Follow by Email

Search This Blog

Oct 18, 2017

Reuters Institute about Journalism

Two interesting papers (study/research) about digital journalism and the publishing business from Reuters.

Their 2017 predictions, now that we are coming close to 2018, are worth a read and most of them are accurate - they predicted the most important topics correctly, although, as always, trends need more time than expected to really become mainstream. But if you're working in or you're interested in digital pubishing, this one is worth a read:

Full PDF: 

A great resource for statistics, numbers and data is the 2017 digital news report with country by country numbers from Europe, Americas and Asia/Pacific

Younger generations as sports fans

We get different studies about how Generation X, Millenials and Generation Z consume sports. But the general rule seems to be: the younger, the smaller the attention spans, the less trust in governing bodies, the less live consumption on TV (legal and illegal streaming), the more social media and highlights consumption.

McKinsey says that GenerationX and Millenials are very similar in their sports attitudes and even in media behaviour, although they also diagnose a) significantly more streaming than linear TV on the Millenial side and b) significantly more social media activity around sports for Millenials.

At the same time, PWC sees total disruption imminent because of younger generations and tech development in their 2017 Sports Survey

Full PDF:

German article:

With regards to business, besides the "media packaging" that needs to be done to attract younger fans, two are also worth mentioning:

65% of 18-34yo stream Premier League matches illegally at least once a month (see  link below). Even McKinsey points out that "converting the pirates" is one major activity to focus on.

And trust in governing bodies is decreasing (PWC study) heavily and should be of major concern (for both the governing bodies maybe intending to stream / sell matches themselves one day, but also for media acquiring rights).

Edit: One more research from Deloitte, Full PDF -

Oct 16, 2017

The state of data journalism

Interesting analysis about data journalists in newsrooms, how many there are and what they do. Next step: Data analysts in newsrooms who work closely with editors not on content as such, but on which content to choose and how to package it for maximum success on various platforms. Plus they need to work with IT to ensure journalists get a tech infrastructure that enables them to perform best, and keeps generating meaningful data the analysts can work with. In my opinion much more needed than "traditional" data journalists.

and here's the full PDF:

Voice is the next big thing

I believe there will be several "next big things", but if we look at the typical exponential curve, voice is in my opinion very likely to currently be in the "valley of disappointment" - but soon reach that infliction point where it "explodes". All the ingredients are  there:

Alibaba invests 15 bn USD in R&D

We talk a lot about exponential and how it makes it impossible for (traditional, linear) competition to keep up. Here's another example: A company like Alibaba can invest 15 billion in researching AI, quantum computing, or anything potentially "foundational and disruptive". I'll make sure to tell this to a local ecommerce company...

Spotify's Deep Learning

"Discover more", the weekly playlist with 30 titles that Spotify suggests for each user individually, seems like some kind of sorcery after using the platform for a few weeks: recommendations are so accurate, Spotify seems like that good old friend who helped build your record collection a few decades back by helping you discover great tracks and artists. The magic behind it is deep learning and narrow AI, and although music is an "easy" field - for example a music track can hardly be "outdated" or "out of stock, compared to journalism or e-commerce - the logic behind it is relevant to almost everyone doing digital business: learn about your customer from usage (not questionnaires) as much as you can, and the be able to apply these learnings in an individually relevant selection from your inventory. Increasingly, this will not be an "if this then that" matching, but an AI / machine learning task. Interesting read: