Authors A.I. Marlowe Analysis
Recently I was made aware of a relatively new analysis tool for authors: Marlowe. Marlowe, created by authors.ai, is an artificial intelligence bot that reads, analyzes and evaluates patterns and trends in a novel, providing a report that is based purely on algorithms and is, therefore (rare in the writing world), not subjective. She compares your novel’s characteristics to hundreds of best-selling novels and provides insight into multiple aspects like pacing, story arc, character personality, dialog, and spelling and punctuation.
Want to see what such a report looks like? You can read a Marlowe analysis of Dan Brown’s The Da Vinci Code here.
I was curious to see how Marlowe would analyze one of my own books, so I chose the one-time report ($89 on the webpage) and uploaded Finding Travis, one of my time travel stories. True to their word, the report came back within an hour or two, 25 pages of graphs and charts with full explanations for how to read them and apply the results to my book. Since my book has done well and has amassed many positive reviews, I was expecting good results, and I was not disappointed.
I won’t detail the entire report here, but will hit a couple of high spots. The first graph is that of narrative arc vs. plot arc.
In the graph above, the green line is the narrative or story arc, which shows the basic journey through the story. The dashed line is emotionally neutral, so you can see that Travis’ story starts at a low point in his life, builds to a hopeful peak, then is dashed to lower than low before rebounding. Marlowe nailed this exactly. The purple line shows the plot turns within the story, and you can see that Travis has a lot of ups and downs along the way. According to Marlowe, the more peaks and valleys—and the greater height and depth of them—indicate greater emotional intensity, which is a characteristic of a successful page-turner. Marlowe suggests that good stories will have a major turning point at approximately the 25% mark, and will have a major peak or valley at roughly 50%, propelling the MC into the final conflict and resolution. I think I came pretty close to both those marks in my book.
The second analysis is the one of narrative beats, or pacing. These can be times when new conflict is introduced (negative beats) or when conflict is resolved (positive beats). According to Marlowe, best-sellers have a regular rhythm throughout the length of the book, roughly every 10% or so.
While the beats in Finding Travis are not perfectly spaced, they are fairly regular, so I was happy with that. In addition, Marlowe gave me several examples of where these positive and negative beats occur in my book, quoting sentences from each occurrence.
Now, just to be thorough, I decided to compare my graphs to the ones for The Da Vinci Code. Here’s the one for the arcs.
While obviously not the same, I do think Dan Brown’s arcs are similar to mine, the narrative arc in particular. I found it interesting that his major plot downturn was at the start of the book, while mine was closer to the end, but we both have similar ups and downs.
The beats in The Da Vinci Code are also similar to mine; not quite every 10% but close, and fairly regular across the length of the book. I was happy to see that.
Some of the other aspects Marlowe looks at are characters’ personality traits (confident, happy, perceptive, responsive, physical, etc.); major subjects mentioned in the book; explicit language; clichés used; repetitive phrases and frequently used adverbs and adjectives. There’s a section on narrative time vs. dialog time, and Marlowe suggests that popular novels have dialog between 25 and 35% of the book. Finding Travis came in at 27% dialog, a good place to be. Another metric was complexity, based on the length of sentences. Marlow says popular books have an average complexity between 2.0 and 3.0, and my average was 2.54. There is also a section on punctuation used, and one for possible misspellings, so pretty much everything from the abstract to the mundane, soup to nuts, is covered.
Want to try out Marlowe? There are several ways to do that. The one I chose was the one-time report for $89. There are two ways to subscribe, a basic plan for free and a pro plan for $29.95/month or $199/year. The basic plan allows you to get unlimited reports, but those reports are not nearly as comprehensive as the one-time or pro reports. For example, the two aspects I focused on here, the narrative arc and the narrative beats, are not included in the basic plan. You can see a full comparison of the plans here.
Now my foray into Marlowe was purely based on curiosity; I wanted to see how my book stacked up to her analysis. I’m actually not sure how helpful it might be for making changes to a book, or for tweaking the arc, etc., but I was gratified to know that my book was within bounds for a well-told story.