Can a computer program help you produce a better book?
I logged on to my paperback printer, Lulu, the other day to find a new button next to each of my titles, labelled ‘Buy a Helix review’. Thinking that this must be a variation on the notorious paid reviews, I clicked to find out how much money my printer wanted to not read my books and still give them five stars. However, it turns out that a Helix review is something different. It’s a computer program which reads your book and gives you a report, so that you can compare your book against “more than 100,000 bestsellers catalogued as part of The Book Genome Project.”
My curiosity and confusion increased in equal measure, mainly thanks to the marketing guff Lulu offered to explain this. According to the site, a Helix report contains metrics about your book under five headings:
1. Words ‘unique’ to your story. Using Moby Dick as an example, they list the words unique in that story. Call me pedantic, but I don’t see how words in common use can be described as ‘unique’ to a story. Certain words may make a story distinctive, but those words themselves are seldom unique.
2. Data worth knowing. You can find out how all of your ‘unique’ words compare with similar titles in your book’s genre and across all bestsellers. In other words, you can beat yourself up by saying “Oh, right, so I’m not a bestseller because I didn’t use enough ‘unique’ words”, instead of the rather more probable “I’m not a bestseller because today’s fiction market is saturated.”
3. Writing style. Now we get to the nitty-gritty. Lots of colourful buttons which Lulu describes thus: “Compiling the metrics of Motion, Density, Dialog, Description and Pacing from your book and comparing them with the same statistics from other books in your genre… is a powerful way to understand and meet the expectations of your genre.” In other words, your aim as an author should be to write as close a facsimile of someone else’s book as you can – for Pete’s sake do NOT try to tell an original story.
4. Story DNA, wherein Helix analyses the thematic elements in your book. This may indeed be useful because, as the author, you will of course have absolutely no idea what themes you wrote about.
5. Comparable titles. Helix will identify and list the titles your book most closely resembles, based on similarities in theme and writing style. Then, presumably, it is only a short step to the angst-ridden author brutally (re)editing their book because, you know, if it’s just a little more like some other book written fifty-odd years ago, it’ll be a bestseller for sure.
I’m a cynic by nature, and I may be sounding a little tough on Lulu’s Helix reviews. However, I don’t see much advantage in shelling out $49.99 (per book) to have a computer scan my words and receive an automatically generated report that is going to tell me my book isn’t selling because it’s too similar to/too different from some other books which were first released in a wholly different publishing landscape. Besides which, fifty bucks constitutes my entire marketing budget for 2014, and I’m not going to blow the lot in one go.
If your aim as a writer is to produce a work for a very specific target audience, it may be helpful to see your book’s content laid out in such a manner. To me, a Helix review appears to be more of a money-making effort: at first glance, it looks informative and slightly sexy; there are lots of tables and charts; and by using modern clichés like ‘story DNA’, it’s clearly trying to appeal to the business person in each author.
However, before buying it must be worth asking if there is any real benefit to be had. Lulu says: “You can use the information provided in the report to refine your work to better meet the expectations of your targeted audience, to draw comparisons with popular authors when marketing your work, or to differentiate your work from current trends in literature.” But if you care enough about your work, it is fair to assume you will have already done some of this research. In fact, such a report may actually work against you by suggesting you should edit your book in a way which your intuition tells you is wrong.
Would that really be worth it?
Interesting post, Chris. I’m curious to see what others think.
Thanks, Al – me too!
If I wanted to write a book just to sell – then maybe. But this makes writing look like an assembly line aimed at generic look-alikes. No thanks.
I don’t think a good content editor can be replaced by a computer program any more than Google Translate can replace translators – neither can take account of the subtleties of language.
This sounded kind of interesting until I read they were gonna charge me for it. I’d put it in the same category as all those “What Hogwarts character are you?” quizzes on Facebook…
Indeed, and we can extend the analogy to “What if I’m not the character I thought I was – OMG!” 🙂
Wow! This sounds really useful! I can’t wait to throw money at it!
Thank you, JK, for a nice drop of New Year’s Day irony 🙂
Cookie cutter – too regimented to appeal for me, like saying nuance or regional or personal differences don’t matter.
Nice layout of the components Chris!
I’d think an exploratory approach, even for beginners, would be more satisfying – for reader & writer.
I guess that leaves the machine, rote production, and God forbid, sanctioned features & sales.
Thank you, Felipe. I think using Moby Dick as an example in the sames pitch worked against them. Great works of literature can’t be compared to what readers want from new, unknown authors today.
I too have a connection with LULU and received the aforementioned ‘Buy a Helix review’ gumph. I am also a bit of a cynic, Chris, and with the words ‘buy and review’ in the same breath I pretty much dismissed it without a second thought.
Excellent article, Chris. Thank you for sharing your in-depth investigative bent.
Thanks, TD. I did think of actually putting one of my books through it, to be able to write a more comprehensive article. But then a voice in my head said “God, no! What happens if it gives you some really awful news?!” and worked those thoughts in above so other authors can make up their own minds if they think it might be worth it.
Great post Chris. Like you I don’t see much value in getting this kind of a review, but it did make my inner conspiracy theorist sit up and take notice. Doesn’t Author Solutions now have some kind of connection to LULU?
Thanks, AC. Yes, Lulu has always had paid services options (right up to $50,000 for the full vanity package) but I use them because they don’t force these services down your throat. The options are there, but you can publish for free, and only pay for the copies you order. The Helix seems a bit more ingenious, like those “writing competitions” where the entry fee is $10 or $25. They’re never going to read your book and you’re never going to “win” – it’s just an easy way for them to make money by pitching the cost low enough so the author thinks it’s “worth it”.
Ugh. Just shows how naive I am – I didn’t realise those comps were just money pits. I’ve only entered 2 so I guess it could be worse, but still. Thanks for the heads up Chris and I hope 2014 is The Year for all of us. 🙂
Well, I paid to enter a couple as well. But since I got nowhere, I can only assume they are cynical money-making operations run by morally bankrupt sharks. But that’s just my opinion 😉
-giggles- how could I possibly disagree?
I think your critique is well stated, Chris. There are writers who will be swept up with the bells and whistles diagrams and catchphrases. That group will form Lulu’s Helix market right there. It won’t last but it’s sure to make them some money in the short term.
Thank you, Jo. It would be interesting to see if any author buys it, edits their book inline with the advice, and then has some/more success, but I can’t see it happening. Which successful author is ever going to admit “Yeah, it was the advice in a Helix review which made my book resonate with readers!”? 🙂
God, that is just nuts.
But it costs a lot less than most of Lulu’s services, and I think they’ll get a few bites
So, just to play devil’s advocate here, I think this sounds like a useful tool that probably still requires some development and tinkering. A 100,000 book database is puny. And the “best seller” concept seems like it would would taint the value of this tool quite a lot.
The real issue to me, though, is cost. I think if I put a thousand or more hours into writing a novel there might be a little value there in getting a Review-Bot to swarm my story and kick back a data profile of all that work. However, a more reasonable price would seem to be somewhere in the neighborhood of $0.99.
One last point: imagine something like this being used by publishers or agencies as a tool to evaluate submissions. If it isn’t already happening, I assure you it will be very soon.
Your last paragraph is a very scary concept.
Excellent comment, David, thanks. I also wonder at the marketing theory behind the $49.99 figure. On one hand, they’ve got to make people think it’s clever and so worth “something”; on the other, if they’d pitched it at the cost of an ebook, say $2.99, I think they would’ve got many more authors giving it a try. What makes me cynical is that the price does seem to be pitched specifically at newbies who don’t yet understand the industry. It’s much, much more about Lulu trolling for easy cash than genuinely trying to help authors market their books.
Now that we’ve got some comments, mostly agreeing with Chris, with the exception of David, I’ll throw out my thoughts. They’re more in line with David in general. I’ll summarize my thinking as maybe not now, but this tool or one like it could be useful, if used correctly.
My day job is in technology which tends to make me pro technology in a lot of ways, yet I’m also a bit of a Luddite in that I tend to resist getting the latest hot tech toy until I’m convinced it is a tool that will actually provide value. This program or one like it has potential, but I’m not convinced it is there yet. There may be other programmatic writing tools out there to analyze some aspect of your writing already that might be ready for prime time already (and I’m sure there are others which aren’t).
Almost any tool like this is going to suffer from two major problems.
How accurate it is in measuring whatever it is evaluating. All programs of this type are difficult because computers don’t do as well evaluating something that is a fuzzy concept which most of the things it is looking at are. In this case it is using comparisons to its database, but as already mentioned, the size and composition of its database are suspect. 100K books isn’t enough and Moby Dick (especially) along with most classic literature isn’t what a modern audience is looking for as a rule. (Kat Brooks and I were had a discussion a couple days ago on this subject which I’m going to talk about in a future post.)
The information given by the program, even if 100% accurate in what it is intended to measure and report, can easily be misinterpreted and misused. The user has to have a certain amount of expertise to avoid being misled.
I’ll make a comparison to the most widely used tool used by authors that is sort of like this tool in that it evaluates your writing and points out errors, but is much simpler to build. That’s the spell checker in Word or other word processors.
The spell checker operates on a database that is actually pretty good. A word it doesn’t flag as in error is almost always (as close to 100% as possible) going to be a real word, spelled correctly. But it may not be the right word I’ve seen some indie books that obviously went through a spell checker, because every word was a real word, but so many of the words were wrong it was unreadable. Word’s grammar checking functions, if turned on, might have caught some of those errors, maybe suggested ‘bare’ instead of ‘bear’, but the user has to have the expertise to know which is the right (write?) word to use.
Thanks for your comment, Al. I would sincerely hope that Helix is quite a long way ahead of things like Word’s spell and grammar checkers. Even if it is, however, I just see way too many potential linguistic problems. How is it going to cope with idiomatic usage, for example? English is a very wide-ranging language – can Helix recognise differences between US, UK, Australian and other variants? What about slang? In addition and as I note in point 3 in the article, even if it can assess and compare differences between dialogue and description in different books, what does (could) it mean by “motion”, “pacing” and “density”?
But the main reason I’ve got little time for this kind of tech comes from my previous jobs. For years I’ve taught English as a second language to some very clever people. But no matter how well they can learn the words, they’ll always struggle with register, idioms, etc. Lots of people mistakenly think Google Translate can answer their problems, but if you use it, you’ll see how hopelessly limited it is. Yes, it can translate one single word quite accurately, but it is absolutely lost when it comes to idiomatical usage. We seldom communicate through single words, but usually through word combinations which can convey all kinds of subtle meanings. Google Translate is not anywhere near understanding language subtleties, and because Google has its reputation, I don’t think someone else is going to come up with a computer program which can handle and process and understand language any better.
Chris, I think we’re on the same wavelength. I used Word’s spell and grammar check as an example because I think most people understand its strengths and weaknesses. Surely everyone uses it in some fashion, but most of us understand where it falls short.
I *think* the tool described would have to have the same difficulties as any language tool. It is looking at a high level and the choices, metrics, etc aren’t as clear cut as spelling and grammar (which still aren’t always clear cut as your examples show). I’d liken what Helix is supposed to do to the kind of things a content editor would do while a spelling and grammar checker are more in the realm of a copyeditor or proofreader. But even if useful all it would do is get the author closer. Usage of a spellchecker doesn’t alleviate the need for a copyeditor and proofreader, right?
I’ve had some fun with Google Translate, feeding it a couple lines to translate from English to another language and then back again. A little of that shows how inadequate it is to the job as subtle differences in words can make a big difference in the perceived meaning.
Indeed, but coming back to Helix, it makes me wonder why $49.99 instead of making it a LOT cheaper so someone like me would give it a try and, who knows, find it useful. By pricing it the way Lulu have done, it means (I think) people are going to be wary rather than saying “I tried it and it’s good at this but not so good at that, etc.” Seems to be self-defeating from Lulu’s POV 🙂
“Seems to be self-defeating from Lulu’s POV.”
Yup. Only being available to Lulu customers (I think that is the case) severely limits their market. If it was available to everyone they could potentially make much more at a cheaper price.
And I think some authors might give it a go out of sheer curiosity.
In all honesty, with a little tweaking, this could be beneficial, especially to new authors. Yes, the price is WAYYY too high and there should be more books to compare with, but I could have used something like this when I wrote my first book and didn’t know which way was up.
I could have used the help with the first book, too.
I think that’s really the main problem, Nicole. If they’d pitched it much lower, we might have some idea of how useful it really is.
Interesting post and comments. I don’t see how a bot could provide this service. It sounds like a critique group, beta reader, marketing analyst and editor all rolled into one.
I heard if you put the word Goddess on every page you can sell millions of books.
And it’s not like any author would really consider using this service in place of betas/editors, etc… Your comment reminds me of that fad a few years ago, when an “entrepreneur” self-published a book called “What men think about apart from sex” and it had 250 blank pages. The guy made a lot of money, too, as I heard 🙂
When you think of all the apps that are on the market for iOs and Android, most priced at 9.99 and less, and all the functionality they include, yes, Helix definitely seems way overpriced. I’m not against paying for something that has good value, but they really need to offer an introductory price to try it on at least one book.
That said, I likely wouldn’t bother with it at any price!