Interesting article focussing on the yearly Average Revenue Per User (ARPU) as a key metric. It is important to see that it's about revenue (Amazon makes 752 USD on average revenue per every user and year, while Apple makes 194 USD with not everyone buying a new phone in retail - so this doesn't tell us much about profits), but the article provides an interesting viewpoint. Publishers like Buzzfeed make 3 USD per user from advertising, while Amazon's advertising is at 15 USD per user, Facebook (globally) at 25 USD, and Apple's service business (music/content, apps) is at 30 USD per device. More here:
Feb 22, 2019
Great feature in wired that is based on the observation that in search, you get a list of results - at least the first page, while on voice, there is basically just one answer. And this is more of a difference than it seems, as this will steer audiences and potential buyers in a way more concetrated way than we know this today (and a little more concetration is what the internet needs right now, huh?). Good analysis:
We have had the news that digital became bigger than print, bigger than TV, and now it will become bigger than all of them combined. eMarketer predicts that by 2023, digital will be two thirds of the overall ad market. There is incremental growth, but we can also see heavy losses for TV (>2% - doesnt sound like much, but will be the lead up to a "waterfall moment") and dramatic losses for print and directories (-17% and -19%).
Publishers always had subscription businesses - but in print. And these were and are different from digital by nature. It is amazing to see how most news organizations learn digital subscriber acquisition and retention from scratch. Now that most have built a basic subscription business (BILD talks of >400k subscribers), retention becomes a key issue.
I have a problem with the term "robot", but this article provides a great overview of automation and artificial intelligen (-like) applications in the journalistic value chain. AP for example increased output about earnings reportings from 300 per quarter to 3700 by using automated text generation (maybe with AI elements). There are more examples in this article, like image recognition helping to tag digital photo archives or software to identify deep fakes.
The first question to ask (basically any digital business, but especially publishers) is where the traffic comes from - and with Facebook delivering less and less, Instagram not showing any efforts to integrate links outside stories, direct traffic and Google search remaining more or less stable, aggregators are there to become the fourth big pillar in traffic sources. Great overview (data based) at Niemann Lab about referral traffic, naming "Smart News" as the up and coming option (along Flipboard, Google News etc.)
With a target of 10mn subscriptions by 2025 (doesn't sound like so much in a Netflix, Spotify world - but is actually ambitious), the NYT has reached 3.3 million paying subscribers now (not only news, also crossword and food apps) with a YoY growth of 27%. But the real message is this: The New York Times has actually ADDED 120 employees to their journalists, totalling more than 1600 now. This will not be a model for every digital publisher as the NYT is kind of international, and the US alone will give room to what - 4,5 nationwide digital subscription news businesses? Still, it is amazing to see how they do it, and every news publisher transitioning into digital can at least learn from monitoring the NYT and draw their own conclusions.