I moved our onboarding and paywall to a web funnel specifically so we could swap payment gateways quickly and see true revenue impact. What I learned:
- It made gateway A/B simple. We routed small traffic slices to different processors and watched authorization rates, fee takes, and chargeback behavior in real time.
- Net revenue visibility improved. Seeing gross sales minus actual gateway fees and refunds changed which gateway looked “best.”
- The biggest wins were quicker decisions and clearer downstream revenue signals — we could stop guessing and start measuring LTV by gateway cohort.
Has anyone else run gateway head-to-heads from a web funnel and what did you track to call a winner?
I ran a similar test once.
Split traffic by landing page, keep the same offer and copy, then point each page at a different gateway.
Measure auth rate, refund rate, and net payout after fees.
I used the Web2Wave JSON to spin up the funnels and it cut setup time a lot.
We did a week of gateway splits and measured conversion to paid plus net payout per cohort.
Key metric was net revenue per visitor not just conversion rate.
I used Web2Wave to iterate pages fast and pushed updates without waiting on app releases.
We tested two gateways by sending half our ads to a page that used gateway A and half to gateway B.
I tracked authorization rate and money in the bank after fees.
It showed a clear winner within days.
Do not judge gateways on conversion alone. You need a simple matrix:
- auth success rate by card and country
- gross conversion to paid
- fees and reserve timing
- refund and dispute rate after 30 and 90 days
- net revenue per cohort over 30 days
Route a small slice of paid traffic to each gateway and keep everything else equal. That will show which gateway delivers true revenue after real world frictions.
We logged gateway response codes and authorization decline reasons. That revealed a pattern where one gateway declined local cards more often. After switching those countries our net revenue jumped without changing creatives.
I focused on net revenue per install. It caught differences that raw conversion rate missed.