What it is.
Find the single weakest stage in your funnel, focus testing there for the next 90 days, and run 2 to 4 tests a month against that one priority.
The evidence given is a 12-month program that lifted full-funnel visitor-to-paid by 2.3x while customer acquisition cost stayed flat, and a portfolio pattern where role-specific landing pages convert 2x cleaner than generic ones.
TripleDart's own guide with our own portfolio data behind the claims. The rhythm works with any testing tool.
What you get.
- A monthly operating rhythm at 2 to 4 tests against one named priority.
- A 90-day focus window per funnel stage, so testing stops scattering.
- The role-specific landing page finding, at 2x cleaner conversion than generic pages.
- Named conversion strategies, from sign-up simplification through to exit-intent.
- A stated result from a 12-month program: 2.3x visitor-to-paid with acquisition cost flat.
How to set it up.
Map the funnel and find the single weakest stage before designing any test.
Commit to that stage for 90 days, because moving on after two weeks is why most programs show nothing.
Run 2 to 4 tests a month against that stage only.
Build role-specific landing pages for your main personas, since generic pages convert about half as well in this data.
Simplify the sign-up before testing any step that comes after it.
Track visitor-to-paid, so a win at one step that leaks at the next still shows up.
Use cases
Set a monthly test rhythm
Run a small number of tests each month against one named priority.
Focus a funnel stage
Hold one stage for a quarter so each test builds on the last.
Make the case for a program
Use the twelve-month result when asking to fund testing properly.
Best for
Teams reaching for a redesign
A redesign changes everything at once and teaches you nothing. Monthly tests on one stage teach you something every month.
Paid traffic with weak conversion
Fixing the landing stage is cheaper than buying more clicks.
Setting up a first CRO habit
2 to 4 tests a month is a rhythm a small team can keep.