Coinciding with the release of The Pauline Conversion, I’ve given my other three books new covers. In two cases tweaks rather than wholesale changes: in one, something entirely new. I want to delve a little into the process.



To state the bleeding obvious, covers are the first part of a book encountered by potential readers. They’re the first muddy hillock you need to urge them over on the obstacle course that ends with page one. If you can’t get them over that first hurdle, it doesn’t matter how perfect the words are. There are plenty of other books out there upon the miles of shelving, real and virtual, to choose from.

(None of this applies to those authors whose names take up half the cover: they could publish a collection of shopping lists called Nnnnggngg and it’d shoot to the top of the bestseller list. I don’t have that luxury, or a collection of shopping lists.)

The only objective metric I have is sales. This measures a lot more than cover quality, but it’s all I have. In an ideal world I’d test two covers (leaving everything else identical) and see which sells more — and then swap the poorer design for a new one, and repeat until Nnnnggngg. Marketing types and those that hang around them call this A/B testing.

It’s tricky to perform an A/B test with print books, but what’s stopping me doing it for ebooks? Amazon, I suppose. They could make it easy: let me have multiple variants of a book in the Kindle store, and use their big black box of twiddly knobs to show different variants to different people. But they don’t let me do that. It’s a shame. It’s a way for authors — and themselves — to make more money. No other ebook vendor has the feature either, as far as I know.

In the absence of A/B testing, all I can do is change the cover of a book every now and then and see if, long term, the graph creeps up. But you never know if that’s due to a better cover, or more books, or a media appearance, or any other variation between the two time periods. When you’re A/B testing you want the only difference to be between A and B: the two covers. You can’t test cover A in country X and cover B in country Y, or any other artificial variation, because you never know if that’s the cause of the difference in sales, not the covers.

I’m left with hunches and feelings and rolls of the dice, and I can’t make spreadsheets out of those, dammit.

For a while I’ve had a strong hunch the old cover for The Pink and the Grey wasn’t doing a great job: I wanted to change that. The hunch wasn’t so strong for Till Undeath Do Us Part or Disunited, but I was willing to explore alternatives.

Accompanying the various hunches were a couple of practical reasons for changing the covers. The aspect ratio (1:1.6) seemed a little off – too tall. More books seem to use the slightly fatter 1:1.5 ratio. More importantly, today there are more devices with high-density “retina” displays, and Amazon’s recommended pixel dimensions for covers had increased. The old images weren’t detailed enough: they’re now 3200×4800, which is a step up from the 256×192 of my ZX Spectrum days.

Changing the “brand”

The previous covers shared certain elements, defining – if you like – the “brand”. Changing all covers at once brought the possibility of updating these shared elements.

All three of the old covers used the same font for both the book title and author name (one of the thousands of variations of Univers), and included a diagonal design element top to bottom. I loved the concept of the diagonal, but it imposes what I shall politely call a “creative burden” on the designer (the “yeah, but how the hell can I shoehorn the diagonal into this one?” problem). Using the same font for all titles linked the covers well: but left no room for genre.

The new covers take a different approach. There’s still a common font (now Avenir Next) but it’s only used for the author name and media quotes (and printed book internals such as chapter titles). The book titles have unique fonts that better fit the story, I hope. There’s no requirement for a diagonal element now (though I’ve kept it for two books, as it works for those independently of branding). Instead there’s a loose theme of rich, solid colours.

Each book now has a tagline too, digging fractionally deeper into plot.

Enough common ground to link the covers when they’re placed together. Enough flexibility to design for the genre and the story.



This cover has changed least.

The player is shifted left to accommodate the great press quote. The title font is much stronger, with a hint of newspaper headline about it.

The tagline tries to communicate the momentousness of the story: the huge change Danny and the sport go through. I’m also using “Out – and outnumbered” with this book. (Perhaps I should have called the book Outnumbered, but there’s a sitcom in the UK with that title, which put me off it. Anyway, it’s not changing now.)

Till Undeath Do Us Part


The new cover reduces the background distraction at top right and presents the faces a little starker. The font is more dramatic and urgent, more typical in books in the horror genre, and the angle emphasises the urgency.

The tagline hints at the two paths of the story. It’s ominous – and also faintly biblical, matching the title. I considered mentioning zombies explicitly, but felt the Undeath of the title served that purpose.

I considered a very different cover, showing King’s College Chapel in a dramatic silhouette. It matches the other three new covers more closely, but tells you nothing about the story. People shown both covers preferred the existing cover: it links the title to two people, the main characters in the story, and hints strongly at plot.

These faces, incidentally, are part of the incredible, beautiful detail in the chapel’s west window. If you get a chance to visit the chapel, do. It’s a tremendous building.

The Pink and the Grey


Now the real change!

I love the old cover. Two quarters of the shield hint at the story, and it’s full of fun if you look at the detail. But nobody sees any of that. People see a cover with a shield on it – and that’s about it. Nobody sees the central circle as a camera lens. The title, of course, gives little away.

The new cover focuses heavily on the surveillance cameras in St Paul’s College (they’re not the whole story, but it’s a mistake to cram too much into a cover). You can now tell at a glance the plot involves cameras, probably a lot of them, with hints of disagreements (cameras looking at cameras). This cover makes the book look humorous, too. It’s a much better fit for the story than the last one.

The taglines for The Pink and the Grey and The Pauline Conversion have the same feel, reinforcing their shared universe (let’s hope I can keep the pattern going for future books). Using four one-word sentences, with the last word slightly off-the-wall, helps convey the humour as well as the plot.

I have a notebook full of sketched ideas for this cover: despite the simplicity of the final idea it took a long time to get here.


The new covers are now published everywhere (the old covers persist only on sites like Goodreads that track each edition separately). I hope they’ll convince more readers to get as far as page one, where the writing can take over. The truth is I have no idea: I’m far too close to the covers to be objective about them. We’ll see. Come back in a year to see if my new book’s called Nnnnggngg.

Nemo scit aliquid

When I was young, my dad worked for the family business. So did my uncle. Until I was about ten years old I thought that’s what happened: you followed in your father’s footsteps. I distinctly remember the conversation with my mum where she told me I didn’t need to do that: I could do whatever I wanted when I grew up.

It was a revelation. The earth shook, and when the dust settled I saw that normal had shifted.

Normal is what you grow up with. And normality shifts.

It shifts slowly, perhaps, and subtly, but shift it does. Normal for a seventies kid (flares, choppers, strikes) wasn’t normal for an eighties kid (ghetto blasters, BMXes, ZX Spectrums) or a nineties kid (consoles, CDs, slow internet). And normal for today’s children, like my two young nephews, is unimaginably different again. Only Blue Peter has remained constant, the unquenching spirit of Biddy Baxter straddling the generations.

This shift, the result of grinding techtonic plates, is unavoidable. You’d be a fool to try to deny it, or to resist it. Times change. Nothing lasts forever. And this, too, shall pass away.

* * *

The screenwriter William GoldmanButch Cassidy and the Sundance Kid, All the President’s Men, etc — famously wrote in his memoir Adventures in the Screen Trade what he said was the single most important fact in the entire movie industry: Nobody knows anything.

Despite decades of experience and armies of accountants and consultants and advisers, studios still commission expensive flops, and still reject what turn out to be blockbusters. Universal passed on Star Wars. Columbia passed on E.T. And every studio in town rejected Raiders of the Lost Ark before Paramount said yes. (Of course, after Star Wars was a success, space films were commissioned by the galactic ton.)

Goldman’s book was written in 1982. The world was different then: cinema normality has shifted. But Hollywood still operates the same way. Nobody knows anything.

Not just Hollywood. Not just the movie business. The Beatles were rejected several times. Harry Potter was rejected several times. Good TV shows aren’t commissioned, and poor ones are.

Across the creative industries, nobody knows anything.

And yet for the last century the studios and music publishers and book publishers and TV companies have been creativity bottlenecks, cultural gatekeepers. They have been the sole arbiters. They have decided which proto-stars to anoint with their dollars and pounds and which to turn away from like a leper on a street corner. And since they know nothing, those organisations that still remain must simply be the lucky ones. Somehow they have retained enough cash to hold onto their gatekeeper crowns, or sold out to someone who did.

* * *

The media companies began their scramble to dominance in primeval form in the late nineteenth century, and they did so because normality shifted: technology advanced to allow the mass production of culture. What had been a singsong around a piano or a topped-and-tailed trip to the theatre — live performance — could now be a recording, and later a broadcast. Shellac, vinyl, film, radio waves, magnetic tape, shiny discs, and so on. Cultural normality shifted.

And the costs of mass production and the risks of mass failure favoured those with bigger pockets, who became luckier and luckier, until by the end of the twentieth century they stood dominant: a landscape of media leviathans, which we see today as normal.

And it became, to a first approximation, impossible to make a film, or to sign a music contract, or to become a published author. The leviathans were in control, the cultural gatekeepers, because they took all the risks.

Even though they knew nothing.

* * *

The internet came along, and normality shifted again. It is no more an inviolable truth that we have a small number of large, lucky cultural gatekeepers than it is that I must work in the family business, or wear flares, or ride a BMX, or listen to the shrill whistles of a 14K4 modem. The internet has democratised, disintermediated, and revealed an untapped human desire for an infinite number of cat videos.

Of course normality will shift again, in a way nobody can foresee. But today’s cultural gatekeepers won’t remain in their present form. They will crumble into separate services like editorial and marketing, or wither away completely despite extensive lobbying of governments to legislate to preserve their business models, like a monk squatting on a squire until he agrees to buy his illuminated manuscripts and put that nice Herr Gutenberg to the stake.

* * *

Normality shifts, and — across the creative industries — nobody knows anything.

How else can you explain the decision by HarperCollins, upon the book’s 75th anniversary, to publish a new hardback edition of The Hobbit in Latin?