Choosing between cover concepts
Three concepts on the screen and no idea which one is right is a normal place to be. It is also the point at which most authors stop thinking and start feeling, and the feeling is unreliable for reasons that have nothing to do with your judgment.
Why you cannot see them any more
You have looked at them too many times. Whatever was striking on the first viewing has gone flat, and flatness reads as "boring" when it is really just familiarity. The reader you are selling to gets one look, cold.
You know what the book is about. You read meaning into an image a stranger cannot see — the object that matters in chapter nine, the color that echoes the ending. None of that transmits. You are the one person on earth who cannot view these covers without the manuscript attached.
You are attached to the one that matches the picture in your head. That is the least relevant criterion available. The picture in your head was formed while writing, by someone who has already decided to read the book.
So take the decision away from taste and give it a procedure.
Look at them the way a reader will
Small. Shrink each one to roughly the height of a thumbnail in a store listing. Most of what you have been deliberating over does not exist at that size.
In a grid, on a phone. That is where the browsing happens, and a phone screen is less forgiving than your monitor.
Beside real covers in your category, not beside each other. This is the single change that makes the comparison honest. Put all three into a grid of eight or ten current, successful covers from the shelf you want to be on. Covers compared only against each other are graded on a curve you invented; covers in a real grid are graded on the one that exists.
Two things become obvious immediately. One concept usually disappears against the competition. One usually looks like it was made for a different shelf.
Ask what each one says before asking whether you like it
Take each concept in turn and answer in this order.
Genre first. What kind of book does this say it is? If the answer is not your category, nothing else about the cover matters. A cover can be beautiful and still be mis-addressed.
Tone second. Funny or serious, warm or bleak, fast or slow, literary or commercial. Tone is the thing covers get wrong most quietly.
Then legibility. Does the title read at grid size, or does it only read because you already know what it says?
Preference last, and only as a tiebreak. When two concepts pass genre and tone equally, then your taste is a reasonable way to choose. Before that it is noise.
In the Cover Maker. Previews are free, unlimited and watermarked, so building three full versions and comparing them properly costs nothing but your time — you are not choosing between sketches, and you do not pay until you want one of them without the watermark. The thumbnail check renders each at the size stores display. Start your cover → · about the Cover Maker
Where other people come in
Your own answers narrow the field. They do not settle it, because you cannot ask yourself the genre question honestly. Testing a cover with readers has the questions to ask, who to ask them of, and how to stage it — do not restage it from memory, the setting decides the answer.
One thing worth adding here: test the concepts one at a time for the job questions, and use the grid only for the click test. Asking someone to choose between your three covers turns them into a designer, and you will get design opinions back instead of recognition.
Feedback that conflicts
It will conflict. That is not a failed test, it is the normal output, and the reading of it is simple.
Look for the pattern, not the strongest voice. Five people guessing the wrong genre is a fault in the cover. One person disliking a color is one person's taste, however firmly they said it.
Weigh wrong answers over objections. "I couldn't tell what kind of book it was" is data. "I'd make the author name bigger" is a preference wearing a suggestion's clothes.
A split is information. If half read a concept as one genre and half as another, that concept sits on a boundary. Sometimes that is the book. More often it is why the cover is not landing.
Three traps
Picking the one that looks like more work. Detail is not quality. The busiest concept often reads as the most effortful and performs worst small, because detail does not shrink — it turns into haze. See The thumbnail test.
Picking the one that best illustrates a scene. How faithfully a cover depicts an event in the book is close to irrelevant to a buyer, who does not know the event happens. A cover promises the experience of reading, not the contents.
Combining the best bits of two. This is the most tempting and the most reliably bad. Each concept is a set of decisions that hold together — a palette, a weight, a way of dividing the panel. Take the type from one and the art from the other and you usually get something worse than either, with no principle holding it up. If a specific element from concept B genuinely solves a problem in concept A, treat that as a new brief and design it properly.
When none of them is right
Sometimes the honest answer is that all three are wrong, and saying so on day four is much cheaper than saying so after the book is listed.
The signal is usually in the test: nobody names the right genre, or nobody can say what any of them is about. That is not three covers to choose between, it is a brief that needs rewriting — see Writing a cover brief, for a designer or for yourself.
Choosing the least bad of three wrong covers costs you the whole launch and feels, at the time, like being decisive.
Make your cover
The Cover Maker is free to use: name your book, bring your artwork, and it draws the full wrap — front, spine and back — to your printer's own measurements. You only pay when you want the files without the preview watermark.
Next: Testing a cover with readers · More: Feedback, revisions, and knowing when a cover is done · The thumbnail test · Comp titles: reading the shelf you want to be on