Data matters to your editors

Introducing editorial analytics into a traditional print newsroom can be difficult, said Sebo Banerjee, Data and Analytics Lead of HT Media, an Indian mass media company.

Banerjee was one of the speakers at the sixth annual Big Data & AI for Media  conference in Hong Kong from 7-8 December.

Because editors with a complete print mindset might not be familiar with data, online media and analytics, the management should start with selected groups who would influence and inspire others, said Banerjee.

Long-tail content volume may contribute up to 60 percent of traffic volume, said Banerjee, who believes that improving median page score is generally a good index of long-tail health.

To quantify editorial efforts, he recommended newsrooms to first collect and track as much as data about editorial productions, by individuals, sub-teams and teams and then compare the data with aggregate page scores. This helps create efficient regression equations in place. Next, they should set benchmarks by genres or sub-teams and use an efficient system to trigger email alerts for fluctuations. Finally, it is crucial to connect efforts to revenues.

Banerjee also shared some hacks and tips with the audience:

1. Academically, KPIs sound great! But other than people at the very top, people do not look at them much. So, you should not put huge efforts into creating them.
2. You may have your preference for data visualization tools, but do not force it on data consumers. If they are in editorial or sales, use simple and uncluttered solutions like Data Studio or Chartio. If consumers are geekier, use Tableau.
3. Tableau Desktop, for all hysteria around it, is nothing more than an advanced data visualization tool. It does not have statistical capacity unless you integrate ‘R’. Plus, it still has a cluttered UI. It does not talk to live data sources such as Google Analytics or Facebook the way Data Studio does. It creates extracts in such cases first, which is cumbersome.
4. It is fashionable to have engineers building data pipes or connectors using service APIs. But an engineer’s one month’s salary will buy you off-the-shelf and custom solutions from platforms such as Supermetrics, Blockspring or Zapier would even worth a year of salary or more.