k
knotCardSwitcher

Product Exercise

How might we increase card-switch conversion in CardSwitcher?

Today, CardSwitcher presents merchants in a generic list. Users have no signal about which merchants matter most to them. By sorting merchants by spend, billing urgency, and reward multiplier, we can surface the highest-value switches first.

Standard

Generic onboarding, no confidence signal

9:41
k
+
A

Update Your Card on File

Choose which merchants to update with your new American Express card.

Saves you time
Securely updates your card at each merchant in seconds.
Thousands of merchants
Works with streaming, delivery, travel, shopping, and more.
Your data is secure
SOC 2 and PCI DSS compliant. Credentials are never stored.

With Personalization

Personalized recommendations build confidence

9:41
k
+
A

Personalized for You

Your merchants, ranked by spend, urgency, and reward potential.

Saves you time
Securely updates your card at each merchant in seconds.
Thousands of merchants
Works with streaming, delivery, travel, shopping, and more.
Your data is secure
SOC 2 and PCI DSS compliant. Credentials are never stored.

Outcome

38%
Standard conversion
58%
Estimated with personalization

Personalized ordering guides users to high-value merchants first. Contextual urgency drives immediate action and builds momentum to complete the full list.

How it works

Where the data comes from

Issuers already share transaction data with Knot. We use that to score each merchant by when they'll bill next, how much the user spends there, and what reward multiplier applies.

Why we think it works

Personalized reordering consistently lifts conversion in similar fintech and e-commerce flows. We modeled a conservative lift on CardSwitcher's baseline — the exact number is less important than the signal from a 30-day A/B test.

How to ship it

A scoring algorithm and a sort parameter on the existing API. No SDK changes for issuers. A/B test with one partner, measure completion rate over 30 days.