Traffic Sources Will Help You Hack The YouTube Algorithm

Understanding this simple data metric will help you reach new audiences

Dear Creator Nation,

Happy New Year! 🥳 

I hope everyone had a wonderful holiday and Christmas break. Now that our team is back together, it’s time to get back to work. But we’re excited. Pumped. In fact, no one’s ever been more pumped to return to freezing cold lakeview Chicago in the heart of winter.

But that’s what makes this company so special. We’re a squad of 9 teenage and barely legal undergrads who like to build cool stuff together regardless if it's -15° and windy AF.

We’re also not shy about our vision for AMA. Three years ago we signed our first client (a personal favorite might I add) with a single goal in mind: help YouTubers grow through data. 🚀 We’ve stayed true to our roots over the years, providing unique insights and clarity to creators in a world of messiness and confusion. 

With that same vision in mind, we strive to help not just a few dozen but hundreds, even thousands of YouTubers leverage data, community-building, and content strategy to take their channels to new heights. To do this requires an incredible amount of strategy, coordination and execution for each and every moving part, including new product launches, a full-fledged content ecosystem, and of course prioritizing our existing clients. 

If you’re new here, I hope you’ll join us on this journey 💜  If you’re a regular, we appreciate you sticking with us.

🚨 Before I get to the data, I wanted to give a very special shoutout to our bestie Nathan Graber-Lipperman who just launched the first ever creator magazine called Creator Magazine.

Of course, we can’t fail to mention that the magazine featured none other than our very own founder Mateo Price. Read more about his incredible vision and at times bumpy journey transforming AMA from a dorm-room idea into a startup generating over $500,000 in revenue.

🚦Understanding the Algorithm Through Traffic Sources

A central flaw in YouTube Studio is lack of accessibility. YouTube displays hundreds of numbers, graphs, and charts but doesn’t explain what each means and why each matters. This is like taking an introductory Spanish class and the teacher doesn’t speak a word of English (ahhh, I think I’m projecting). Anyway, traffic sources are a difficult concept to grasp at first, but once you understand them, you’ll learn a lot about who your audience is and where they’re coming from.

First, let’s go over some definitions:

• Browse traffic → Viewers of a video that come from the YouTube homepage, subscription feed, or the watch later icon

  • = engaged subscribers and people who frequently visit the channel (aka the algorithm knows them)

• Suggested traffic → Viewers who click on a video from the “up next feed” 

  • = inactive subscribers or new audiences (i.e. the algorithm recommends videos they might enjoy)

  • increasing suggested traffic is a measure of success in reaching new viewers

*Note: The ratio between suggested and browse traffic typically hovers around 1, meaning for every 1 browse view you receive, you also receive 1 suggested view. When the ratio dips below 1 and towards 0.5 (i.e. browse > suggested), this is an indication that content being posted is not as engaging for new viewers and thus is not recommended to new audiences.

What does this mean for you? Well, in order to reach new audiences, you need to determine which types of videos have high suggested traffic.

Let me show you ⬇️

We analyzed data from a client of ours that specializes in motorcycles (hey, we support a diverse clientele). After grouping their content into different categories, we observed a breakdown of their traffic sources and noticed some interesting things:

1. The channel drives a large # of viewers from YouTube search and browse traffic, which indicates that the majority of their viewers are loyal fans who know what they are looking for

2. Beyond the graph above, when the channel posts more videos, their suggested/browse ratio dips even closer to 0.5

3. Browse traffic > suggested traffic for every video type except “Bike Building”

4. THIS tells us that the algorithm doesn’t know what types of people to show these videos to OR their content is not accessible to new viewers

To reach new viewers, we recommended that this channel makes “limited series” and “stunts” videos more digestible, consistent, and viral to ensure that these new viewers don’t leave the channel after discovering their videos. 

😴 TL;DR

1. Traffic sources = where are your viewers coming from

   • Browse traffic = viewers from YouTube homepage, subscription feed, or watch later

     • ↳ driven by loyal and regular viewers 

   • Suggested traffic = viewers from YouTuber’s “up next feed”

     • ↳ driven by new viewers who were suggested this content

2. Usually, suggested traffic = browse traffic

   • BUT, when suggested traffic/browse traffic ~ 0.5, content is not being recommended as much to new audiences

3. Prioritize content that has high suggested traffic or revise underperforming content to make it more accessible and digestible for new viewers

🪜 YouTuber To-Do List!

If  you are a YouTuber, here’s how to learn more about where your audience comes from:

1. Sign in to https://studio.youtube.com/

2. ↳ from the left menu, select ‘Analytics’

3. ↳ from the top menu, select ‘Reach’ 

4. ↳ scroll to ‘Traffic source types’

5. Repeat our analysis above: group your content into categories → note where traffic comes from → observe videos that reach new views and those that don’t → plan accordingly 😈

🎥 Video of the Week

🚰 Youtuber: Matt D’Avella

👀 Views: 415k

👍 Likes: 27k

💬 Comments: 777

🔔 Subscribers: 3.27M

📋 Content for Creators:

1. Exploring video length best practices for social media by Jenn Chen

2. How the YouTube Algorithm Works (Or Why Your Videos Aren’t Getting Views) by Amir Jaffari

🎬 Action Steps:

1. Follow us https://twitter.com/amagrowth

2. Think this would help a YouTuber or creator you know? Forward it to them!

Have a great rest of the week creators!

Sent with 💜,

Noah from AMA