Hello there,

We’re probably not the only ones being bombarded with 'don't fall behind, use these mind-blowing AI tools!' type messages right now.

It’s tricky. While you may feel the need to engage with developments in the field of Artificial Intelligence, you’d be entirely forgiven for experiencing a case of AI-fatigue thanks to all the substance-lite chat on the subject in the year since ChatGPT went mainstream. Waiting to see if the big hype subsides is an understandable and very relatable response. 

What we’re waiting for, really are tangible, concrete use cases, which cut through the noise - and this was something that publishers of news sites were talking about at length this week at the INMA Media Innovation Week in Antwerp. We were there and were pleased to be one of those few who could already demonstrate one-on-one what smartocto is currently doing to bring editorial analytics into the AI era.

So, after an extensive experimental and developmental period, it's time to share our findings and showcase product improvements to a wider audience. In our latest webinar, we provided a glimpse of that, but now smartoctober is around the corner, and our full focus is on intelligently leveraging Artificial Intelligence and making it work effectively in the newsroom environment. So, this month stay tuned to see us unveil new, smart applications in our tool, such as headline testing, article readability measures, automatic user needs recognition, and timing optimisation. 

We invite you to closely follow these developments. Follow us on LinkedIn and be ready to receive an extra newsletter in two weeks. 

Webinar about headline testing and AI

Editors are generally adept at crafting alternative headlines for A/B tests, but if the editorial team lacks the time or energy to brainstorm alternatives, with just a few clicks, ChatGPT can be a valuable collaborator. That is one of the key conclusions we highlighted during the webinar 'Why headline testing matters, how AI could help.'

Whether editors feel that modifying prompts is more work than generating alternatives through AI-free brainstorming is perhaps a personal preference, but there are A/B testing tools that can automate this task. 

Client case: headline testing with ChatGPT

If you’re short on time, but rich in curiosity, we also wrote a client case about the pilot we did with Omroep Brabant on this subject. And just to make clear: the goal there wasn't to determine whether ChatGPT can outperform journalists in headline creation, but rather to see if ChatGPT can be beneficial in helping to make suggestions. One good example of doing so:

"We were quite critical of the suggestions that ChatGPT came up with, but often we still saw a new perspective that we hadn't thought of ourselves," says Omroep Brabant editor Janneke Bosch in this client case.

Blog: what is the optimal number of words in a headline?

As you can see, we’ve done a lot of work on headline testing recently and between client-specific workshops, consultations, and data investigations we’ve acquired a good sense of what makes a successful headline. Of course, it’s not possible to be prescriptive - a lot depends on the media brand, the topic, the timing, and so on. But in our latest blog, we’re showing data that could send you in the right direction. For instance: did you know that longer headlines aren’t as troublesome they’re sometimes made out to be?

Event: INMA and b future festival

The INMA Media Innovation Week is not the only place where our team indulges their thirst for new knowledge and information about the publishing industry. We also attended the first b future festival at the Bonn Institute. There was a strong emphasis on the importance of constructive journalism—a significant development in the response to news fatigue. 

Nic Newman himself came to explain once more that an increasing number of people are avoiding the news. There is a growing sense that news is mostly negative and hopeless. Most of those present agreed that constructive journalism can turn the tide of reader apathy. Check out this summary of the interesting readings from that conference.

Reading tips

  • How does AI actually work? The Financial Times tries to explain it in a smooth and interesting way.
  • Just to emphasise once again that headline testing is only the beginning with AI: The New York Times had chatbots write essays for Harvard, Yale, and Princeton. The apparent proficiency of the bots continues to be astonishing.
  • Another great reminder that newsletters could be very popular… The local news outlet Manchester Mill is preparing to expand across the UK after being valued at almost 2 million euros by a group of investors including the former New York Times boss, Mark Thompson. And by the way, the same Thompson is the new CEO of CNN
  • To immediately mention CNN and share some news about... news! Publishers see a dramatic drop in Facebook referral traffic as the social platform signals an exit from the news business. 'The recent changes reduce the already lackluster levels of referral traffic even more,' writes Oliver Darcy.
  • The developments from Meta are not entirely gloomy. WhatsApp has launched 'WhatsApp channels’. This allows publishers to once again use a channel for direct contact with their readers.

So, that’s all from us for now. Remember: we’ll be sending you another newsletter in a couple of weeks, mid way through smartoctober.

Stay tuned!

Team smartocto