Case Study: BorrowerCheck™ Integration

Dealertrack | Lead UX Product Designer

I Integrated an AI-powered third-party fraud analytics solution into the existing F&I compliance workflow, where deal accuracy, compliance, and speed all carry real weight. The goal: make fraud detection feel native to the dealer’s workflow, not bolted on.

Learn More

The Problem

F&I managers were navigating compliance checks manually: a fragmented, time-consuming process that created friction at a critical point in the deal.

Without real-time fraud signals in context, dealers made decisions with incomplete information, and compliance stayed reactive. There was no prior UX research on these workflows, and significant technical constraints to design around.

Integration Goals

  • Reduce customer churn
  • Satisfy demand for advanced fraud mitigation
  • Keep pace with competitors
  • Streamline compliance workflows

Three Fraud Screening Tools to Bring into the Deal

BorrowerCheck™

Point Predictive's proprietary AI screens for fraud across five checks.

  • Fraud Alert
  • Synthetic ID
  • OFAC
  • Red Flags
  • Military Lending Act

License Scanning

Identity verification sent to the customer's phone, returning one of three results.

  • Verified
  • Not verified
  • Pending

One-Time Passcodes

A 6-digit code the customer enters within a time limit to confirm they control the phone on file.

Core Design Pillars

Minimize friction, maximize trust and efficiency.

Fraud Insights inside a Familiar Workflow

Real-time signals that reduce reliance on manual compliance checks and improve decision confidence.

Seamless Hand-Off to Mobile Verification

One step sends the license scan and selfie capture to the customer's phone, with status tracked inside the deal.

Proactive Compliance

Intelligent alerts and verification outcomes that guide dealers toward safer, faster deal closings.

Built to Scale

Adapts to evolving mandates and future additions like auto-filled deal jackets and multi-verification flows.

My Process

  1. Heuristic Audit

    Mapped pain points across the existing F&I workflow to set a baseline.

  2. Competitive Review

    Studied how RouteOne approaches identity verification and fraud checks.

  3. User Flows & Wireframes

    Repositioned fraud detection as a natural step in the deal, not an interruption.

  4. Prototyping & Testing

    Clickable prototypes tested with F&I managers on states, verification, and alerts.

  5. High-Fidelity Design

    Custom components and a modular widget framework, built for the MVP and ready to scale.

Discovery & Flows

Current Experience Flaws

The heuristic audit showed compliance was a manual detour. Each finding shaped a design decision.

Compliance checks were manual and fragmented

Dealers pieced verification together outside the deal flow.

Solution: Fraud checks embedded in the credit application.

No fraud signals at the moment of decision

Dealers made decisions with incomplete information.

Solution: Real-time result cards on the ID Verification screen.

Compliance was reactive, not proactive

Problems surfaced after the fact, not when they could be acted on.

Solution: Result cards paired with step-up tools that point to the next action.

Ambiguous states and dense compliance language

Dealers had to parse jargon under time pressure.

Solution: Color-and-label status pills with plain-language reasons.

High-Level Workflow Entry

Dealers opt into fraud checks inside the Credit Bureau Application they already complete on every deal. Then one validation, one submit, and the bureau request processes. For deeper checks, including the driver's license scan, dealers use the dedicated ID Verification step.

  1. Select Bureaus

    All bureaus, or Equifax, Experian, and TransUnion individually

  2. Verify ID

    Synthetic ID and Red Flags / OFAC checks, opted in with a checkbox

  3. Enter Applicant Info

    Personal, address, identification, and contact details

  4. Confirm Permission

    Confirm permissible purpose, or capture the customer's signature

Five Verification Paths

Five paths, one verification results screen. Pick a path to walk through it.

Driver's license scan

  1. DealerEnters the customer's phone number and continues.
  2. CustomerScans their license and takes a selfie on their phone.
  3. DealStatus is tracked right inside the deal.
  4. Outcome
    • Verified
    • Not verified
    • Pending

Result Cards Architecture

Results guide the dealer to the next action. Each automated check gets its own card with a status pill. Cards appear only when relevant, and optional step-up tools sit in their own column beside them.

ID Verification · Results

Automated checks

Fraud AlertMedium risk
Synthetic IDNo risk
OFACPassed
Red FlagsPassed
Military Lending ActCovered

Step-up verification

Scan Drivers License Send One-Time Passcode Ask Out-of-Wallet Questions
Illustrative recreation of the results layout.

Design Artifacts

Solution Architecture

One verification screen that flexes to every subscription tier, with the license scan as an opt-in add-on to BorrowerCheck™.

Why
Point Predictive sells three tiers: license scan only, comprehensive scan only, and both. One layout serves all three.
Bonus for dealers
On the combined tier, dealers choose per deal. F&I managers said they'd add the scan for remote and out-of-state deals, and skip it in person.

Key Design Decisions

  • Clarity of state

    Distinct UI states with clear hierarchy, so dealers orient at a glance instead of parsing compliance language under pressure.

  • A clean hand-off to mobile

    Send the scan to the customer's phone in one step, then track verified, not verified, or pending status inside the deal.

  • A guide, not a gatekeeper

    Fraud signals framed as decision support. Messaging is directive without being alarmist.

  • Results that document themselves

    Every result is stored automatically in the deal's digital document repository. No uploads or screenshots.

Wireframe to High Fidelity

Wireframe
High Fidelity

Edge Cases & Iterations

Every result state is designed for each fraud check. Fraud Risk, for example, returns one of three levels, each pairing a color, a label, and a plain-language reason.

Switch the result

Fraud Risk Low Risk
  • The borrower's credentials matched via independent verification.

Low risk. Reads as clear at a glance, so the deal keeps moving.

Explored several iterations for the verification results screen.

A · Verdict first. A banner says how many checks need review, then a severity-sorted list.
B · Triage lanes. Checks sorted by what the dealer must do next, with step-up tools inside the flagged card.
C · Master–detail. A list on the left, one result at a time on the right, so it scales as checks are added.
D · Guided review. Verification as a step in the deal, with a progress stepper and a recommended next action.

Prototype

Interactive prototypes and UX Blueprints are under NDA and remain the property of Cox Automotive. Request access to view.

Designed to Validate

The verification experience end to end, from data entry to results, built to test comprehension, decision speed, and confidence.

  • Entry from the existing IDV workflow
  • Mobile license-scan hand-off with live status
  • Fraud results cards and recommended next steps
Request Access

Research & Testing

What F&I Managers Said

Enthusiasm for the tools, with sharp questions about fit.

What We Tested

  • Understanding of BorrowerCheck™ capabilities
  • Clarity of the "Include Driver's License Scan" option
  • Tooltips, field masking, and validation
  • Reactions to passed, mixed, and failed results
  • Discoverability of step-up tools in result sets

What Landed

SSN verification
The most valued new feature in the session.
Color-coded status pills
Trusted, as long as each card says how to proceed.
License scan for remote deals
Most useful for remote deliveries and out-of-state buyers.

What Worried Them

License scan timing
Unsure where, how, and when to upload the license.
Redundancy
The license may already be in VinSolutions.
Passcode value
A code sent to a customer-provided number proves little.

For remote deliveries, I'd definitely use the driver's license scan and text verification because fraud risk is higher.

Finance Manager 3, on where the license scan pays off

If the salesperson already collected the license and entered it into VinSolutions, why do I need to upload it again? It feels redundant.

Finance Manager 1, on workflow redundancy

I don't see the passcode as a strong security check. Anyone committing fraud can easily control the phone number.

Finance Manager 3, on one-time passcodes

Outcomes & Recommendations

Testing de-risked the build before engineering committed.

Reduction in time-on-task vs. the current F&I experience
16%
Improvement in task completion rate
12%
Single Ease Question score, above the ~5.5 benchmark
6.2/7
Usability rating for result states and secondary tools
78/100

3 workflow risks caught before build: scan timing, VinSolutions redundancy, and passcode value. Testing also validated the core pattern: card layout and colored pills keep fully passed deals moving.

Strategic Recommendations

  1. Keep the card layout and colored pills

    When every check passes, dealers keep moving. Failed checks are easy to spot.

  2. Explain the license scan up front

    Add context on the data entry page so that dealers know when and why to scan.

  3. Rethink the one-time passcode

    Reconsider it, or clarify when it adds real compliance value.

  4. Spotlight SSN verification

    Dealers saw it as the biggest fraud-prevention win, so keep its results prominent.

  5. Connect stipulations and proof of address

    Uploading them in this workflow serves the most common Red Flags use case.

The biggest design challenge wasn't the complexity of fraud detection; it was making something inherently compliance-heavy feel like a natural, low-friction part of a dealer's day. Working without existing research and within tight technical constraints, the heuristic audit became the foundation everything else was built on.