customers left before completing payment.

Drop-off observed

Historical stage-level conversion rates are unavailable. No funnel percentages are inferred here.

Research

The missing context was costing trust.

Customers

Interviews and usability testing helped us understand the experience after someone selected Tinkoff Checkout. The transition led into a new environment without enough recognizable branding or transaction context.

Merchants

We paired that research with business-side interviews, including Golden Apple, and quantitative analysis. This connected customer uncertainty with the merchants' checkout problems.

The root cause

In a financial flow,
losing context means
losing trust.

Customers needed to understand who was handling the payment and what they were paying for.

The process

Evidence changed the direction.

Research, design and validation were connected. The process below adapts the supplied UX/UI guide, with Checkout-specific evidence alongside each stage.

From question to release

  1. 01Discovery
  2. 02Ideation
  3. 03Interaction design
  4. 04Validation
  5. 05Design & specifications
  6. 06Review & UI testing
Evidence can send us back to exploration.

01 / Discovery

Understand the task

Business goals, user stories, constraints and deadlines come before a solution.

  • Business goal
  • Analytics
  • Constraints

Checkout: identify where customers abandon the funnel.

02 / Ideation

Find the reason

Combine data analysis with internal and external research. Map the user journey and challenge the first explanation.

  • User interviews
  • Merchant research
  • Journey

Checkout: customer testing and merchant interviews revealed a trust gap.

03 / Interaction design

Explore alternatives

Work through information architecture, scenarios and navigation. Agree on a direction with product and engineering.

  • Scenarios
  • Variants
  • Alignment

Checkout: explore entry wording, brand recognition, order context and hierarchy.

04 / Validation

Test, then revise

When a solution needs validation, build a prototype and return to research. The answer can send the design back for another iteration.

  • Prototype
  • Experiment
  • Feedback

Checkout: use multivariate testing to inform the selected direction.

05 / Design & specifications

Resolve the whole flow

The process guide includes the main flow, edge cases, loading, empty, error and success states, plus adaptive layouts.

  • Main flow
  • States
  • Engineering

Checkout: bring the selected approach into final product design with the cross-functional team.

06 / Review & UI testing

Check and release

The guide closes with team review, scenario and UI checks, and a bug-fixing loop before release.

  • Review
  • Fix
  • Release

Checkout: the optimized experience launched. Conversion improved by approximately 30%.

Entry point

Make the next action clear before the click.

We explored wording, visual treatment and the value proposition of the entry button. Multiple alternatives went into testing rather than choosing one on appearance alone.

Original button specifications: spacing, width, height and color alternatives.
Original button specifications: spacing, width, height and color alternatives.

Original button specifications: spacing, width, height and color alternatives.

Original entry-button explorations: labels, benefits and personalized variants.
Original entry-button explorations: labels, benefits and personalized variants.

Original entry-button explorations: labels, benefits and personalized variants.

Trust & purchase context

Keep the merchant, the bank
and the order connected.

Stronger brand recognition and visible purchase details addressed the uncertainty after the transition. The explorations compare how authentication, merchant identity and the order summary fit together.

Original Figma comparison: existing authentication and alternative brand / order-context treatments. These are explorations, not a claimed test winner.
Original Figma comparison: existing authentication and alternative brand / order-context treatments. These are explorations, not a claimed test winner.

Original Figma comparison: existing authentication and alternative brand / order-context treatments. These are explorations, not a claimed test winner.

Hierarchy & experimentation

Compare alternatives.
Then choose on evidence.

We explored the arrangement of delivery options, purchase information and the next action. Multivariate testing informed the direction that went into final product design.

  1. Concept
  2. Design variants
  3. Multivariate testing
  4. Selected direction
Original mobile checkout variants from Figma. Individual experiment results and the winning variant are not documented in the supplied materials.
Original mobile checkout variants from Figma. Individual experiment results and the winning variant are not documented in the supplied materials.

Original mobile checkout variants from Figma. Individual experiment results and the winning variant are not documented in the supplied materials.

The experience

One purchase.
A continuous path to payment.

  1. 01Merchant checkout
  2. 02Tinkoff entry
  3. 03Recognizable checkout
  4. 04Verify order
  5. 05Authentication
  6. 06Payment
  7. 07Success

Structured user flow reconstructed from the case brief. Product explorations are shown above; this diagram is not a screen-by-screen release specification.

Launched & learned

A clearer checkout.
More completed purchases.

~30%

relative improvement in checkout conversion

Optimized and launched

The work moved from funnel analysis and research to tested alternatives and a released checkout experience.

For me, the lesson was to treat trust as part of the transaction itself: the customer needs continuity, not just fewer steps.

Approximate relative improvement recalled by Davit. The historical analytics report is not currently available. This is not a 30-percentage-point increase, and no revenue or stage-level uplift is claimed.

Next case studyiCredoKeep scrolling to continue