Sunday, 1 July 2012

Biography Bashing

Fact: I spend far too much time on social media.

For a couple of days I've been trying to find statistics on what percentage of the market is celeb-biographies and what is non-celeb. I was curious to find out whether you have to be a celebrity to write a successful biography, as it's a common saying in writing circles that your biography won't sell if you're not a household name. 
You would have thought this would be an easy Ask Jeeves issue by now. But, after consulting Twitter, Wiki and  a 9k-strong writing forum, I was none the wiser.

So, I rolled up my sleeves and tried hard to remember the one post-graduate quantitative research lecture I managed to stay awake through.

I thought that if I blogged my progress it would: 

  1. Push me to continue making progress and 
  2. Add something to the online knowledge base

First problem - where to source data? Especially when I'm not about to fork out to access any large databases. They talk in silly money, and my bank account is only little.

Word association sometimes helps:

Think books -

- think Amazon.

So, off I trundled and, sure enough, they categorise their biography section:

(click to enlarge)

Each time you click on a category, it tells you how many books are listed with that label.

A quick break down of current stock works out like this:

(click to enlarge)

There's nothing in row 1 - that's just me being messy. The number at the very bottom right is the total number of labels: 845,062.

But it's really important to remember that this is the number of labels, not books - as a book might have more than one label. That's why 'general' has so many more than any other label, because books categorised as 'historical' or 'religious' biographies are often also given a 'general' rating. Books on the Holocaust would also come under War & Espionage.

This makes things a bit tricky, so I ignored the 'general' label and this brought the total labels down to: 471,204.

I then decided to go for a trusty visual aid. To try and tidy things up a bit, I cut out everything with a 0% share of the market: any label that's percent of the market is below 0.45% - thus rounded down to 0% rather than up to 1%.

If you're not so familiar with 'rounding', 5 > (five or more '>') is usually rounded up, whereas 4 < (four or less '<') is rounded down. So, with 0.45 you round five up, which rounds 4 up to 5, which rounds 5 up to 1.0 which is 1%. Whereas with 0.44 you round 4 down to 0, which also rounds the next 4 down to 0, which leaves you with 0.0, which is 0%

See how the wonders of maths can cause entire stock piles to magically disappear?

Yes, yes, *yawn*, I know. So, anyway, this means we sadly have to wave goodbye to 'reference' biographies (compilations and photo collections, that sort of thing), 'tragic life stories' (thank heavens for that!) and 'artists, architects and photographers'. British Royalty barely scrapes through with 0.47% of the bookshelf - I'm being generous.

(click to enlarge)

I'm working from as I'm British, and I'm lazy. If anyone would like to throw me a US or EU comparison, that'd be great.

So, as of 30th June 2012, the biggest biography label by far in the UK is Historical (22%), followed a long way behind by Politics (11%) with Religion and Social & Health Issues battling it out for 3rd place (9%).

A few things surprised me. I expected Holocaust (1%) to rank a little higher, with all the interest in war memorabilia. Also with History having such a large share of the market. I also thought Sport (3%) would do better, given the number of sports fanatics and big names - but then I had to ask myself whether sports fans are also big readers...?

Multi-label data isn't the best to use, it's far from clean, but it is interesting nonetheless.

Still, that's the easy part. The hard part, and the question that started all this, is getting a sense of the celeb-biog proportion of the market verses the non-celeb - Joe on the Street - section. Are people mostly interested in reading about big names, or is there a market for everyday people with interesting life stories?

A quick glance at the biography bestseller list is quite encouraging (relatively speaking):
  1. Slave Girl - Sarah Forsyth
  2. Secret Diary of a Sex Addict - Amber Stephens
  3. Against All Odds - Paul Connolly
  4. I Feel Bad About My Neck - Nora Ephron
  5. Open - Aundre Agassi
  6. Confessions of a GP - Benjamin Daniels
  7. How to be a Woman - Caitlin Moran
  8. Life...on a High - Nick Spalding
  9. Diary of a Menopausal Woman - Cheryl Reid, Toby Williams, Lizzi Eastburn
  10. The Sugar Girls - Duncan Barrett & Nuala Calvi

If you class a 'celeb-biog' as one written either by or about a celebrity, then it's not looking as bad as I thought it would. Roughly a 70/30 split in favour of the average Joe.

But it is difficult. Where do you draw the line on celebrity? How famous do you have to be? Well, for my calculations 'pretty famous'. Preferably a household name, living or dead. You certainly need a Wiki entry. But then, what if a celeb writes a book about the history of Norwich? Is it the content or the author that's important for celeb-rating? I say it's still a celeb-biog because sales will increase off the back of the name of the author.

Not easy, huh? The rules need to be clear.

Next question, though - does this statistic hold true through the categories?

Method: choose three random pages from the 'most popular' category in each listing, count the number of celeb/non-celeb titles. Total sample number each category: 36 (12 per page).

Good: Random sample groups
Bad: Not proportionately representative of size of category (why? I just haven't got the time)

I decided to knock out the Royalty section as it was a bit unfairly skewed (100% celebrity) for such a small slice of the market. Thought it would be more interesting to concentrate on the slightly ambiguous labels.  As I suspected, there were quite a few repeat titles between categories.

At first glance, the gap wasn't as big as I'd expected it to be. Though it was a lot smaller than the bestsellers list led us to believe: 45% celeb, 55% non-celeb.

I had this image of biographies, and the biographical market, as being swamped by celebrity culture with no room for anything that didn't take place in Hollywood or involve a dead bard. I wasn't expecting to see a higher average per category for non-celeb biogs.

Little bit of visual magic might assist:

(click to enlarge)

And on that helpful histogram, I can conclude that it's fairly evenly matched, except in the obvious categories you'd expect celebs to dominate: Theatre & Performance Arts and Film, Television & Music. Whereas non-celebs appear to be cornering the expert markets such as Medical, Legal & Social Sciences and the more ethereal realms of Religion.

It's only a small data set, and you'd need to clarify the definition of celebrity a bit - but it's an interesting indication.

No comments:

Post a Comment