Journalism is in a tough spot. For decades, money and jobs in the industry have been in decline. The government’s Cairncross review, released last year, revealed the scope of the problem.
It found that total press industry revenues had declined by more than half over the previous 10 years, and that the number of full-time journalists had fallen by over 25 per cent since 2007. This problem isn’t unique to the UK – US newsroom employment has also dropped by a quarter since 2008.
The review also found that a quarter of all regional and local newspapers have closed in the past decade. This should ring alarm bells – these news organisations do an essential job of holding local government to account in a way that is difficult for national newspapers, and can also uncover important stories.
What’s to blame for this decline? Obviously, a major culprit has been the growth of the internet, with audiences switching from buying print newspapers to accessing news online for free.
But if technology caused this problem, perhaps it can also help fix it. One company, Dataminr, is doing just that. It uses artificial intelligence (AI) to help organisations find breaking news stories faster and more easily.
Trawling the internet
Dataminr works by trawling through social media, blogs, and other public sources of data, such as from aeroplanes and maritime shipping. Its algorithm, which has been developed and honed over the last 10 years, looks for patterns and events that are relevant to journalists based on their preferences, and sends real-time alerts and live updates to them.
Of course, this data is already available, so you’d think that journalists could simply find it themselves, especially on platforms like Twitter. But as Jonathan Barrett, managing director EMEA of Dataminr, points out, around 500m tweets a day are posted on social media.
“How do you find the right valuable information for news entities in that huge sea of data?” he asks. “Unless you know exactly where to go to, the probability of you finding it is very small. Some of our news company partners refer to it as not just finding a needle in a haystack, it’s finding a needle in multiple haystacks.”
In practice, Dataminr could spot several tweets from different accounts that are from similar geographical locations and all mention an event – perhaps through a keyword such as “fire” or “protest” – and alert a local journalist, who could then decide whether it’s a story worth investigating.
Barrett recalls one client case study about a journalist with the publishing house Reach who reported on a security scare in Cardiff in December 2018. Police had evacuated part of a shopping street after threats were made to a hotel. The reporter received alerts about the incident from Dataminr, contacted the police to check its veracity, and was able to break the story.
“It was Reach’s biggest story that month, and one of its top five stories for the year. That’s a pretty strong endorsement of how Dataminr can support local news,” Barrett says.
Winning over journalists
Journalists may worry that this algorithm is out to replace them, but Dataminr does seem to winning them over. It is now being used in 600 newsrooms around the world, and journalists appear to find it helpful. Jonah Bromwich, a reporter for The New York Times, wrote an article last year about his experiences with the service. He described it as “handy”, adding that it is his most important tool for breaking news.
I also asked Reach about its experiences with the platform.
“When an incident is ongoing, we are able to set searches and geographical boundaries in Dataminr to bring us all the tweets and other social conversation from that place or about that event,” says Karyn Fleeting, head of audience engagement at Reach.
“It acts as our virtual eyes and ears on the ground, and helps ensure that we don’t miss a thing.”
By helping reporters, Dataminr is also providing a way for news organisations to build their readership with more breaking news stories. Barrett points out that by attracting the right audience, sites can generate more revenue from advertising.
“Yes, big stories like Brexit are important to someone in Scotland or Cornwall, but equally a story breaking in their community that is relevant to them and with a local angle is compelling,” he says. “If there’s a breaking news story, and you’re the first to break it, people will then keep coming back to your publication.”
Journalism is at a crisis point. The combination of squeezed budgets and declining ad revenues are causing many – if not most – publications to struggle. Dataminr’s value proposition is that it can help these beleaguered institutions use their limited resources efficiently, and report stories that are potentially more relevant to their core readers – that should sound appealing, especially to local and regional press.
“We’re enabling news organisations to have many more feet on the street,” says Barrett. “You don’t have to have a journalist on every street corner, instead you can harness the power of technology.”
Main image credit: Dataminr