AAPL Apple Inc

By Katie Deighton 

Media companies used to rely on voice artists or basic text-to-audio software to turn written articles into audio stories. Now they are developing humanlike reader technology and training reporters in the art of reading out loud to help people with overflowing "to-read" lists.

The BBC, Apple News+ and The Washington Post have in the past month rolled out new ways to listen to their written articles, hoping to give busy subscribers a flexible way to explore stories and to attract new subscriptions, executives said.

"We conducted user research and learned that users want to stay informed but are busy, so they appreciate an option to get up to speed on the latest news developments while cooking dinner, running errands or exercising," said Emily Chow, director of site product at The Washington Post.

The Post said it began producing audio articles as an experiment "several years ago," but text-to-audio or click-to-listen story formats have been available for over a decade.

The Economist Newspaper Ltd. began producing an audio edition of its weekly magazine in 2007 and little has changed since then, said Tom Standage, the company's deputy editor and head of digital strategy. Professional newscasters record every story in the magazine in a weekly session.

The Economist sees audio articles as a way to retain subscribers and find new ones by publishing select content for free as podcasts, said Mr. Standage, who noted the offer "is not about advertiser revenue."

Unlike podcasts, which are often free and include advertising, publishers tend to keep most of their audio articles ad-free and behind a paywall. They are also cheaper to make than podcasts because the reporting has already been done and they don't need production add-ons such as music.

Apple Inc.'s Apple News+ last month began producing around 20 audio stories a week across its portfolio of partner publishers, which include Esquire, New York Magazine and Time. It has also developed a user interface that tracks how far someone has read in a piece and lets them start listening to it in the app from the same place, and vice versa.

Apple will create audio articles predominantly for feature-length reported pieces rather than breaking news, according to the company.

The New York Times Co. has also based its read-aloud strategy around longer, narrative stories. It began producing audio stories last fall with Audm, which provides publishers with audio recorded by voice artists.

The Times in March acquired Audm, which will continue to work with publishers such as The Atlantic and The New Yorker. The Times' target audience is not necessarily news addicts, but more likely readers without time to get through stacks of magazines on their coffee tables, said Stephanie Preiss, vice president of TV and audio at The New York Times.

"This is a way in which long-form journalism can fit better into your life," she said.

Other publishers are trying audio for shorter articles as well.

The Washington Post, for instance, initially used reporters and voice artists to create audio articles. Now, the Post is using text-to-speech features on Android and iOS mobile operating systems, allowing the newspaper to offer an audio version of every article it publishes, including quick news stories, Ms. Chow said.

The BBC's Global News has gone a step further by developing an automated "voice" as part of the click-to-listen strategy it began testing last month. Its developer team built a reader bot meant to sound as human as possible.

"You can't have somebody producing a new audio version of one article every time it's updated," said Andy Webb, head of product for the voice and artificial intelligence team at the BBC. "But with this synthetic language, there's hardly any additional cost to production at all."

The BBC, which does not charge for access to its content, will make the audio versions of its articles available free with ads that run before them, and possibly later as part of wider sponsorship deals. It will spend the next three weeks studying how readers in a test group respond to the voice of the reader bot, which is of a man with a soft northern English accent.

The BBC aims to eventually own a stable of synthetic voices with different accents, genders and nationalities to appeal to its global audience, Mr. Webb said. It also hopes to introduce voice "moods" to match the tone of stories, toggling between a peppy voice for a sports report, for example, to a serious voice for a feature on the coronavirus pandemic, he added.

Not all publishers are ready to let go of voice talent. Apple's operating systems include text-to-speech technology that users can apply in a variety of ways, but the company hired an audiobook director to cast voice artists for its Apple News+ audio articles.

The New York Times has also stuck with real human voices. Along with its work with professional voice artists, Audm is training reporters and columnists to read their own stories.

The Times has no plans to automate the audio article system, said Ms. Preiss.

"We're interested in building products worth paying for," she said. "We feel like 'worth paying for' doesn't include computer-sounding voices."

Write to Katie Deighton at katie.deighton@wsj.com

(END) Dow Jones Newswires

August 12, 2020 06:14 ET (10:14 GMT)

Copyright (c) 2020 Dow Jones & Company, Inc.