You wanted a headline test, not a second funnel
You had a line you wanted to try, and the honest path was duplicate components or another funnel branch. So you postponed the test.
The enemy is treating every copy tweak like a new build.
On coded pages you can paste an <inline-split-test> block: each <variant> gets a name and optional weight (weights sum to 100). Save once; the host validates, stores the record, and can hand back an id for later edits.
Equal split (no weights):
<inline-split-test name="Headline Test">
<variant name="short headline">
<h1>Buy Now</h1>
</variant>
<variant name="long headline">
<h1>Buy Now and Save 50% Today</h1>
</variant>
</inline-split-test>Weighted (70/30):
<inline-split-test name="Headline Test">
<variant name="control" weight="70">
<h1>Original Headline</h1>
</variant>
<variant name="challenger" weight="30">
<h1>New Headline</h1>
</variant>
</inline-split-test>At request time the server picks an arm with weighted rotation, sets a 7-day cookie so return visitors see the same variant, swaps the tag for a wrapper with data-sid / data-cid, and the tracking script fires split-test-view from those attributes so conversions stay attributable.
Try it: paste the weighted block into a staging coded page, save until an id appears, open two incognito windows, refresh until both headlines show while the URL bar stays identical.