When a hero section isn't working, the instinct is usually to redesign it — new colors, new imagery, a bigger button. That instinct treats the problem as aesthetic. Most of the time, the actual problem is hierarchy: the page is presenting five things with equal visual weight, and a visitor given five equally loud things to look at processes none of them.
Visual hierarchy isn't about making things bigger or smaller for style. It's a sequencing decision — deciding what the visitor's eye encounters first, second, and third, and making sure that sequence matches the order in which they need information to make a decision.
The one-thing test
Cover your hero section and ask: if a visitor only absorbed one element before scrolling away, which one would move the needle most? That element should be visually dominant. Everything else — the subhead, the secondary CTA, the trust badges — should be sized and positioned to support it, not compete with it.
- One dominant element — usually the headline or a single strong visual, never both fighting for attention
- One clear next action — a secondary CTA is fine, but it should look secondary
- Supporting proof sized to be noticed, not to compete — a stat or logo bar shouldn't out-weigh the headline
- Whitespace used deliberately to signal what's related and what's separate
If everything in your hero is bold, nothing is. Hierarchy only works when most elements agree to be quiet so one can be loud.
Why this gets missed in review
Hierarchy problems are easy to miss internally because the person reviewing the page already knows what matters — they read past the clutter because they know what they're looking for. A first-time visitor doesn't have that shortcut. They see everything at once and default to reading whatever is visually loudest, which is often not the thing that would actually convert them.
Before redesigning a hero that isn't converting, run the one-thing test with someone who's never seen the page. What they notice first tells you what your hierarchy is actually communicating — which is often not what you intended.