I Made 31 Instagram Reels Until 1 Went Viral. Here’s What I Learned
TLDR: I wanted to concretely know how a short form video “going viral” might affect an unestablished social account, as I work with a lot of smaller EdTech creators. Although there are many variables, my initial findings indicate an account with no previous traction can expected 7 followers for every 100K views one reel receives, but no followers for total Reel views.
More nitty gritty below:
Background: I work for an EdTech company in San Francisco, and my responsibilities include producing cool stuff with edtech creators. To be as helpful as possible, I started pondering the questions creators ask with their brands — what does the sales funnel look like for creators on social media, and how many views translate to followers down the line? How long does it take to gain traction, and does that rate change with generic or niche content? How organic can someone expect that growth to be? What does short form video mean for a creator’s strategy, both short and long term?
I wanted to approach some kind of answer to at least one of these questions. Given I was most familiar with Instagram and knew how to make a Reel, I decided to more concretely define an experiment.
Question: What happens to a previously unknown Instagram account when one of their Reels goes viral?
Defining Viral: For the purpose of this experiment, “going viral” means a piece of content gains wide exposure in a short period of time, say 100K+ views in 7 days. Frankly I didn’t know if these were realistic numbers or not, nor did I know what size account generally sees growth like this, but it seemed like a good place to start.
Choosing an Account: I decided to use my personal Instagram for this exercise; it’s an account that doesn’t have promotional history but does have a long enough engagement record for the algorithm to know I am a real person, not a bot.
Why Instagram?: It’s the platform I knew the best and one of the most important platforms for edtech creators. Not that TikTok isn’t, but I thought resources were best spent on a tool I already knew.
Aren’t there tools for this already? If you have a Creator or Business account, Instagram provides Reels Insights (a tool launched in May 2021) that tracks the engagement data bigger brands need. If you’re just starting to validate an idea, the best tool is trial and error. And starting November 17, 2021, my own trial began.
Hypothesis: With 1 viral video, largely nothing will happen. People will engage with the video, and not the account.
Prediction: This is a quantity play; If I post reels at least once a week using only trending audio and filters, I predict it will take 6 months to get 1 reel with 100K+ views, and I will get 0 followers from this action.
Knowns & Variables:
Steady States & Knowns: I’d only use sounds that were already trending (aka had little arrows next to the audio); I used my personal account and never posted to my feed or stories (aka couldn’t count on friends for promotion or engagement); I’d only post a few words in the description; I’m a blonde-ish, young white woman, which may inform whether someone chooses to follow me or not; for better or worse, I consume a lot of reels and generally know when a trend is on the rise — plus, I’m not afraid of a bit.
Dependent variables: Views, Likes, and Follows
Independent variables, with a reminder that this is “scientific method lite”:
- Duration: Reels varied from 7 seconds to 58
- Location: I’d include a location if it seemed relevant to the content (ex. skiing in Utah), otherwise I left it blank
- Filters: If a filter was trending (including “beauty enhancing” ones) I would use it
- Content: The video’s subject entirely depended on the audio or filter trend; usually it would be my face, but sometimes it would be just nature or something else
- Time of Posting: I (initially) split between posting at 9 AM PT, 1 PM PT, and 7 PM PT. This did not last long.
- Instagram’s Algorithm: I saw several tips floating around that Instagram favored videos edited on Reels (an effort to staunch the flow of re-uploading TikToks) and that the algorithm favored videos 6 seconds or under. My experiment doesn’t provide a large enough sample size to know, since most trending audio is short regardless. I don’t know how much my own reels consumption informed the exposure my videos received. I don’t know if being an individual account versus switching to a creator account informed the algorithm either. A deep dive for another time.
Method: On average, I would make 1–2 reels a week. I didn’t broadcast the reels or otherwise inform anyone about my experiment, which made it all the more fun when IRL friends came across them organically. For the most part, I posted consistently (with two notable exceptions) from late November to earliest April, around a 19 week period or just under 5 months.
Observations: Accepting that performance is measured in views, likes, comments, and follows (excluding saves for now), here were some initial observations:
- Not including the outlier, each reel averaged 1,866 views
- While most reels got the most traction right after posting, that was not always the case — so don’t delete anything! You never know when something is going to hit.
- Since most conventional wisdom suggests making videos below 10 seconds gets the highest level of engagement, I decided to examine how different lengths of video compared. In this sample, reels under 10 seconds and reels 10 seconds and over had similar engagement rates (averaging 1.45% and 1.41% number of viewers liking the video, respectively) but fairly different volume — even without the outlier (which had 146,000 views) videos under 10 seconds acrued 31,246 views while videos 10 seconds and over totaled 25,754 views
- I quickly learned posting in the morning (PT) garnered more views than the afternoon or evening
- Reels posted after what functionally reads as a 4 week hiatus underperformed even the earliest reels (though that might be due to the fact that earlier videos had more time for folks to engage with them)
- Posting too close together proved worse for performance — all four days that had two reels posted on the same day, the first reel performed 2 to 12 times better than the second
And then, of course, there was 1 reel to rule them all …
A Viral Result: The outlier received 146K views, 746 likes, 2 comments, and led to 10 follows. The video in question is 6 seconds long, features a video of an Ocean Beach sunset, and text that matches the beat of the music. Though I can’t strictly attribute 10 followers to the viral video, they followed my account in the week it went viral.
Unlike every other reel, this video gained traction weeks after posting. It was one of my earliest videos, and sat at around 500 views (similar to another early nature reel) until all of the sudden … it wasn’t. Every other reel had its highest traction right after posting, but not this one.
Conclusion: I wanted to explore how viral short form videos affect small accounts. I hypothesized that nothing would happen, even if something gained spectacular traction. After concluding this experiment, 10 strangers decided I was worth their time. I’m happy to discover I was wrong, but why? How did it happen?
I only created 31 reels over the course of just under 5 months, and I wasn’t nearly as disciplined as a standard brand would be. Which of the variables was the biggest informer of success? A second experiment around a brand new creator account (one that’s actually eligible for Reels Insights), with much stricter records on when a post is made, how long a reel takes to make, and when it receives the most traction may be needed to find out.
I have a different hypothesis for this next experiment: Short form video is a great resource for brand awareness, but not an effective strategy for niche accounts; anything other than lifestyle brands and personalities may have a hard time seeking the conversions they want, and their resources are better spent elsewhere. (However, lifestyle brands and personalities can make for great distribution — a post for another time). Stay tuned.
Feel like drawing your own conclusions? Check out the sample data here.
Follow along for more thoughts on EdTech, the creator economy, continuous learning, books, and designing your life.
Are you an EdTech creator in the design, data, cyber, or software engineering space? I want to talk to you! Reach me here.