Whose recommended content is it anyway?

A successful recommended content strategy can be priceless for publishers, yet getting it wrong can be a real turn off.

by Rob Corbidge

Published: 15:09, 24 August 2023

Rob Corbidge is Head of Content Intelligence at Glide Publishing Platform, applying the latest knowledge about advances and ideas in the publishing industry to our own product and helping clients get the most from their content.

Around eight months ago I noticed some content dissatisfaction had crept into my media consumption world. This information is of little consequence unless you are me of course, yet the cause of the dissatisfaction was a misstep by YouTube - arguably the masters of content recommendation - which only goes to show that anyone can get it wrong.

What is interesting about the cause of this dissatisfaction was that recommended content was at the root of it. 

Getting recommended content right is hugely important for many publishers. The chimera of "personalisation" was all abuzz in the industry for some time, and remains an aim for some. A good recommendation can set a reader, viewer or customer off on a mutually beneficial additional journey within the bounds of your content. They'll be satisfied and so will you. 

Just to clarify, here 'personalisation' has become the general term for, well, content that seems like what you will want to see based on what you watched before - NOT 'true' personalisastion of knowing who you actually are.

Anyway, it seems that in the push of of a new YouTube video format, Shorts - a TikTok-style video snack versus its more common long videos - YouTube started to recommend trash to me. 

There's a whole content hive out there producing video material of highly questionable quality, and suddenly YouTube thought I would be interested in it. Equally, it started to recommend content outside of my interest areas, videos with very few views.  

What part of recommending low viewer count content outside the viewer's areas of interest sounds like a good idea? Foremost, it's a really bad idea for the content creators, who will rely on Click Through Rates to rise up the recommendation algorithm.

The only reason to think Shorts were responsible is that they appeared on the homepage at the time things started to go askew for users. However it's a reasonable assumption given the way Big Tech responds to threats - in this case TikTok - by simply attempting to mimic the threat and going all out on it. As a premium subscriber who views most YouTube content on a widescreen TV, maybe I'm not the representative YouTube content consumer.

There sits an issue for almost all platforms. As they've increasingly tried to be all things to all people, and the novelty of the sheer amount of content available through them starts to become stale, is a stratification such as we saw naturally in old media likely to occur? 

To use a menu analogy, the current situation is one where fine wines and single malts are presented on the menu along with Mad Dog 20/20 and supermarket cider as being of equal value, when objectively they are not. 

I'm a content consumer who likes properly researched history, technically revealing engineering explanations and competitive mountain biking vlogs. I do not wish AI-narrated nonsense about Russian super weapons to even cross my gaze.

It is possible of course to set notifications for particular channels, or look at the Latest section to see what's new from the channels you know and like. Yet that destroys a vital aspect of digital media - discoverability. Correctly recommended content can lead to a good place.

The other aspect of this is that YouTube can afford to make a mistake. It wouldn't be a proper piece about a Big Tech platform without the statutory rant about distribution control and of course in this case YouTube is an advertising business that controls the content distribution channel, with no serious rival - it has the room to make mistakes that few actual publishers enjoy.

Who knows, maybe the revenue from Shorts more than makes up for lowering the IQ of people's feed by some magnitude.

Now, here a pause is required. When thinking about publishing, we have to take our own content consumption behaviour into account. 

But equally, everyone's experience is different, so the danger comes from some talking head - in this case, me - assuming that what they are seeing is indicative of something broader or a widely-shared reality. Checks are required before any assumption can be made.

Yet it's a fact that the issue is becoming noticed more broadly: here is one such video of a number which are gaining traction lately. Video content not by us etc.

What do you think?