Improving KYC Capture by Reducing Steps and User Confusion
Revisiting an earlier design to improve the KYC camera capture experience—transforming a basic upload flow into a more guided and intuitive interaction through research, exploration, and iterative refinement.
4 days
Concept to refined KYC capture design
6–8 concepts
Explored before selecting final direction
2–3 iterations
Refined through stakeholder feedback
30% fewer steps
Reduced interaction steps in final flow
My Role
Lead Product Designer (Individual Contributor)
Approach
Explored 6–8 concepts (AI-assisted), selected the most efficient flow, and refined it in Figma to reduce steps and improve clarity
Stakeholder Collaboration
Worked with internal and client managers; iterated through feedback cycles and secured final approval
Photo capture in KYC is a high-risk step—users need guidance, not just an option to upload.
KEY INSIGHT
01 — THE PROBLEM
Why Users Were Failing Selfie Verification
KYC verification was a mandatory step before users could proceed, requiring a government ID and a selfie photo. While the overall flow worked, the selfie capture experience created noticeable friction.
The existing interaction relied on a basic upload approach—users were shown instructions and asked to upload or capture a photo. This passive setup placed the responsibility entirely on users to get the image right, without guidance or feedback during the process.
Key Observation
The issue was not the requirement of a selfie—but the lack of guidance while capturing it.
Breaking down the rejection reasons
Face not properly aligned or partially out of frame
28%
Blurry or low-quality images
19%
Poor lighting or shadows on the face
18%
Incorrect orientation or tilted angles
15%
Use of caps, masks, or face obstructions
20%
02 — Discovery & Problem Framing
Redefining the Problem for an Out-of-the-Box Solution
The data showed a high rejection rate for selfie submissions, but the expectation from stakeholders was not just to fix the issue—they explicitly pushed for an “out-of-the-box” solution that would go beyond standard industry patterns.
Given limited time and user access, I adopted a lean discovery approach to quickly frame the problem and move toward solution exploration.
Direction & Constraints
Stakeholders set a clear expectation:
-
The current experience was underperforming
-
Competitor solutions were not benchmarks to follow
-
The solution needed to be innovative and differentiated
This reframed the challenge from improving usability to rethinking the capture experience entirely.
Signals & Insights
53% selfie rejection rate pointed to a systemic issue
This provided enough clarity to move forward without extensive research.
Hypothesis & Opportunity
Based on available inputs, key hypotheses emerged:
-
Users may not understand the need for a live photo
-
Lack of guidance during capture leads to errors
-
No feedback before submission causes avoidable failures
This revealed a clear opportunity:
Shift from a passive upload flow to a guided, system-driven capture experience.
Key Takeaways
Direction & Constraints
The problem was not positioned as a usability fix but as an opportunity for differentiation. Stakeholders explicitly required a solution that goes beyond standard patterns. This shifted the focus toward rethinking the experience rather than optimizing it.
Signals & Insights
The high rejection rate indicated a systemic issue affecting a large portion of users. AI-assisted synthesis enabled rapid identification of likely problem areas without extensive research. This provided sufficient directional clarity to proceed confidently into design.
Hypothesis & Opportunity
Users were likely failing due to lack of guidance, not lack of intent. The absence of real-time feedback created reliance on guesswork. This opened up an opportunity to design a more guided, system-driven capture experience.
03 — Design Solution
From Passive Upload to Guided Capture
The experience was redesigned from a fragmented, instruction-light upload flow into a structured, guided capture process. Instead of relying on user judgment, the system now supports users at each step to ensure a correct photo is captured the first time.
Step 1: Prepare to Take Photo
Users are guided before opening the camera with:
-
Face alignment illustration
-
Clear selfie guidelines (lighting, eyes visible, no obstructions)
-
“Open Camera” action
Why this works
Setting clear expectations upfront reduces errors and prepares users for a successful capture.
Step 2: Take Your Selfie
Users capture their photo within:
-
On-screen face alignment guides
-
Consistent guidance panel for reference
-
“Capture Photo” action
Why this works
ReaVisual guidance during capture removes guesswork and helps users correct mistakes in real time.
Step 3: Review Your Photo
Users validate their photo with:
-
Zoom-in/out preview for quality check
-
Clear quality indicators (lighting, face clarity, eyes visible)
-
“Save & Continue” action
Why this works
A review step allows users to verify and fix issues before submission, reducing avoidable rejections.
Key Design Decisions
-
Shifted from upload-first to camera-first to guide correct behaviour
-
Introduced step-by-step guidance instead of static instructions
-
Added visual alignment + quality checks to improve accuracy
-
Enabled photo review before submission to prevent errors
Before design:
The flow included two capture screens with a “Capture Photo” button and minimal instructions to ensure camera access and enable location services. The “Save & Continue” button remained inactive until a photo was captured.
After design:
The flow was redesigned into three guided steps. First, a “Prepare to take photo” screen provides visual alignment guidance, an illustration, and clear selfie tips, with an “Open Camera” action.
Next, in “Take your selfie,” users align their face within on-screen guides, with the action updated to “Capture Photo.”
Finally, a “Review your photo” screen lets users inspect and zoom the captured image, alongside quality checks (lighting, face clarity, eyes visible), with the final action changing to “Save & Continue.”
04 — Testing & Validation
Validated through stakeholder reviews + AI evaluation
Due to project constraints, the redesigned flow was validated through structured reviews with 4–5 managers and stakeholders, along with an AI-driven evaluation using Floto. The objective was to ensure the solution addressed key usability issues and aligned with business and technical expectations before development.
Stakeholder Review
The high-fidelity Figma prototype was presented to cross-functional stakeholders to assess usability, clarity, and feasibility.
-
Strong alignment on shifting from upload to guided capture
-
Positive feedback on face overlay and real-time feedback improving accuracy
-
Consensus that the flow reduces dependency on user judgment and minimizes errors
-
Validation that the solution is feasible for implementation and aligns with KYC requirements
AI Flow Review (Floto)
The flow and UI were further evaluated using Floto to simulate user interaction and identify friction points.
-
Confirmed that the guided experience reduces ambiguity and improves task clarity
-
Highlighted smooth step-by-step progression with minimal cognitive load
-
Identified minor improvements in instruction clarity and edge-case feedback (e.g., low light, blur)
Outcome
Stakeholder alignment combined with AI validation provided confidence that the redesigned experience is clear, reliable, and implementation-ready, significantly reducing the risk of errors in the KYC photo capture process.
Next
Back to Home
View Details