Features instead of outcome
The home page explains the product instead of naming the visitor’s problem.
SaaSIndustry
A/B testing for SaaS.
More sign-ups are not always better. Measure further along, at real use.
- Main goal: Completed sign-up
- Segments by source checked
New product page
RunningAll goalsThe main goal decides
VariationVisitorsCompleted sign-upRateMain goalButton clickRateFirst real useRate
AOriginal4,800Baseline3.2%Baseline11.0%Baseline1.9%
BNew page4,800+12.5%3.6%+20.0%13.2%+10.5%2.1%
Segments by sourceMain goal, change
Organic search+11.8%
Paid search+13.1%
Email+12.4%
The gain comes from more than one source, not from a single campaign.
The pattern
Common mistakes
on SaaS sites.
Price hidden
“Price on request” filters out not just bad leads, but good ones too.
Signup as the only goal
Counting only registrations optimises for dead accounts.
Concretely
Tests to start with
Turn the hero to one outcome
A concrete outcome instead of a category. Goal: registration started.
Show prices
Starting price versus request. Goal: qualified enquiry, guardrail: total enquiries.
Shorten the form
Four fields instead of nine. Goal: completed registration.
Demo versus trial
Two routes as a redirect test. Goal: whatever actually counts downstream.
Measuring
The right goal
Take the latest step you see often enough during the test, usually the completed sign-up. The click rate on the button only runs alongside.
Matching platforms
Other industries