On Page Tests vs. Full-Page Tests

Not every problem needs a full redesign. This class teaches you when to run small element tests versus when to test an entire page, and how to avoid turning A/B testing into “random redesign roulette.”

Element tests are best when:

• The page structure is mostly correct, but a key piece underperforms (CTA, headline, proof, pricing clarity).
• You have a specific hypothesis tied to one component.
• You want speed and low risk.

Full-page tests are best when:

• The page is fundamentally mismatched to the user intent (wrong narrative, wrong order, missing sections).
• Heatmaps/replays show widespread confusion (scrolling, backtracking, abandoning multiple sections).
• Your “fix” requires a new layout or a different flow.

Rule of thumb: Start with an element test whenever possible. Earn the right to redesign. If small changes can’t move the metric, then escalate to a full-page test.

AB Split Test workflow: Use element tests inside blocks/builders for quick iterations. Use full-page tests when the variation needs a different page narrative. Validate with heatmaps and session replays so you’re not guessing whether the redesign actually improved comprehension and decision-making.