MetriFi
Research report

9 winning patterns from auto loan experiments

Nine winning patterns we’ve discovered from running A/B test experiments on auto loan pages.

Below are 9 winning patterns we’ve discovered from running A/B test experiments on auto loan pages.

List of auto loan experiments

Below is the list of tests we’ve run on auto loan pages.

Test Page Result
#11 Auto loan lead form Won
#15 Auto loan product page Likely won
#22 Auto loan (rates + calculator) Likely won
#39 Auto loan landing page Won
#63 Loan overview (includes auto loan tiles) Likely won
#66 Multi-product (auto loans, home equity, etc.) Likely won
#78 Auto loan product page (main hero test) Won
#87 Auto loan calculator page Won

Total: 8 auto loan–related A/B tests

  • 🟢 Winning / likely winning: 11, 15, 22, 39, 63, 66, 78, 87
  • 🔴 Losing: none (no negative results)
  • ⚪ All eight tests either statistically significant positive or positive trends

Winning patterns: what consistently worked

1) “Get Pre-Approved” outperformed “Apply Now” as the first action

What the tests changed

  • 78: Hero CTA changed from “Apply for an Auto Loan” → “Get Pre-Approved Today.”
  • 39: Primary CTA pair changed from “Apply Now” + “Check Status” → “Apply Online” + “Calculate Payment.”
  • 15: Hero button: “Start Application” replaces “Learn More.”

Why it worked: Auto loan visitors are often shopping, not yet committed. A lower-friction verb (“Get Pre-Approved”) increases click intent because it promises progress without full application commitment.

Evidence: [78, 39, 15]

2) Rate transparency near CTAs drives more starts

What the tests changed

  • 78: Adds “APR as low as 6.00%” and “Flexible Terms up to 84 Months” next to CTA.
  • 15: Introduces visible View Rates button under hero.
  • 22: Integrates rate info into calculator results + “View All Rates →” link.
  • 66: Adds “as of [date]” label next to rate disclosures for trust.

Why it worked: Visible, date-stamped rate and term details near the decision area made users feel informed and increased CTA click confidence.

Evidence: [78, 15, 22, 66]

3) Compact payment estimators boost engagement, but only with a next-step CTA

What the tests changed

  • 22: Introduces an inline loan estimator with amount + term inputs and “View Rates / Apply” directly below.
  • 87: Adds a “Get Started” CTA immediately below the calculator output; calculator placed in a bordered card.
  • 39: Promotes “Calculate Payment” as co-primary CTA beside hero Apply button.

Why it worked: Users could experiment, then seamlessly continue with “Get Pre-Approved / Apply / View Rates.” Estimators without next-step CTAs (common in older variants) led to dead ends; these didn’t.

Evidence: [22, 87, 39]

4) Scannable card and grid layouts beat paragraphs

What the tests changed

  • 15: Converts a paragraph-heavy section into icon + benefit cards (3 columns).
  • 78: Restructures the whole page into white cards (benefits, features, FAQs).
  • 63: Removes one irrelevant loan tile (Unsecured Loan), focuses cards on Auto, Credit Card, Mortgage.
  • 22: Groups rate table, calculator, and disclosures into logical cards.

Why it worked: Breaking loan content into cards improves scanning and moves users faster toward CTAs.

Evidence: [15, 78, 63, 22]

5) “Why choose us” and security cues reassure users before commitment

What the tests changed

  • 87: Adds compact “Why Choose Us” block (icons + short lines) right below estimator.
  • 11: Adds a security reassurance line (“Your info is safe”) and labels on each field.
  • 15: Adds member testimonial strip below the hero, not above.

Why it worked: Trust reinforcement near hesitation points (before data entry or next-step buttons) reassures without distraction.

Evidence: [87, 11, 15]

6) Short, segmented application flows increase completion

What the tests changed

  • 11: New micro-steps for lead capture (Email → Name → Phone).
  • 15: Simplifies “Start Application” to open a short, in-page form instead of full multi-step.
  • 87: Moves “Get Started” CTA inside a short modal instead of a full redirect.

Why it worked: Segmenting or simplifying forms reduced perceived effort and boosted completion.

Evidence: [11, 15, 87]

7) Benefit-led, plain copy (not lifestyle messaging) increased clicks

What the tests changed

  • 15: Headline from “Drive what you love” → “Low Rates on New, Used, and Refinance Auto Loans.”
  • 39: Copy from “Apply Online for Your Next Car” → “Get Pre-Approved and Shop with Confidence.”
  • 78: Adds subheading “Flexible Terms up to 84 Months” under rate line.

Why it worked: Removing abstract “dream car” messaging and replacing it with direct benefits (rate, term, speed) improved clarity and clicks.

Evidence: [15, 39, 78]

8) Contextual but muted imagery supported readability

What the tests changed

  • 78: Added background photo of person with keys but overlaid a white card to keep text legible.
  • 15: Reduced background image opacity and added left-aligned text.
  • 39: Switched to neutral tone car imagery with more whitespace.

Why it worked: In lending pages, high-contrast text and clear buttons outperform colorful lifestyle photos.

Evidence: [78, 15, 39]

9) Focused navigation + removing irrelevant loan types helped users choose faster

What the tests changed

  • 63: Removed “Unsecured Loan” tile and “Apply” from exploratory cards; left only Auto, Credit Card, and Mortgage.
  • 94: (global pattern) Reordered nav so Loans appear before Accounts when auto promotions are active.

Why it worked: By reducing irrelevant options, auto loan CTRs improved.

Evidence: [63, 94]

Things to watch out for

No auto loan test in this dataset statistically lost, but below are some patterns to watch out for that neutralized wins or introduced risk of noise in similar contexts.

A) Too many CTAs clustered together

  • In some early variants (e.g., 15 A-control, 63 A), multiple Apply buttons (“Apply Online,” “Learn More,” “Calculate Payment”) appeared side-by-side.
  • The winning versions reduced to one or two clear CTAs.
  • Lesson: Avoid CTA clusters in heroes; show only one main action and one secondary (View Rates / Calculate Payment).

B) Calculators without CTA or buried in long flows

  • Some control variants (22 A) hid the calculator below long benefit text or didn’t attach a “Start Application” next step.
  • Lesson: Keep calculators close to hero and make sure they end in an action.

C) Generic imagery or decorative icons

  • Auto variants that kept bright, busy lifestyle photos in the hero (cars on highways, colorful motion blur) underperformed vs white-card designs.
  • Lesson: Imagery should support, not dominate, the call to action.

What to ship on auto loan pages (best-practice blueprint)

Hero

  • H1: “Low Rates on New, Used, and Refinance Auto Loans”
  • CTA: Get Pre-Approved Today (primary); View Rates (secondary)
  • Add “APR as low as X.XX% · Terms up to 84 months · As of [date]” next to CTA [78, 15, 66]

Benefits (cards / short list)

  • 3–5 bullets: “Fast decisions,” “No prepayment penalty,” “Flexible terms up to 84 months” [15, 78]

Estimator

  • Simple 2-field calculator (amount + term); output: “Estimated payment: $___/mo”
  • Button right below: Get Pre-Approved / Apply Now / View Rates [22, 87, 39]

Trust and reassurance

  • Compact “Why Choose Us” icons; one-line security note near form fields [11, 87]

FAQ / disclosure

  • 6–8 key Qs, “View more FAQs” toggle [15, 78]

Nav

  • When running auto promos, show Loans first and list Auto Loans before others [63, 94]

Summary table (auto loan tests)

Test ID Product Outcome Key winning change
11 Auto loan lead form ✅ Won Micro-steps + security reassurance
15 Auto loan product page ✅ Likely won Hero simplification + benefit copy
22 Auto loan calculator ✅ Likely won Interactive estimator + next-step CTA
39 Auto loan landing page ✅ Won Hero CTA simplification + calculator focus
63 Loan overview (includes auto) ✅ Likely won Removed irrelevant loan cards
66 Multi-product (includes auto) ✅ Likely won Date-stamped rate blocks
78 Auto loan product page ✅ Won “Get Pre-Approved” CTA + APR & term next to button
87 Auto loan calculator page ✅ Won “Get Started” CTA added below estimator
Let’s talk

Want to know what would work on your site?

We’ll look at your website with you and show you what the evidence says to test first.