Monday, 28 March 2016

Algorithm Blues

I'm sure not many people knew what an algorithm was before Easter weekend, but thanks to everyone losing their shit on Instagram, the word is now pretty mainstream.

In case you have no idea what I'm talking about, last week there were whispers in the wind that Instagram were changing the format of news feeds. So instead of seeing all posts of everyone you follow in the order they posted them, we will only see posts from people we view most/interact with most, or brands who have paid for the pleasure.
Celebrities and influencers had a meltdown, frantically telling people to turn on notifications so they wouldn't miss out on their posts.


In some ways, this is good. No longer will you see 12 consecutive photos of Susan's "night out with the gals" which prove that she cannot: a) work a camera, and b) grasp the concept of art direction.. (It's quality Susan, not quantity.)

But in many ways, this is bad. I like waking up in the morning and scrolling down all the photos that were posted whilst I was foolishly asleep, so I know I didn't miss out on anything.
Artists and entrepreneurs will no longer get a fair chance at their content being seen, and instead be elbowed out of timelines by brands with big budgets.

Anyway, it's not happening. Yet.


"Nothing is changing right now" - aka, not now but soon. And I know this because Instagram is owned by Facebook, and Facebook constantly love to ruin our lives with changes.

Whingeing aside, looking at these stats you'll probably agree that algorithms which filter out content are a good thing:

  • 3 hours of content is uploaded onto YouTube every minute
  • 3 million photos are uploaded onto Facebook per day
  • 4.75 billion pieces of content are shared on Facebook per day
  • 6,000 tweets are tweeted every second
  • 70 million photos uploaded onto Instagram per day

Algorithms are literally saving us from our own excess.

So we can see what we really want to see: photos of our ex's new ugly partner

However, it's hard to trust algorithms because they are not always accurate. Sometimes they get it right, like when I've searched for murder conspiracies online, and later an advert will pop up for a crime autobiography on Amazon.
But sometimes, like every time I watch a video on YouTube and get shown an ad for 'ClearBlue' pregnancy test, they get it wrong, and the targeting can be pretty sloppy. Yes, I am female, yes I am of child-bearing age (cry), but no I have never shown any interest in baby-related things or babies themselves; please stop assuming I have..

break my heart why don't you

Thanks to cookies, there is a wealth of data on each and every internet user. But it's an immature business with insanely vast amounts of information which lacks human intervention, so it's near impossible for the algorithms to get it right every time. It's just a computer. Instagram is more of a personal platform than other social channels - people have specific behaviours and feelings towards it and the content. That's why they like it. How can a computer formula understand this?

Like all social updates, we will adapt because we have no choice. Perhaps users will like the new format and feel that being shown more tailored content brings richness to the platform. And let's face it, life will definitely be better without seeing any posts from Susan.

No comments:

Post a Comment