MODEL CONNECT

RUNWAY INFLUENCE

LESSON 2:
HOW THINGS GO VIRAL ON IG

To go viral on Instagram, your content needs to be featured on the ‘Explore’ page.

Instagram features content on the ‘Explore’ page based on 4 main things:



(1) Your Saved Photos

The short version: Using image recognition and 'lookalike' audiences, you're shown posts similar to ones you’ve previously saved.  

Instagram analyzes the characteristics of the content that you’ve saved. The theory here is that you save content you really like, so they want you to see more of this type of content.

Why? Because if you keep seeing content you like, you’re going to keep coming back to their platform and engaging on it.  

How? Simply explained, through page categorization and image recognition. 

 

Instagram categorizes pages into niches like travel, memes, nature, models etc. If you’ve linked your Facebook page to your profile, it will also use this information to put you into a category. To do this, Instagram will use data like ‘Page type’, enabling the algorithm to find similar content it already knows you will like. It also looks at both your personal Facebook's and Instagram's digital behavior to make assumptions on interest too.

Instagram also runs image recognition to show you content similar to the posts that you’ve saved. First, it scans all your saved content to find common characteristics between them. Then, it searches for similar content and
 shows you these on your ‘Explore’ page. 

EXAMPLE


Facebook automatically scans every single image you upload, as well as the ones you save. Not only can it detect objects, like mountains or cats, it can also detect activities like surfing or dancing. 

Facebook/Instagram is scanning all your photos and videos and running it through image recognition. 

For example: the above, "1 person, standing, phone, selfie and indoor",  was detected from the below image of the lady in the red dress (@annelisejr). This is a small example of how Instagram and Facebook use image recognition to learn about the media you upload. If you're interested in seeing first hand what Facebook can see about your photos click here. 

As you can see, the algorithm can also detect multiple elements to understand context. The algorithm is able to figure out that this image shows a "person" "standing", with a "phone" taking a "selfie" "indoors". 

Using this technology, Instagram is going to show you more pictures on your Explore page of a person standing with their phone taking a selfie indoors!

 

Instagram also categorizes you as a user based on this computation. For example, if you upload photos and videos of trees and mountains, you'll be put in the "nature" category. 

 

They don't stop there. Instagram puts you into a subcategory based on the specific features of this content. So if the pictures you uploaded of trees and mountains are mostly taken during the day, it'll add the "day" and put you into further subcategories.

(2) Posts that you’ve liked
 

The short version: Same concept as image recognition, except this time the algorithm is run on photos that you’ve previously liked and videos you previously watched to a high completion rate. It’ll find the common characteristics between these to find new similar content to this to show you.

This works in exactly the same way as image recognition. Instagram looks for common traits among videos you’ve watched to a completion rate of above 75%, in order to try and show you more of these types of videos. The assumption is that you will stay on Instagram longer watching more videos if they do this - and their ultimate aim is to always keep you on the platform and provide you with content you enjoy, for as long as possible. 

(3) Post similar to accounts you interact with
 

The short version: The same thing with image recognition, plus Instagram finds posts from accounts you’ve previously engaged with, but didn’t follow after you engaged.

Instagram finds accounts that share similar traits to those you routinely engage with. It does this through tools like image recognition, page categories, and hashtag usage.

 

In addition to this, the algorithm also keeps a log of all your history. If you’ve previously engaged with some accounts but didn’t end up following them, you’re more likely to see their future posts in your own newsfeed.

(4) Based on people you follow
 

The short version: You are shown content that the people you follow engage with. This is why our engagement group is so effective! 

Instagram shows you content from that the accounts that you follow have engaged with. For example, say you follow Justin Bieber. If he likes + comments on a random picture, it's likely to appear the Explore page, (if it has excellent engagement ratios too). Since we understand this formula, we run our engagement groups where accounts in the same niches with different followers engage with each other - helping to drive each other’s following, giving you mutual growth. We offer these groups at Model Connect.  Below is a walk-through example of this works.

EXAMPLE

Let’s say you and I are both in the same engagement group and we’re both fitness models. In this example you have 100,000 US-based followers and I have 100,000 UK-based followers. You post a picture. I "like" this post. Because I did that, my followers see your post in their Explore feeds. Now if you do it for me, I benefit from exposure to your audience. This is how you can get easy, targeted, mutual growth.

 

At Model Connect, we run the largest model like groups.