I think we need to have a talk about our algorithms.
Karl Allen-Muncey
Experienced Entrepreneur & Leader. Specialising in Marketing, Design Thinking, Brand, Product & Team Development, Innovator, Investor & Motorcyclist 🏍️
Are you thinking for yourself? Or Is Instagram doing it for you?
I’ve always been fascinated by social media and the science behind these platforms. I’ve been working with social media my entire career. Six Degree’s, Friendster, Myspace - I’ve had accounts with all of them but as we’ve matured - So have they.
Outside of the work I do in the innovation space with the Accelerator Centre, I have a keen interest in other things; Motorcycles, coffee..travelling - so, like a lot of other people, I have an Instagram account that I play around with - and I think we should have a chat about it
As a marketer with a decent amount of experience, I’ve profited from knowing about how the social media game works, even personally, I apply that same knowledge to grow my own accounts and opportunities.
In 2019 I set up a side project Instagram account (@zenmotorcyclemaintenance), and in 3 years it has just around 10,000 followers and generates a decent amount of engagement. I’ve leveraged the audience to work with brands and partners, and been able to do some really cool stuff - All because of demographics, being consistent and creating meaningful engagements.
For 3 years, the account has been growing in traction, increasing reach and engagement. But one day, it stopped. Instagram changed the rules again, and the tap got turned off.
So how do you go from getting likes in the thousands, and posts topping a million views one week - dropping to hundreds and post engagement failing the next, all just with the introduction of a new algorithm?
The elusive algorithm changes that has everyone talking, comes from the fact that Instagram is shifting from it’s photo sharing focus, to having a video content (Reel) focus… and that has a lot of people shrugging their shoulders - you either change your strategy to work with the new rules, or you’re old news.
Honestly, If you aren’t lip syncing to this weeks trending audio, or pumping out post after post - your content strategy is going to take a hit.
To find out why this happened - we need to find out how we got here.
The Birth of a Behemoth.
Founded by Kevin Systrom and Mike Krieger in 2010. Instagram is a Silicon Valley fairy tale. With over a billion international users in just over 10 years it has easily become one of the most influential social media platforms in the world. But it had really humble intentions…
The first prototype of Instagram was a web app called Burbn, which was inspired by Systrom's love of fine whiskeys and bourbons - so you have to respect that - but it was too close to apps like foursquare, however the photo sharing element was very popular so they changed the primary focus of the app to feature photographs, specifically those taken on mobile devices - and Instagram was born - and boy, did it grow.
Just imagine launching your photo sharing app on October 6th 2010 and onboarding a million registered users in two months, 10 million in a year, and 1 billion as of June 2018 - That’s a lot of people. And a hell of a lot of selfies, cat pictures and well photographed plates of food.
Everything was going fine - and then Facebook came along..That sounds a little harsh doesn’t it, and it is.
In April 2012, Facebook Inc. acquired Instagram (at the surprise of the Instagram team) for approximately US$1 billion in cash and stock - and that’s when things got really interesting.
If you’re up on your ‘history of the internet’ or seen The Social Network by Columbia Pictures, Facebook (or Meta as we know it now) is an American online social media and social networking service. Founded in 2004 by Mark Zuckerberg with fellow Harvard College students and roommates. Facebook became the blueprint for how to build an intelligent social media network - and all of this data came from the content being generated by its primarily youthful audience.
Facebook has been around for a good while. In ‘Internet time’ anyway, and the young audience is all grown up - but that did a couple of things;
Facebook leveraged its vast, demographically rich audience - to become one of the most intelligent advertising platforms of all time.
In 2020, about 97.9 percent of Facebook's global revenue was generated from advertising, whereas only around two percent was generated by payments and other fees revenue. Facebook ad revenue stood at close to 86 billion U.S. dollars in 2020. So Ad’s make up a significant portion of that income.
In order for Facebook to stay at the top, and remain relevant with the youth audience (an audience that has steadily been declining in recent years, it needed something new. People just weren’t spending enough time on site - it had actively been trying to create hooks to retain or pull in younger users. but it wasn’t working - and that is why it acquired Instagram.
Squeezing the juice out of the Instagram fruit.
“Increase engagement, increase time spent”
The ‘Algorithm’ get’s referred to with the same mysterious awe as the Great and powerful Wizard of Oz behind his curtain.
There’s always this talk about how we are targeted as users, and what (or who) determines how and what content gets served up to the users by ‘the algorithm’ - but it’s a lot more complicated than that.
Algorithms use hundreds, if not thousands of machine learning models, and they all interact with each other. The way that they interact with each other is determined by a guiding metric or goal of the ranking model - and in Instagrams case, the guiding metric is ‘Meaningful Social Interactions”.
The term ‘meaningful’ can be misleading - because it doesn’t mean meaningful in the emotion sense, but more so in the engagement sense. What Interests, people, content types or categories lead you to engage with a piece to increase your time on site.
These machine learning models work together to effectively manipulate your behaviors and ultimately get you to perform tasks or actions - and that manipulation can have some nasty consequences over time.
Recommended by LinkedIn
“Toxic for teenagers, and facebook knows it.”
In 2021, Former Facebook data engineer and scientist, product manager turned Instagram whistleblower Francis Haugen posed the question “Are we going to let algorithms rule us, or will people rule algorithms?”. She disclosed tens of thousands of Facebook's internal documents to the Securities and Exchange Commission and The Wall Street Journal. Stating that “The companies leadership knows how to make facebook and instagram safer, but they have put their astronomical profits before people”
The greatest trick that Meta ever pulled was convincing the world the algorithm didn’t exist
The whole point of targeting - is for it to be relevant and appear natural..
Ken Ammi followed up this wonderfully butchered quote with “The second greatest trick the Devil ever pulled was convincing the world he is the good guy” - Now that for me is the epitome of how engagement algorithms succeed, they convince you that they are benefiting you, when in truth, they are narrowing your bandwidth of diverse content and ads.
User behavior, predictions and consequences dictate what we are served and in what order - to have the best likelihood of engagement - a comment, a double tap, a save or forward. It uses that knowledge to base its decision on as a platform.
Remember when I mentioned that Instagram is shifting from it’s photo sharing focus, to having a video content (Reel) focus…Its for that precise reason - People consume more video content than scroll through pictures, the engagement is simpler and faster, and you can display more ads and sponsored content to a user in any given time slot.
“If an algorithm is making decision for you, then it has the ability to take your control away” - Bailey Richardson, Former Instagram employee.
The human brain doesn’t distinguish between what is authentic, and what is fake, and this is where the algorithm lives. When it sees something it likes, or it learns from your behaviours - That’s where it targets - Leveraging opportunities to appear natural, to create an event and predict a possible outcome, all for the purpose of converting advertisements into click thoughts for their media buyers.
"It's your kids, Marty. Something’s gotta be done about your kids!" - Dr. Emmett Brown.
We actually have very little understanding of the long term implications that this targeted refinement might have, and no way of winding that machine learning clock back from a place of hindsight - for example, what decisions were entered into the system, and what paths did we choose that resulted in the content and ads that we currently see.
What we do know is that this refinement and relevancy algorithm and machine learning models has the capacity to provoke a reaction or to change a behavior and because of which is having a profound effect on mental health. It’s affecting our youth in a whole host of ways; body dysmorphia, eating disorders and anxiety to name a few.
Instagram is quite different from other platforms, tiktok is about performance, snapchat is about faces, augmented reality, facebooks internal research says that instagram is about bodies and social comparison - peering through a window into other people's lives and making constant comparison.
And when it comes to our kids - we need to become far more aware about the impact social media has on their development, and our society.
According to love organization, a group that supports youth to overcome the challenges they face through healthy relationship and emotional intelligence programs, in Canada in 2015 (the same year Instagram opened its platform to ad’s, and we lost the divide between user generated content and advertisers.)
That means around a quarter of all of our 6th graders are susceptible to the intentions and action of the algorithms, and begin the molding of their demographic profile. This could become very dangerous.
We don’t fully understand what the consequences of this period in society will be - this is just the beginning. The use of instagram, and let’s say social media in general has risen to have a ubiquitous impact and poignant place in our society - we hear about how it’s affecting our mental health, destroying the connections between people - and even changing the narrative of our communities and attributed to affecting political appointments and elections - It’s pretty important.
(for additional and awesome reading check outI: The Shallows: What The Internet Is Doing To Our Brains by Nicholas Carr, Ten Arguments for Deleting Your Social Media Accounts Right by Jaron Lanier & The End Of Absence: Reclaiming What We've Lost in a World of Constant Connection by Michael Harris)
So are we just screwed then?
I don’t think we are screwed, but we are certainly on a path that we need to be acutely aware of, and a lot of work needs to be done to educate not just our kids, but our leaders and decision makers on the effects these algorithms are having on our culture and longer term impact on our society.
We live in a weird world. Even when brought to Congress to testify against allegations of Facebook targeting and exploiting its users, the senate actively looked like they didn’t understand either the questions they were asking, or answers they returned. Ex US President Donald Trump would actively troll international leaders on social media and increase political tensions, allegations of Russians getting involved with election tampering via targeted ads, a mental health pandemic brought on by fear based reporting, and a potential bill on internet censorship by the Canadian Government that could infringe things like free speech…
We’re not looking good folks.
We need more education, and active involvement in what we want the digital future to look like for us, our kids, and our society.
Be the change you want to see (literally).
We know that the path of content we see changes but becomes more regular and refined. You just have to go to your ‘explore page’ on Instagram to see who Instagram thinks you are..
The system doesn’t take a break, you are constantly being graded, targeted, tested and quantified. Every move you make feeds the machine learning model's vision of you as a user - and your content will reflect that - because the system WANTS you to spend more time on site.
Former Facebook Responsible AI Researcher Manish Raghavan says “If you want to understand why instagram is the way it is, you need to understand how the algorithms have shaped it to be that way - they are the result of people's choices.”
That means it knows when you react or engage on a post that has dogs in it - or it knows that you get into lengthy debates in comment sections about vaccines..so guess what - that’s exactly what you’ll see more of, more regularly, and at the expense of other more rounded content.
Perhaps It’s time for us to stop for a few minutes and make some decisions on the ethical dilemma that’s unfolding in front of us and take responsibility for our actions to look at this thing we’ve created and ask ourselves if we’re on the right path.
Because if you look at the stats.. We’re not.