Checking the first scatter proofs of the Fluxx Book (Scatter proofs are where you test the printing, to see if images and design elements will work)

How we turned our Medium stories into a beautiful paperback book

We wanted to find out if making a book could be a fast, simple experiment, not a big slow project. This is how Fluxx made our second book happen with the help of one startup, two local small businesses and a talented freelancer.

In 2015 Fluxx published our first book Unthinkable. We did it the traditional way; writing 178 pages from scratch, a publisher in New York to handle printing, selling through Amazon (we’re currently number 256 in the Books/ Business/ Management/ Creativity chart). It was a huge investment of time and a real success in terms of meeting interesting people.

For our second book, Whatever happens, we don’t want people to write to the Daily Mail, we wanted to try something different, to live by the advice we offer people every day. Instead of a big investment upfront, we decided to create a minimum viable book. Something small and simple, just enough to explain what we do and get our stories into people’s hands.

Could we produce a book locally, working with people like us? How quickly and cheaply could we do it, while still producing something we were proud of? What would a Minimum Viable Book look like?

Bored already? Just want to read the book? You can download a copy here.

Step 1: Start writing

At Fluxx, we tell stories all the time. We help organisations change their internal stories (here’s a classic piece about sandpaper, whiskey and how big companies tell stories).

Our process of running experiments, spending time with customers and incubating businesses tends to generate a lot of anecdotes. When we re-tell those stories, we magnify the impact of all that activity. We also write them down.

We started Fluxx Studio Notes in October 2015 with a writeup of a Dorkbot meetup in Limehouse Town Hall.

Since then, we’ve had about half a million views from a few big hits (Paul’s lifts, my Pricing bit, Alice’s Dubai memoir, Nic’s ethical future for banking) and a few flops (how is it possible that only 2,515 people have read Minimum Viable Cat?) and one piece that made everyone in the office smile.

If you’ve been writing on Medium for a couple of years, you probably have enough material for a great book.

How many words do you need? Our 64 page A5 book contains about 18,000 words. Normal 1,200 word features are spread across four pages. You could squeeze in more words with a denser layout.

Step 2: Why you want to make a book?

We’ve definitely met interesting people and got work through our Medium posts. It’s a great platform to reach people like us around the world.


It’s not quite the same as having a coffee with a potential client or someone you want to work with, reaching down into your bag, pulling out a copy of your book and dropping it on the table.

The idea of turning a bunch of Medium posts into a book was hard to visualise; how much work would be involved? How many pages did we need to feel like a real book? Would it make any sense?

The prototype — bound together with clips — and the finished product.

Step 3: Make a prototype

So, in December 2016 I opened a blank InDesign file and made a A5 template with a column of 9pt text (Droid Serif 9 with 10.7 spacing, copying the text style another book).

Copy/pasting articles from Medium, it started to come together and feel like a rather ugly book.

Most of the articles worked without pictures. Some were more difficult — Alice’s piece on Louis Theroux is illustrated with animated gifs, which are tricky to print and would require clearance from the BBC, so we bought a portrait from Rex Features and had to rewrite the copy.

The pages started to fill up with pull quotes and highlights and lines and boxes. The idea is that if you open at a random page, you’ll find something interesting enough to pause on, then flip to another random page. We wanted to reduce the amount of time required to get something of value from the book from hours to seconds.

I printed the pages out on an old single-sided colour printer, folded them in half and used binder clips to make a little book.

Suddenly, it felt almost real, and was something people could pick up and flick through. I lent the prototype to other people in the team borrowed it and it returned full of scribbles, question marks and post-it notes.

For an A5 book to feel like a ‘real’ book, you probably need 64 pages, perfect binding (square and glued, rather than stitched or stapled) and 120gsm paper, which is thicker than pulpy paperback paper. That will still be very skinny — even a slim paperback has 200+ pages. Our book has 64 120gsm pages, a 240gsm cover and a 120gsm slip cover.

Step 4: Find a physical book you like

So far there were only two certainties about the book. It had to be

  1. Small — we didn’t have the time or inclination to write a huge tome; a 64 page book is probably 90% as impressive as a 256 page book, but only 25% as much work.
  2. Cheap to produce in small quantities — this wasn’t a book to sell for money, but a calling card. Lovely hardback books are relatively affordable if you make 10,000 of them in China, but that wasn’t our plan.

Beyond that, there was a bewildering array of possibilities; from traditional pulp paperbacks to A4 colour magazines. It was important to narrow down the options.

In magCulture, a great shop selling short-run magazines from around the world. I found a book/magazine called Paper, produced by Erik Spiekermann, the German typographer who redesigned the Economist and wrote Stop Stealing Sheep.

Paper is A5, had a bright flouro cover and just 64 pages. Inside, it looks like a book, not a magazine. It fits nicely in your hand and is full of funny stories.

The prototype and a copy of Paper were useful physical props. They meant I could go to the partners at Fluxx and say “This is what I want to make, can I have some money?” They said yes.

The Risograph machines at Hato Press, London.

Step 5: Find a printer

Paper was printed on a Risograph machine. It’s curious eighties-era Japanese printing press that looks like a photocopier but uses brightly-coloured soy-based ink.

Gif via Hato Press

Normal printers use CMYK (Cyan, Magenta, Yellow and Key — Black) dots to create colours. Risographs print in layers of differently coloured ink. The output feels more like a screen print than a slick laser print.

Books printed on Riso machines feel a bit hand-made, because they are a bit hand-made.

Hato Press is a Risograph printing studio a short ride on the 55 bus from Fluxx’s office; they make books and posters on their Riso MZ 770E machines, and were prepared to help us through the process.

Riso printing feels lo-fi and slightly DIY, but it’s not the cheapest way to print a small book. A service like BookPrintingUK will make 200 black-and-white A5 books for under £300. The quality looks pretty good and the turnaround time is fast, but the output doesn’t feel very distinctive — that’s why we used Hato Press for this project.

Step 6: Find a designer

Designing for print is an old-school specialist craft skill, and should be done by someone who loves fonts and ink and dots, understands things like paper grain and sends emails saying “I have added 1mm to every fold line to allow for folding.”

Fortunately, a friend introduced Fluxx to Jay Prynne, who did all those things for us.

Jay created the ten different templates used in the book. It takes some discipline to work within templates; cutting our copy to fit the space, rather than re-designing every page to fit the copy, but it meant we could pay Jay for design, not for fiddling with our words.

Step 7: Find an editor

It took a couple of weeks of cutting, pasting, editing, and working with Jay to complete the 64 pages. We then sent the InDesign files to Annabel at Whitefox — a London startup that provides all the services that old-school publishing companies used to deliver, from editing to PR to printing.

The books being bound in Hackney by ChapmanWiley.

We used Whitefox to get the whole book proofread and sense-checked by a professional. This cost less than £300. They found lots of little mistakes and a few big ones, and made the whole document feel much more professional and consistent.

Step 8: Wait

Finally, with a bit of hesitation and trepidation, we sent the files to Hato Press. Jay’s expertise helped get past technical pitfalls with the eccentric Risograph process (“There is no need to halftone your artwork files, the Riso will create a halftone automatically.”)

We had a hard deadline for the book (hundreds of people invited to a launch party) so watching the book come together, element by element, was painful; at one point the flouro orange ink on the covers had to sit and dry overnight before anything else could happen.

Step 9: Celebrate

The books were ready after what seemed like weeks but was actually a few days, and just in time for our book launch.

As with any of our projects, what we did wasn’t as important as what we learned – where to invest, where to scrimp. When to do-it-yourself, when to hire a professional.

And how doing something properly doesn’t have to mean embarking on a long expensive project.

You can download the full digital version here. And if you’ve published a book from your Medium posts, get in touch, and we’ll swap you for a physical copy of Whatever happens, we don’t want people to write to the Daily Mail.

Previously: 52 things I learned in 2016, We’ve been running business experiments since 2011, this is what we’ve learned.

Tom Whitwell is Senior Consultant at Fluxx, a company that uses experiments to understand customers, helping clients to build better products. We work with organisations such as Atkins, National Grid, the Parliamentary Digital Service and William Hill.