Data vs judgment

With the adoption of the digital-first strategy in many newsrooms, social data is increasingly being used to assess story performance, modify placement of stories, headline selection and so on. In fact, postmortems of news articles conducted by editorial teams also include discussions on time spent on individual stories by website visitors, shares, tweets etc. While hiring data scientists - who can take data and turn it into pageviews - may still be some way off for media houses in India, experts discuss the merits and demerits of the traditional and algorithm-based decisionmaking

Don't let analytics overrun the newsroom: Shane J Cummings

In this era of data, there are two very different reasons to analyse content, and this, coupled with the traditional editor-selected content model, means we are experiencing schizophrenic newsrooms. The first reason to analyse content is an attempt to tap into what the public want. The second is to understand what content works and how it can integrate with or assist with advertising. Commercial content is imperative to ensure the media's ongoing survival. Given that inflexibility, let's look closer at the first reason and more intriguing reason to analyse and measure content.

Last year, News Whip conducted an experiment that compared actual newspaper front pages (the editor's content picks) with the most popular content as measured by social media reads. The results are striking. Where editors of the Wall Street Journal looked at international affairs, WSJ readers preferred to read about US domestic issues such as veteran's pensions. Instead of a media-focused story about TV ratings (it's a weakness of many editors to focus on their own industry, the media), readers read a whimsical human interest story about rappers and dragons. Local and quirky most often won out with readers over international and overtly political.

In many of the examples, readers eschewed the editor-picked stories about international politics in favour of local issues. The notable exception was the Irish Times, where readers chose to peruse a story about the worsening Ukraine conflict rather than national debt. I can't blame them for that: war worries me more than accountants.

While News Whip's experiment is fascinating and proves editors and readers don't necessarily think alike, there is one weighty consideration editors have that readers do not: history. The media has the responsibility to serve the public interest, and part of that is chronicling the important issues for future generations. The public interest for publishers these days extends to light and fluffy celebrity gossip. Most of the time, Kim Kardashian will score better analytics than Kim Jong-Un, but celebrities are ephemeral, and history will be better served by reporting on issues of international significance.

There will always be a place for measuring data and content shareability, and with the blending of publishing commercial models and blurring divide between editorial and sales, this is inevitable. Publishers have mechanisms in place already to cater for 'most popular' content through the use of designated sections on the homepage (and it could be argued editor-selected content has the same opportunity with an 'editor's picks' content section). However, we owe it to our children and grandchildren not to let analytics completely overrun the newsroom.

There must always be a place for content selected by the editorial gatekeepers to speak to future generations, not just the current one and whatever trends it follows today.

Shane J Cummings
Publisher, Brimstone Press

 
A blind data-led approach can commoditise news: Prasanna Singh

In the current environment that we have in India and elsewhere, it doesn't need data analytics to impact editorial control, or editorial gut feel for that matter. I believe that editorial control was already wearing down under attack from other forces, not the least, commercial pressures. For editors, data analytics offers a final chance to preserve a degree of discretion and some control over their work. By embracing data analytics, they can build a better than working understanding of it. Resistance to it would simply mean being considered behind the curve in the new news ecosystem. Editors who can combine their experience, or gut feel with insights from data analytics will have an opportunity to do their stories better than their competitors, creating a real opportunity for their products to stand out with unique perspectives.

After all, what we call 'gut' feel is usually backed by extensive experience of understanding a story's value, especially readers receptivity to it. The big sales challenge is, of course, to monetise this more effectively. Surely, a better alternative, as compared to suffering 'advice' led by the self-styled'market experts' in news organisations - the sales and commercial departments.

Another key aspect is that audiences have exploded even as data tells us that their attention is divided over many more options. What that means is that editors need to define their readers much more sharply and focus on the appropriate data inputs to meet their expectations. This does not necessarily need a reduced audience as the ability to reach virtually anyone who is receptive, anywhere and anytime, is a huge multiplier now. A blind data-led approach to please all is only going to further commoditise the business. Something that the horizontal portals are discovering. 

Yes, there is a real fear that data analytics might cause profound changes in the news we read. And not all of them would be desirable. For example, data analytics will rarely throw up 'bad news' like everyday crime as a data 'winner'. Unless it is an abomination like the Nirbhaya rape case, for example. We can already see the consequences in the form of some news organisations bestowing on themselves a 'duty' to be 'positive', and even share 'happy'stories, since that is what data points to as popular with readers.

For editors and news organisations, it will take a fine line of balancing between data-led output, and at least some output that is created because it matters, or is considered important to be placed in the public domain. Perhaps, this will take the form of a ratio like say 80:20. We need to accept that in something as volatile as the news business, one must always leave space for a contrarian approach. This might be the best way to create a differentiator and keep up with the 'breaking news' pressure.

Finally, keep in mind that different news media have different strengths in terms of creating big stories. A medium like television is far more likely to ignite interest in say, a massive traffic pile up or raise an issue to a new level with a live discussion, while print is far more suited to exposes of corruption in detail and other more research-backed stories. Digital, of course, offers tremendous opportunities for serendipity. For editors today, this probably means experience matters even more to ensure that they have an opportunity to attract the clicks, and not follow them.

Prasanna Singh
COO, Afaqs.com

Data can only indicate, not dictate: Mariam Mammen Mathew

People are constantly changing the way they access information in this digital age. Therefore, it's imperative that the readers' appetite for social media, emails and reading apps continue to be satiated. It's no more just the debate between the traditional and online media. Thankfully for India, both have their own niche and coexist despite the "if the news is important, it will find me" trend among news consumers.

For the news industry, the necessity to carefully segment and target, in order to provide personalised experience to the reader/viewer has taken precedence. Enter data journalism. Publishers and news channel owners are increasingly feeling the need for journalists to develop a high-level understanding of the consumers, including their behaviours, consumption patterns and preferences. Data journalism plays an important role here in processing the vast amount of disparate 'bits and bytes' that exist and churning them into conceivable, usable and analytical information to build repositories for current and future use.

Although data churned out by analysts can be used to improve the credibility of stories at many levels, details in isolation lack the capability to breathe life into a developing story. That is where creativity and instincts kick in. When the numbers dug out by our data analysts join forces with the content and creativity of journalists, the reader benefits. At the same time, media players must know when to rely on data and when to take a new, pathbreaking approach. Maturity lies in the ability to comprehend that data is influenced by many external factors and essentially, data can only indicate, not dictate.

At Malayala Manorama, we are extensively using data analytics to provide an enriching and compelling experience to the reader. Data journalism is one of the latest exciting ideas we are experimenting with. We have tied up with Adobe Marketing Cloud to drive our digital transformation initiatives.

Mariam Mammen Mathew
Chief Operating Officer, Manorama Online

A unified newsroom is the need of the hour: Jehil Thakkar

With the rise of an omni-channel reader who consumes content on multiple devices such as desktop, laptop, mobile, along with the newspaper, media companies that adopt a digital-first strategy are set to gain from increased readership. Analytics is increasingly used in newsrooms now to understand what kind of stories are widely read and to get maximum clicks and shares. It has become a critical metric in companies opting for a digital strategy. The story format is also getting customised as per the device used by the reader. For example, short-format stories are best suited for mobile medium while detail analysis pieces are preferred for print.

Since news has limited shelf life, publishers are now looking at the right target audience for maximum coverage. They are studying various parameters like how long a reader stayed on the page, how many times the page was shared, is it the right story length for the given medium and how many crossovers it led to and so on.

Creating a unified newsroom is the need of the hour. Media companies need to cut down silos and all the concerned departments and vertical heads need to come together to approach digital in a unified manner. The focus should shift to localised and customised content. Remember mobile is a personal medium and requires a customised approach.

Media houses are sitting on a huge pile of data but so far they have not leveraged it. Now some newspapers have created content for specific geographies.

The downside of algorithm-driven decision-making is it can create a reader-led newsroom. Apple's former CEO, Steve Jobs said, "A lot of times, people don't know what they want until you show it to them." Similar is the case for media companies. As long as you present the content in the right format, on a preferred medium, there is no question of editorial losing to algorithm-based decisionmaking.

Jehil Thakkar,
Partner & Head, Media and Entertainment, KPMG