Cover Test Results: How Cover B Beat Cover A by 31 Points and Doubled CTR for a Thriller

# Cover Test Results: How Cover B Beat Cover A by 31 Points and Doubled CTR for a Thriller

*Note: Author details have been anonymised at their request. All vote data is real.*

A thriller author came to CoverCrushing with two cover variants for their debut novel — a psychological thriller set in a corporate law firm. They'd spent three months going back and forth with their designer and couldn't decide. They ran a 500-reader test. Here's what happened.

The Setup

Cover A: Dark navy background, corporate building silhouette, gold serif title treatment. Professional, polished, slightly generic.

Cover B: Close-up of a woman's face partially obscured by a glass reflection, red accent colour, distressed sans-serif title. More unsettling, more personal.

The Results

| Metric | Cover A | Cover B |

|---|---|---|

| Overall Vote Score | 52 | 83 |

| Purchase Intent | 41% | 67% |

| "Would stop scrolling" | 38% | 71% |

| Genre accuracy | 78% | 89% |

Cover B won by **31 points** — a decisive result. But the more interesting data was in the reader comments.

What Readers Said About Cover B

> "That face in the glass immediately made me want to know who she is and what she's hiding."

> "The red feels dangerous. I'd click on this in a heartbeat."

> "This looks like a book I'd actually read. The other one looks like a business book."

What Readers Said About Cover A

> "It's fine but it doesn't tell me anything about the story."

> "Too corporate. I can't tell if this is fiction or a business book."

> "The gold title is nice but the cover doesn't make me feel anything."

The Key Insight

The author had assumed readers wanted the more "professional" cover. The data showed the opposite: readers wanted to feel something. Cover B created an emotional hook — a face, a secret, a sense of danger — that Cover A's architectural composition couldn't match.

What Changed in the Final Design

The author used Cover B as the base but made three refinements based on the reader comments: (1) increased the contrast on the face to make it more immediately readable at thumbnail size, (2) moved the title to the bottom third to give the face more visual space, (3) added a subtle tagline that addressed the most common reader question ("who is she?").

The final cover launched with a 2.3x higher Amazon ad CTR than the author's previous book.

FAQ

How many readers do you need for a reliable cover test result?

For a decisive result like this one, genre-matched readers is sufficient. For closer results (within 10 points), we recommend 1,000 readers for statistical confidence.

Can I test more than two covers?

Yes. CoverCrushing supports up to 4 cover variants per test. Testing 3–4 variants gives you more data but requires more readers to maintain statistical significance per variant.