Novel Survey Results, Part III
Update: The full survey results and the raw data are now posted at http://www.jimchines.com/2010/03/survey-results/
Last month I collected information from 246 professionally published novelists on how they made their first pro novel sale. This was rough, Mythbusters-style science. It’s not a perfectly controlled study, but it provided much more data than I usually see when we talk about these things.
I’m wrapping up my results, and will be working on compiling everything into a single essay, to be posted on my web site along with the raw (anonymized) data. Today I’ll also be examining the weaknesses of my survey, as well as other data sources for those looking to learn more.
Can You Boost Your Odds?
Survey Flaws
Other Resources
Final Thoughts
Can You Boost Your Odds?
We looked at networking and the myth that you have to know somebody in Part II. The idea that you must have connections to break in? Busted. But as some people (including my agent) have pointed out, connections can still be helpful. I wanted to know what other steps authors took to try to improve their chances, and asked whether participants had done any of the following:
- Attended conventions
- Attended one or more writers groups
- Earned an undergrad degree in English/Writing
- Earned a graduate degree in English/Writing
- Attended a weekend writing workshop
- Attended a week-long writing workshop
- Attended a longer writing workshop
- None of the above
By far, the two most popular choices were conventions and writers groups, both of which were reported by more than half of our novelists. The least popular choice? The graduate degree in English/Writing. (As someone who holds an MA in English, I’m trying not to be depressed about that one.)
The full breakdown looks like so:
Remember, this is correlative data, not causative. However, I decided to take a look at a few more correlations, taking the writers from each of these categories and examining how many years it took to make that first pro novel sale. I bolded the highs and lows.
Full Group: Average 11.6 years, median 10, mode 10
Conventions: Average 10.5 years, median and mode unchanged
Writers Groups: Average 10.5 years, and median drops to 9.5
Undergrad Degree: Average 9.8 years, median 6.5, mode 3.5
Graduate Degree: Average 11.8 years, median 10, mode 6
Weekend Workshop: Average 10.7, median 8.5, mode 3
Week-long Workshop: Average 10.7, median 8.5, mode 6
Longer Workshop: Average 11.6, median 10, mode 6
None: Average 15.7 years, median 15, mode 9
I’m reluctant to draw too many conclusions from this, or to say that any one category will definitely help you break in. But looking at the “None” category, I think it’s safe to say that writers who are more actively trying to get out and build their careers — in any one of a number of ways — tend to break in faster than those who aren’t.
Survey Flaws:
This was not a perfect study. It wasn’t meant to be. I wanted a large enough sample to start to see some trends, but I’m not qualified to run a full-scale, controlled study. Nor do I have the time. In the interest of full disclosure, here are the flaws I’m aware of.
1. Sample bias. I’m a fantasy author. When I announced the survey and asked for authors to participate, I knew the results would be heavily skewed toward SF/F writers in my network. I did some outreach to spread the word to other writing groups and blogs, which helped, but the results are still weighted toward SF/F and may not apply as strongly to other genres.
2. Question imprecision. Several questions were imprecisely worded. For example, one question asked “How many times, if any, was your novel rejected before it sold to a professional publisher?” I received enough comments and questions about this, asking whether I meant publisher rejections, agent rejections, or both, that I did not include the final data in my write-up. I’m also unhappy with one of the networking questions which asked if you were introduced/referred to your agent or editor. “Referral” is fairly broad, and could mean everything from a personal letter of recommendation to an author saying “Oh yes, Bob’s my agent and I think he’s open to queries right now.”
3. Can’t prove cause/effect. This is a weakness of correlative data. I think the data worked well for busting certain myths, but if I catch anyone saying things like “Jim Hines proved that if you get a degree in English, you’ll sell a novel faster,” then I will personally boot you in the head. See here for a good example of correlation =/= causation re: pirates and global warming.
4. Limited scope. I restricted this survey to authors who had published at least one novel with a professional ($2000 or higher advance) publisher. Not everyone shares the goal of publishing professionally. For those who prefer the small press, non-fiction, script writing, short fiction, or other forms of writing, the path to breaking in might be very different.
I’m sure there are other flaws. However, it was my goal and my hope that even with these problems, the data I gathered would be useful in talking about how writers break in, and would be much better than the anecdotal “evidence” usually cited in such conversations.
Other Resources:
Tobias Buckell’s Author Advance Survey: Data from 108 authors about novel advances, showing trends over time and over the course of authors’ careers.
Megan Crewe’s Publishing Connections Survey: Data from 270 authors on whether you need connections to break in. Her results tend to match my own on this one.
SFWA’s Online Information Center: Includes essays, resources, and advice for new writers from the Science Fiction & Fantasy Writers of America. (Thanks to Charlie Stross for the link.)
If you have other examples of studies/resources like this, please let me know. I’d love to have more to share for the final write-up.
Final Thoughts:
My thanks once again to everyone who participated in the study, who spread the links to other writers, and for all of the support and encouragement. I’m quite pleased with the way this turned out, and I hope it’s helpful to others.
In conclusion (and in true Mythbusters style) I present you with this artistic rendering of my editor when she learns how much time I’ve spent on this survey instead of working on my next book:
Lynn Flewelling
March 24, 2010 @ 10:29 am
Great work! It’s been really interesting seeing the results. Not always what I expected. Sure makes me wish I’d known about conventions *before* I got published!
First (Pro) Novel Survey Results at SF Novelists
March 24, 2010 @ 10:40 am
[…] http://www.jimchines.com/2010/03/novel-survey-results-part-i/ http://www.jimchines.com/2010/03/novel-survey-results-part-ii/ http://www.jimchines.com/2010/03/novel-survey-results-part-iii/ […]
Jim C. Hines
March 24, 2010 @ 10:57 am
“Sure makes me wish I’d known about conventions *before* I got published!”
You and me both 🙂
Ken Schneyer
March 24, 2010 @ 11:17 am
Again, this is fascinating, and I’m delighted to read it.
But again, with all the “improving your chances” data, you haven’t actually even established a correlation, because you don’t have a control group. Unless we know what percentage of those trying to sell novels who met each one of these categories, we have no notion of whether it appears to improve your chances.
I am fascinated, though, by the datum that those with undergraduate degrees in English or writing showed a markedly lower median success time than all the others. I know it’s not causative, but it really does make me wonder about what could be bringing about this effect.
Tweets that mention Jim C. Hines » Novel Survey Results, Part III -- Topsy.com
March 24, 2010 @ 11:36 am
[…] This post was mentioned on Twitter by pauljessup, Lisa Shearin, Jim C. Hines, MG Ellington, Michelle Muenzler and others. Michelle Muenzler said: RT @jimchines: New Post: Novel Survey Results, Part III http://bit.ly/atxdaG […]
karen wester newton
March 24, 2010 @ 12:46 pm
Fantastic idea! Thanks so much for doing all this work and sharing the results.
Steven Saus
March 24, 2010 @ 3:12 pm
Ken, you mean he hasn’t established causation. He’s noted that within the sample, more authors have attended conventions. Whether or not this had any effect on them being published is not demonstrated.
Unfortunately, as Jim noted earlier, the “number of rejections” question can’t be used (because the question was not answered uniformly), otherwise we could run regressions against that and see if these factors actually improved chances of sales.
Satima Flavell
March 24, 2010 @ 7:18 pm
A fascinating project, Jim. Many thanks.
conrad steenkamp
March 26, 2010 @ 2:34 am
Absolutely excellent! Thanks a lot for this. It really clears up what has been quite a muddle in my mind up to now. I am busy doing a second draft of a SF novel and am looking around for the way forward. Perhaps you could also do more detailed research about the submission process! Thanks, Conrad
Jim C. Hines
March 26, 2010 @ 7:31 am
Thanks, Conrad. I’m a little burnt out on the research at the moment, and should probably be working on my own novel, but there’s a fair amount of info out there on the submission process. If you’ll be querying agents, I’d look for some of the agent blogs. A few of them even post and critique query letters, which would have been incredibly helpful had it been around when I was trying to break in.
Steven Saus
March 26, 2010 @ 8:09 am
As the guy who’s doing some secondary analysis on Jim’s data, let me second what he said. Just to break down my involvement, it’s been about four hours of hard-core analysis work (running stuff through SPSS, figuring out what I needed to look at further, figuring out what wouldn’t work, and the like) plus another two hours of work just formatting the results I’ve had so far and presenting it properly.
(I’m not whining – I volunteered to do this. Just pointing out that it’s not a trivial time investment – and I’m doing the *easy* stuff. Jim already did most of the hard work.)
Of course, if SFWA (RWA, etc) wanted to pay me to do the research, I’d be thrilled.
Jim C. Hines
March 26, 2010 @ 8:11 am
Sheesh — what is it with writers always wanting to be paid for stuff? 😉
Writer Wednesday Blog Tour #11 « By W. J. Howard
March 31, 2010 @ 7:24 am
[…] Novel Survey Results, Part III […]