Unintended consequences: how online change is hard to predict
And even harder to respond to.
One of the many regular themes of this newsletter is about the level and frequency of change that is our new normal, and how we can all find ways to plot a path and find opportunities within this environment.
There have been a few examples of late which demonstrate how multifaceted change is in our environment, and how whole new mindsets need to be adopted to be able to operate effectively in this space.
Not only is our era of change disruptive, it is also not linear, nor is it easily predictable. In addition, the human response reaction to this change can be also just as unpredictable too, thus creating multiple moving parts to plan around and respond to. And another key behaviour is that while there is always something new, this doesn’t mean that the older activities are no long valid or of use, nor does it mean that your audience has moved to the new thing in the same time frame as early adopters.
To give a generalised example of what this means. For the last 15+ years, companies and brands have been establishing all sorts of online presences as new platforms and trends emerged. So, first it was Myspace and Bebo, then along comes Facebook and Twitter and YouTube, then Instagram and Pinterest, Second Life, and later comes TikTok, BeReal and BlueSky. While some of these have disappeared (RIP Bebo), for many of the others, these were added incrementally to a brand or company’s portfolio. At the same time, often companies applied some sort of subconscious ‘one in, one out’ rule, where the new platform was exciting and fresh, and took up a whole host of time which had the consequence that there wasn’t the capacity or interest in whatever was perceived as the older more boring and stale platform. Bluntly, it wasn’t uncommon for TikTok to push Facebook and the company’s email newsletter down the priority list.
The problem with this approach is it means that the personal online behaviours of the team managing this activity can end up determining an organisation’s strategy. Fine if this aligns with your target audience’s behaviour or opinions, but not so fine if it doesn’t. It certainly can be common where brands invest heavily in platforms like Instagram or TikTok as they were fresh and new to their teams, plus of course these platforms are important routes to reaching particular audiences: TikTok for younger audience looking to be entertained by highly shareable video content or Instagram for slightly older women. Again, if these are the audiences a brand wants, then this is a fine strategy. However, if the brand sought a larger, older and broader audience, then that audience is less to be found on TikTok, and instead is still on Facebook and hasn’t gone away.
It is important to understand that we aren’t in a stable world, and there are three semi-independent moving parts that need monitoring to be able to plan, adapt and course correct in this environment.
Part 1 - the platforms
While we talk about the direct to consumer world, in reality we are still mediated through companies as we are dependent on social and video platforms to reach audiences. The difference for TV production companies is that this relationship with audiences is much more direct when operating on the various tech platforms like YouTube in comparison with the B2B world where the management of audiences is left to the studios, networks, distributors and broadcasters.
These platforms we are using to reach our audiences are constantly evolving. This is thanks to the volume of metrics and performance data within each company, and the armies of product managers, analysts, engineers and developers tasked with enhancing and improving whatever bit of the platform they are responsible for - both to improve audience experiences as well as obviously to make more money.
To give an example, yesterday YouTube announced a whole host of new features at their Made on YouTube event.
The next 20: Powering the future of entertainment together at Made on YouTube
YouTube official blog: Announcing our biggest updated to YouTube live
As part of the announcement, they said that 30% of daily logged in YouTube viewers watching live content in Q2 2025, and therefore they are making their largest ever upgrade to the live streaming experience as a result. In other words, as a platform, YouTube noticed a user behaviour and therefore they prioritised enabling this trend to give a better experience to their users as well as encouraging more ways to make money. As an aside, they’ve framed this live announcement as reimagining TV and formats, while also emphasising further tools for video podcasts:
We’re bringing creators and their fans closer by reimagining TV and formats driving today’s culture.
This is happening constantly within every platform and tech company the world over, and is why there is no end point to the change era we are in.
Part 2 - audience behaviours
Meanwhile, audiences themselves are constantly changing and evolving their behaviours, and often that doesn’t happen in a monolithic shift, but different demographics and territories can move not only at different speeds, but also often in divergent ways.
You’ll be able to see this in your own life; think about say your online social activities from five years ago, how are they different to today? Or when did you last buy a physical newspaper? Did you start using an eReader for books and then stop and instead go back to print? What about the cinema - do you still go as much as you used to or has that changed too?
Then think about all of the different demographics and territories there are in the world and multiply these changing behaviours by all these groups, plus of course the zillions of hobbies, interests and subcultures out there.
The point here is what was normal a year ago might have changed again. And that new trends might emerge in specific demographics which then move like osmosis to other groups, or indeed might stay localised to that group and die out like a passing fad.
Facebook is a great example. Facebook’s audience was always significantly larger than any other social platform by some margin, and while usage might have declined with a younger demographic, older audiences have stuck with it for far longer. However, rather than it being primarily about connection with your social graph (family, friends and colleagues), it has quietly become the home of global community groups and pages of people with shared interests - so cooking or fitness, parenting to reading - and a million more besides.
Another great example is Blue Sky, which emerged as a platform in large part as a response to the changing ownership and policies of Twitter/X. Here we’ve seen the network effect of a whole bunch of people and brands shifting from X to Blue Sky, however that hasn’t resulted in the same level of cultural salience for it as a platform in the same way that Twitter had back in the 2010s. Rather it appears that this has fragmented the audience of what was previously a single public square, with a further consequence of reducing the wider cultural impact of trending on a Blue Sky and/or X as a result.
Part 3 - broader internet and cultural trends
This is the nebulous soup that we are all swimming in, and covers a whole broad range of trends to be aware of, while also constantly being attuned to what might be a signal and what might just be noise. Change is not linear, isn’t uniform and doesn’t happen in a bubble; reactions and counter swings happen against new trends, and these are often unpredictable in nature as pendulums swing. When you add in particular tech inflection points involving new product launches or new technology arriving, then the cat and mouse game of chasing upsides and mitigating the downsides becomes the norm.
To give some examples of what I’m referring to here.
Firstly, the example of AI slop. This week YouTube announced that Veo 3 is going to be integrated into YouTube Shorts, meaning anyone and everyone can use this tool to create content for the platform. Here is an example below made with Veo 3, this one mocking the trend of YouTube challenges by creators:
For some, this will be an explosion of creativity, for others however, it signals yet another step into a world where we will be drowning in low quality audio visual AI content. There is a believe that this will create split audiences, triggering a backlash where some users seek safe havens of curation, accuracy and authenticity as a response.
The rise of AI content on YouTube is already quite apparent, and while there might be a novelty about this trend at the moment, it is clear that we are in the foothills of this technology and there will be both growing sophistication as well as a high likelihood of a rejection of it from certain demographics.
This piece from The Current is worth a read on this issue:
It highlights this channel below, which has generated 65m views and has 360k followers (as well as a whole host of paid subscribers), where it makes a broad range of AI videos often in the style of famous adverts. This example below is similar to the Dos Equis beer’s famous ‘Most Interesting Man in the World’ campaign. The Current say they approached the Dos Equis company about these spoof adverts, and didn’t get a response. This type of activity can be viewed by brands as tongue-in-cheek fan content which could be seen as enhancing, or alternatively they (or indeed other brands) might see it as damaging to their reputation.
This example is a great encapsulation of the challenges presented by AI content.
Firstly, consumers are increasingly wary of AI slop, with a recent US Raptive survey saying that people were less likely to purchase products on webpages that they believed to be AI generated as they considered the content ‘less premium’. And interestingly, this trend isn’t limited to older audiences, indeed, the same study suggested that Gen Z respondents were even more untrusting of ads that appeared on websites with AI generated content.
Secondly, it is likely brands will be increasingly wary of having their ads appear around AI generated content on YouTube, if it is perceived as junk and not premium. Here you can see two sentiments pulling in different directions: some believe AI tools might increase the appearance of quality by offering up a slicker visual veneer (and therefore, this will help YouTube sell more ads at a higher CPM because the lower quality content has improved). Conversely, if this visual veneer acts as a flag to certain audiences that this content is AI generated (and in turn, audiences perceive AI to mean less premium) then brands might not value advertising around this content.
Thirdly, the rise of AI slop also is starting to chomp into many creator businesses both directly and indirectly. Below is a video from creator Cory Williams who has the Just for Kids Silly Crocodile who was also flagged in The Current article above.
Here he explains how other users are using AI to rip off his work and regurgitate it over and over - he so far has found over 200 knock offs of his Silly Crocodile IP and he expects more to appear.
Indeed, many of my social feeds are full of people talking about how to find what is successful out there and copy it using AI, usually under the guise of ‘creating faceless channels’ however they aren’t completely explicit the level of replication this involves.
Fourthly, this also has a really dangerous factor around online safety and disinformation. These low slash no barriers to entry means that anyone can make anything and use the platforms to try to reach audiences.
Here is an example, where a YouTube channel called ‘True Crime Case Files’ (it is no longer available on YouTube), used AI to generate videos of crime stories that he had made up.
