Self-publishing: design the cover pt.2

Following on from my post about how to set up InDesign and Photoshop when creating your book cover, I thought I'd add a few more tips and observations.

Think small

Many of the book covers I like use subtle details and small type. That's not going to work in this web age.

A very large proportion of the people who come into contact with your book are going to do so via the web. If you're adopting a web-based business model, that proportion becomes 100%. And, most often, they're going to see a thumbnail version of the cover. So your design has to work on a small scale.

A cover that seems elegant, classy and understated on the bookshelf may look drab, fuzzy and unexciting when it's only 100-ish pixels wide. To see what I mean, I chose a book from Amazon.co.uk: I didn't spend a lot of time searching, so you'll find much worse examples, but this gives some idea of what I'm talking about. (And to show there's no malice, I selected a book and an author that I admire. In spite of what I say here, I strongly recommend that you buy this book - click on the cover images to visit Amazon.co.uk.)

Here (left) is the main thumbnail of Daniel Dennett's Freedom Evolves. It's a tasteful cover, befitting a leading philosopher. But at 125 pixels wide, the background image (of tree branches) is already starting to become somewhat murky and indistinct. There's no impact to the image at all. It would be hard to pick this book out of a line-up.

Amazon also makes extensive use of 40px-wide thumbnails - eg, for the 'People who bought this also bought...' feature. Here (right) is the 40px version.

The author's name is still readable - just. That's important because Dennett is a well-known writer and that name recognition is important. But the book title has broken up, so you'd be hard-pressed to know which of his books this is. This hasn't been helped by the use of a serif font reversed out of a dark background. Serif faces can suffer when being reversed out at the best of times, and don't stand up to this kind of scaling down treatment at all well.

The background image has become entirely abstract and, well, messy. The book has lost its unique identity. There's no recognition factor.

Key elements

That doesn't mean that everything has to be bold and brash. It does mean that the design needs to retain some impact when it's viewed small. Critically, it has to remain recognisable, and not become a jumble of random pixels.

As a general rule, then, this means using:

• Simple, bold colours - and not too many of them
• Large, simple type for the title - avoid fancy fonts
• A simple graphic element that scales well, retaining its recognisable shape as it gets smaller.

I'll use my own book as an example. That's not because I think the design is perfect - far from it. It's just that I have access to all the design elements.

Here's the cover of my novel, Lady Caine, at 200 pixels wide.

I used Gill Sans for the title. It's a very open, simple typeface. As you can see, it's in black against a white background.

There are several elements to the illustration, but the key one is the aeroplane (a DC-3 I photographed at an airshow back in the early 1990s). This is used large and has been treated to make the image more graphic, including the use of a bold colour. The effect is to make the whole thing simpler while retaining the recognisable form of an aircraft. To see what I mean, here's the original image:

So what happens to these elements when they're scaled down?

Here's a 120px-wide version (right). The title and author name are both still readable. And the graphic is barely affected. The cover remains instantly recognisable.

And much of the same applies when the cover is further reduced, to 40px wide. Okay, the author name has gone, but for a first novel, author recognition is less important - it's the book title that's important.

Don't be so clever

Browsing the Lulu website, you'll soon find examples of classic mistakes. People cram lots of images on the cover, presumably in the hope that people will find something they like. They use fussy designs with no strong graphic elements. They use muddy or dull colours. Or they use complex background images that turn into a frenzy of random pixels as soon as the image gets reduced to web resolution. There are lots of examples of fancy fonts, too - for example, script faces that, I assume, the authors believe are classy or sophisticated or quirky but which, alas, fast become unreadable.

So, you can take away at least one lesson from their examples: KISS.