Instagram Reels: when a view is not a 👀

25th April 2021

Hello, I’m a marketer and I’m an addict. Before you hand me the number of a professional and exit the page, hear me out on my addiction. I’m addicted to 0’s. In the marketing world, vanity metrics. And Instagram Reels is making those views worse.

More 0’s means more ROI! Right? As the cynicism drifts from the pixels on the page I think not one of you agrees with that statement.

But here we are.

The platforms, particularly Facebook, enable my addiction on the product side by ever-increasing dopamine releasing features, and on the marketing side, addiction makes good business sense (for them, not so much for my budget) so they love it even more.

And, they keep feeding us numbers to satiate that addiction which translates into spend with the promise of Reach.

We keep gobbling it up. It’s a cycle based on questionably thin air, though. TikTok is making it worse.

When a view is not a view, or how I learned to love my audience.

Firstly – what is a view, anyway? Well….depends who you ask:

Platform views per second

Subtle shifts in how we use channels causes big changes to how we perceive the importance of that platform.

Instagram Reels was rolled out in response to the rocket ship that was pandemic TikTok. There’s nothing more to add really, it’s a clone.

It’s not available to advertisers, so our commentary below is more related to how the culture of obfuscation is confusing us about what our audience is interested in.

Numbers are where we want to examine, as the key metric of what attention can stand for is unwittingly on display.

As we know, this is especially true in the Arabic-speaking world, where we tend to be the earliest and strongest adopters of new features.

For example, Look at this chart based on Instagram’s own metadata across a selection of Arabic Reels measured in the last month – these are the totals of what is displayed as a ‘View’ (the icon you see on the feed) vs. what the definition of a ‘Play’ is in Instagram’s language (3+ seconds):

Views vs plays

Put more explicitly – here’s an example Reel of what a user, and creator (remember, Facebook/Instagram Insights do not include Reels data) might see:

Instagram reel views example

372k Views! Awesome. But…is it?

On a scan of the metadata sitting behind the image we have a different picture of how many people saw the post vs. people who view the post (presumably using Facebook’s defined 3sec v. a TikTok play of counting from load):

Views and plays of instagram reel

So that juicy 372k is probably a good indication of a load, but the difference in the two figures is stark. In fact, it’s 48.2% different.

That 192k is certainly not what the consumer sees, nor an analyst deciding what platform to place content on for say an influencer campaign. And with influencers, the results can mean that we’re not getting the full picture.

It fundamentally skews what is trending, what is important to the consumer, and importantly, how much attention we are able to command. In this case, it appears about 66% in total of what we get reported (again, for 3 seconds).

Popularity is the signal next to the play triangle. The dopamine is somewhere, as we seen in our examples, between 48-66%.

In fact, maybe, maybe, we could call it an impression. It’s certainly not 3 whole seconds of someone’s undivided attention.

You could argue the 3sec mark isn’t really a good measure either. The missing attention metric is surely watch time.

While advertisers get access to certain pieces of this data, the decisions we make ahead of the buy are influenced by the topics, channels and creative we produce. It becomes a cycle.

Implications

The implications of this are far-reaching – ponder this:

  • Creators think they are delivering more audience than they are;
  • Reports are showing trends that are not trends;
  • We get a better idea of what the true reach of any influencer is;
  • We don’t fully understand the potential, or limitations, of a new platform;
  • If they’re doing this for the largest content creators, what are they doing with ad money?

I’m not hating on Facebook here, TikTok is guilty of starting this all to begin with, their view metrics are load-based as well.

What I’m advocating is that we need better, clearer metrics on attention than we currently have.

Take-away, please.

We need to do better as marketers. We need to question our advisors, internal teams and experts on what’s behind the numbers that are continuing to be reported. Particularly in the Arabic-speaking world.

Attention is the commodity we need to break through – you can’t do that unless you understand an audience and properly strategise for them, not a bunch of 000’s on a twerking Reel.

Back to basics, we need to consider the totality of the consumer’s needs in their digital world.

What interests them – it’s probably not your brand. So in understanding how they behave in that world, putting aside noise metrics like views, we can better position our communications with the valuable audience we define.

Ambient feedback is one way to do this, but human smarts are needed more than ever to connect the dots. While at D/A we do both, I’m not selling here, but advocating for a shift in how we speak.

I’m here to implore us to do better – use the layers to understand what drives the audience – that’s the only way to create a connection in 2021.

I challenge you to question your metrics at every turn.

The above was published as a newsletter was sent in April as a part of D/A’s monthly series: subscribe here.