This cheat sheet outlines seven winning patterns from MetriFi’s data set of 75 A/B tests, and cites the test IDs that prove them. Knowing these patterns will help you quickly say, “We should test [X] on our own website because it’s been proven to work.”
1. Single, specific hero CTA repeated at decision points
Have one main CTA in the hero (e.g., “Open a checking account”, “Apply now”) and repeat that same CTA where users decide again (below benefits, on a calculator, end of page).
Why: Tests showed better movement when the primary action was singular and re-surfaced at natural decision spots.
Tests: [1, 2, 49, 53, 55, 87, 93]
2. Match CTAs to buying stage
E.g., pre-approval before full application
On overview content, show “View [product]” or “Learn more” (not “Apply”). On specific products, show “Open”/“Apply”. For loans, consider “Get pre-approved” before “Apply”.
Why: Tests won by matching the commitment level of the CTA to where the user is in the journey.
Tests: [62, 63, 67, 68, 72, 78, 91]
3. Scannable structures
E.g., cards/grids/accordions
Turn long blocks into card-type layouts with short headings and bullets, on a 2–3 column grid. Use accordions for FAQs and disclosures.
Why: Tests showed improved forward movement when information was easier to scan.
Tests: [1, 2, 31, 33, 40, 41, 43, 49, 62, 72]
4. Rates/terms surfaced with date stamps
Put APY/APR, term, minimums right next to the hero or a calculator and include an “as of [date]” label. Keep a nearby “View all rates” link that shows more rates without leaving the page (e.g., in a pop-up modal).
Why: Tests indicate users decide faster when key numbers are visible and current.
Tests: [62, 63, 64, 66, 70, 81, 90, 91, 92]
5. Calculators paired with rates and calls to action
If users must estimate (payment/earnings), provide a compact calculator along with rates and put “Get Started / Apply / View rates” directly under the result (plus a short “estimate only” disclaimer).
Why: Wins occurred when calculators didn’t dead-end; they led straight to the next action.
Tests: [22, 30, 33, 39, 40, 43, 49, 53, 55, 72, 87, 90]
6. Trust at the hesitation point
E.g., security line, “Why us”
Add a one-line security note right by sensitive fields; place a small “Why choose us” block (3–4 bullets/icons or a stat strip) near the hero or calculator, not as the main focus.
Why: Tests reduced perceived risk at the exact moment of commitment.
Tests: [1, 9, 10, 11, 33, 66, 85, 87]
7. Navigation as a lever
E.g., focus first; vertical lists
Make products easy to find via vertical product lists; put the priority line of business first in the navigation (“Accounts” if deposits are the goal, “Loans” if lending is the push).
Why: A test showed better findability when navigation reflected the business priority and listed specific products, not just categories.
Test: [94]