June 01, 2015

NEXT POST
How curation automation is going to disrupt content consumption The best content curators have extensive topic knowledge and a knack for reader interests and preferences. That sounds like something only a living, breathing human can do, right? While that’s largely the case today, I believe technology will drive the biggest advancements in content curation tomorrow. Narrative Science is a terrific example. I met Kris Hammond of Narrative Science a few years ago when he spoke at a Tools of Change conference I helped produce. If you’re not familiar with them, Narrative Science is one of those companies that develop tools to automate story writing. You may have read a computer-generated article or two this week and never even realized it. Think you can tell the difference between human- and auto-generated content? Stick around and take the quiz at the end of this article… Data is at the heart of the stories generated by Narrative Science but what exactly is “data”? In the current model, data typically consists of numbers, tables and other highly structured information. For example, the narrative summary of last night’s baseball game could be auto-generated using nothing more than the game’s box score, the data from the event. As platforms like Narrative Science’s evolve, so will the definition of data. Last week I wrote an article about why all-you-can-read subscriptions need curation. We’re drowning in a sea of content and we need better tools to help us uncover and consume the must-read content. There’s a big difference between what you and I consider must-read though and that’s where the curation element comes into play. A number of industry pundits criticized my thinking and pointed out the high cost of this sort of curation. I agree. Curation today almost always requires human intervention. But what happens when that’s no longer the case? What happens when an application is able to rewrite and summarize the sea of daily content that’s most important to you? What happens when this tool, which knows your interests, your job responsibilities, etc., is able to deliver a fully-automated Cliffs Notes version of everything you need to read that day? I think that will be a game-changer and will become an extremely important, real world application for artificial intelligence. Will it put writers out of business? No, not necessarily. After all, most of the original content still has to be written by someone. But it will help amplify the content that needs to be read, enabling it to rise above all the noise that surrounds it. Still think this is nothing more than sci-fi and wishful thinking? Take this short quiz and see if you can figure out whether each of these excerpts were human-generated or computer-generated.
PREVIOUS POST
What questions do your reader analytics need to answer? In my book publisher days I recall saying the following to our Amazon rep: “You guys are capturing a ton of reading data from our customers. When are you going to start charging us to access that information?” She looked at me like I just arrived from another planet and declined to answer the question. A few years have passed since that encounter but some things never change. Amazon is still the dominant ebook retailer and they continue hoarding reader data, sharing only bits and pieces from time to time. I’m still convinced once day they’ll offer a detailed reader analytics service to publishers…for a price. The data will be anonymized, of course, but it will benefit publishers by shedding valuable light on reading habits and preferences. In the mean time, Olive Software, the company where I serve as director of strategy, is in the process of revamping its ebook reader app. We want to take our analytics to the next level and we’d like your input. You see, at Olive, we believe in providing publishers with every bit of data about their readers and we do so at no additional charge. As a former publisher these are the types of questions I would want the data to answer: How many people opened the book they bought? Did the typical consumer read from beginning to end, in chronological order, or did they jump around a lot, reading out of sequence? When readers didn’t finish the book, at what point did they tend to abandon it? What are the most popular phrases searched for when reading the book? I’m sure there are plenty of other questions publishers, editors, marketers, etc., would love to see answered with analytics. What are the questions you need data to help answer? Click here to email me the reader behavior questions you’d like analytics to answer. I’ll gather all the input and will summarize it in a follow-up article. That’s probably yet another thing Amazon would never do for you. :-)

Joe Wikert

I'm Chief Operating Officer at OSV (www.osv.com)

The Typepad Team

Recent Comments