Press Gazette has partnered with BuiltWith to create a unique picture of the technology which underpins more than 2,000 of the world’s leading news websites. Last week we looked at content management systems. In the second of our series we take a look at the most popular newsroom analytics tools.
Although BuiltWith’s data cannot give us an exhaustive picture of the tech stack used by a newsroom, it can inspect a page’s source code to pick up known tech tools in use.
Our analysis revealed that the majority of newsrooms use industry-standard tools to deliver essential information and data, although there’s room for differentiation in the detail. In the news industry where cross-site comparisons are a critical currency for publishers and their advertisers, tracking codes from third-party audience measurers were also commonly found, with codes from Comscore, a company which provides marketing data and analytics to media and advertisers, appearing for example on 70% of sites.
Understanding who is on your site, when and which kinds of content work best is key to building loyal audiences for publishers. Knowing how traffic and engagement vary across channels such as social media, search and aggregators helps determine where to target efforts. A 2018 global study from the International Center for Journalists found that almost 70% of newsrooms in 130 countries consulted analytics at least daily – with these more likely to be checked even more in mature news markets.
“Analytics are usually more about getting the basics right so that you’re tracking everything the right way,” says Gabe Karp, EMEA director at digital agency, 10up. “It’s making sure you’ve got the right tool and getting it implemented well.”
Google Analytics, rather predictably, dominates the analytics space appearing on 97% of news sites.
But while it’s excellent at helping understand trends over time, many publishers use analytics tools developed specifically for newsrooms. Chartbeat, which counts the BBC and Reach as clients was found on 31% of sites. Parse.ly, which we have recently adopted here at New Statesman Media Group, meanwhile was in use on 26% of sites.
While these publisher staples for analytics were found across all kinds of sites this was even more so the case with the top names in our list. Using traffic data from Similarweb and Comscore as a rough proxy for top sites, of the top 500 URLs on this measure, 347 (69%) were either using Chartbeat, Parse.ly or in some cases, both.
Analytics tools specific to the news industry or other content-driven businesses of course provide standard insights on traffic sources and top pages, but where they come into their own is on things such as how they measure engaged time or in providing publishers with a one-screen view on how a piece of content is performing across various channels. Chartbeat for instance returns different engagement numbers to Google Analytics because it "pings" pages every few seconds to measure active visits at any given time.
A/B testing to optimise content
Although often bundled up in more generalised analytics software, a significant number of sites were also found to be using dedicated A/B testing tools. A/B testing allows analysts to test out real-time alternatives to see which option works best with users. Common uses in a newsroom might be showing groups of website visitors different headlines for the same article to see which performs best - something which the New York Times has in the past said it has done using in-house tools. Reach Plc also has a dedicated content A/B testing team.
Google’s premium A/B testing tool Optimize 360 appeared on 20% of sites, while competitors Optimizely and AB Tasty appeared on 144 (6%) and 124 (5%) of sites.
Tracking - everything
Analytics tools form part of a wider suite of tracking tools commonly used by publishers. An investigation last year by Press Gazette found that on average, news sites had 17 third-party reader tracking tools.
Tracking codes which help website owners increase their revenue by targeting advertising, are also common on news sites.
Facebook’s “Pixel”, which lets companies see how many of their Facebook and Instagram ads result in conversions, allows them to target people who’ve visited the website before and opens up Facebook analytics was found on 55% of sites in our list.
Google’s Conversion Tracking which shows website owners what happens after a customer interacts with ads on its sites -- was found on 268 (11%) of sites.
While understanding audiences to maximise advertising will remain important, reader revenues are of course only becoming more important for news businesses, exceeding advertising revenues for the first time in 2013 according to WAN-IFRA.
While analytics have come a long way in re-pivoting newsrooms to focus on getting to know their readers and their behaviour, combining knowledge on how users interact with a news site with upcoming technologies such as AI could enhance reader revenue strategies.
Canadian publisher the Globe and Mail for example have built their own AI-based automation and prediction engine, Sophi that analyses data on article performance to programme the homepage and key landing pages. The system also uses AI to work out which stories users will pay to access and which can deliver more advertising revenue.
Machine learning drawing on detailed data of how readers use a news site, says Karp, will also play a bigger role in helping publishers tailor their subscriptions.
“Even with some of the more effective paywalls or data walls, most of the time they're ‘a one size fits all’ or ‘one size fits a segment’ thing. I think what we'll see is much more automated, AI-driven single user journeys,” says Karp. “So if the machine can figure out that I only read one article a day from The New York Times but their subscription is valued based on me reading five articles a day, can they give me a different offer at a different price point?
“It's those types of things that I think we still haven't seen yet.”