How did I get millions of extra views on YouTube without any ads?

In a word—annotations.

In this YouMoz, I'll explain how following a 5-step process enabled me to increase my annotation click-through rate (CTR) by 22,400% (from 0.2% to 45%), how I received 150,000 views from annotations, and how this resulted in millions of views.

All the screenshots and examples I’m using are from one of my channels, DoctorMadScience:

The most important indicator on YouTube: watch time

Before we jump into optimizing annotations, let's talk about why annotations matter. With the help of YouTube annotations you can increase your viewership, grow your subscribers, sell merchandise and drive traffic to your website.

But one of the most powerful uses of annotations is to increase your watch time.

YouTube describes watch time as one the most important metrics for measuring the success of channels and videos. One way to understand the importance of watch time is to take a look at the new analytics dashboard. YouTube recently revamped its analytics dashboard and one of the big changes is how prominent watch time has become as a metric. Watch time has, for the most part, replaced views. The old "Views report" is now "Watch time report," and the first element of your analytics dashboard is the watch time report.

What's watch time?

In 2012, YouTube shared the following message:

"In particular, we’ve discovered that time watched is one of the best indicators of a viewer's engagement. As a result, we’ll be focusing more prominently on time watched in providing Related and Recommended videos starting next week. [...] What matters is that your audience stops clicking away and starts watching more of your videos. " YouTube Blog

What this means is that if I get my viewers to watch more videos and stick around on YouTube longer, this sends a strong signal that they're happy with the content they just watched. As a result, YouTube will send free traffic to my videos by showing them as suggested and related videos.

That was 2012. Since then, watch time has continued to become a more important metric. Just last month, YouTube launched their subscription service, YouTube Red. The revenue generated from the subscriptions of YouTube Red will be split based with creators on their watch time and not views.

My results

It's been more than a year since I have started using this strategy on this channel. Below is a snapshot of my yearly traffic sources after I implemented my 5-step annotation optimization strategy. Since then, Suggested videos as a source of traffic source has grown to become my largest source of organic traffic, with more than 4.2 millions views. I attribute 150,000 of those views to my annotation strategy alone.

My starting point

Here are some of my metrics before I started using my 5-step plan:

Annotation click-through rate (CTR): 0.02%

Annotation close rate: 5.62%

More people were closing my annotations than clicking on them.

My end results

After months of tweaking and modifying my annotations, not only did I drastically increase my CTR, but I also managed to decrease the annotation close rate.

Annotation CTR: 48.6%

Annotation close rate: 1.57%

My 5-step annotation optimization strategy

Step 1: Find your best opportunities.

Step 2: Understand viewer behavior (retention optimization).

Step 3: Create your annotations.

Step 4: Write an effective call to action (pitch).

Step 5: Measure and optimize your annotations.

Step 1: Find your best opportunities

1. Visit your 28-day analytics views report.

2. Find out which videos are getting 80% of your views.

3. Select one or two videos to get started.

➜ [Video Guide]

Don’t add annotations to your videos just yet. First, let’s find an annotation structure that works for one video. The best place to start is with one of your most popular videos, and then scale from there. It’s very common for one video to receive a significant majority of your total new views. To find your most popular videos take a look at your 28-day analytics report. It allows you to see the breakdown of views and watch time per video for the time period selected. The 28 day measure is long enough to give you meaningful data and short enough to provide recency.

Since 70% of the new views (and watch time) are happening on the “Milk + Soap = Magic ” video, I chose to only focus on that video as my starting point. Once you figure out a successful structure for the annotations on one video, then you can scale the structure to other videos.

Step 2: Understand viewer behavior (retention optimization)

1. Go to the Audience Retention tab of your analytics dashboard.

2. Select the videos you chose in step 1.

3. Find the drop-off point for your views on the videos.

➜ [Video Guide]

Now that we have a video, let’s figure out the time when annotations should appear. The two most common ways channels use annotations are by placing a button over a simple screen or multiple b-rolls with annotations, like the picture below.

These types of annotations almost always appear at the end of the videos. There is no doubt they look cool and very engaging, but they have one big problem: Timing.

Most of your viewers will not watch until the very last second of your videos. Your drop off point is usually long before the last second of your video. So a good chunk of your viewers don’t even see your annotations, let alone get to click on them.

YouTube does a great job at is providing super detailed analytics and powerful tools to help you understand your viewers' behavior. By taking advantage of these, you can find the best time to show your annotations and have the information you need to optimize them even further.

One of my favorite tools is Audience Retention, as it allows you to track viewers' drop off points on each of your videos.

Audience Retention is very similar to scroll depth on a website. The lower the content is on a page, the fewer the number of viewers who see it. The same is true for videos. The closer your annotation is to end of the video, the fewer viewers see it.

How to read your audience retention metrics

On your analytics dashboard, click on the Audience Retention tool. If you’re logged into your YouTube channel, click here to go there directly. Now, select the video you plan to include annotations on. Now you should see a chart similar to the one below.

The X-axis shows the time on the video and Y-axis shows the percentage of viewers who watched that part. A number higher than 100% means your viewers replayed that part of the video. As you can see, there are less than 50% of my total viewers left by end of the video. Your drop-off point is when you start losing views at an increasingly faster rate.

I included mine at the 1:32s mark:

Simply by showing an annotation before my drop-off point (20 seconds earlier), my annotation will be viewed by 20% more viewers.

Also, at this point, my viewers haven’t decided on what they want to watch next, so they’re more open to suggestions.

The best part is the data tells me the exact time to show my annotations.

Step 3: Create your annotations (placement and size)

1. Go to the annotation editor for your video.

2. Add the annotations based on the drop-off time.

3. Position the annotation on the lower right and make sure it's noticeable.

➜ [Video Guide]

By this point, you have found the best video to start with, figured out the time to show the annotations and simply have to create the annotations.

You do this by simply going to the URL of your video and clicking on the annotations button. Or you can just put your video ID at the end of youtube.com/my_videos_annotate?v=[VIDEO ID]

In step 1, I decided to show my annotations 20 seconds earlier, and this comes with a compromise. When you have your annotations at the end of the video, you have the luxury of space. The video is finished and there is nothing else going on except your annotations. But now we are showing the annotations much earlier and in the middle of our prime content. So our options for the placement of the annotations are limited to the sides of the video.

Annotation location

Where you place the annotation has all to do with your viewer’s attention pattern. This is very similar to the way you think about placing your call to action button on a website. By understanding your viewer’s attention patterns, you are able to place the annotations in the area where they are most likely to be noticed and effective.

After doing a lot of testing on the position of my annotations, I found that the lower right section of the video gets significantly higher click-through rates compared to the left side, 40%-60% more in some cases.

Recently I came across an eye-tracking study done by Sawyer Ricard on eye tracking on the layout of video websites. They used eye-tracking technology to understand the attention patterns of viewers on video platforms, including YouTube. One of the main findings was that many of the viewers fail to focus on the video and get distracted by the suggested content on the right side.

Using the results from the eye-tracking study, I ended up with what we see below:

Viewers are getting distracted by the content on the right side of the page, and this is where they will choose the next video they watch. Placing your annotations within the viewer’s natural attention path will grab their attention much easier. So annotations on right side do have an advantage.

Annotation size

Choosing the right size is tricky. My rule of thumb is to start with the largest size you can get away with it, as large as you can without taking away from the main content. Then look at the close rate vs. CTR of the annotations. As long you have more people clicking on them rather getting annoyed by them, you’re golden. Then let the annotations run for a few days and check the results.

But how can you tell if your viewers find the annotations annoying?

Annotation close rate

With YouTube analytics you can almost track everything! YouTube allows you to track and compare the percentage of people who click on your annotations vs. the percentage of people who closed them.

Based on steps 1–3, I completely changed the way I use annotations, and now focus on a single large button that’s timed to appear well before the video actually ends.

Structure of the annotation

Call-to Action: The best call-to-actions (CTA) are super simple. I limit mine to one word.

Title/Pitch: I’ll use this space to pitch and sell the viewer on watching the next video. (I’ll explain how to write an effective pitch in the next step.)

Step 4: Write an effective CTA (pitch)

1. Find the shared context of your audience.

2. Write a simple title based on the shared context.

3. Paste the next video's link and publish your annotation.

➜ [Video Guide]

The most important element of a great pitch is context. Context defines how your copy is understood. First, you have to find out what’s the one thing all of your viewers have in common. Is it their age, location, interests, or language?

Probably not.

By the time my annotations appear, they will all have one thing in common: They just watched my Milk + Soap = Magic video.

So the most effective way to write a call to action that's relatable to the largest number of viewers is to use the context of the video they just watched.

Milk + Soap = Magic is a simple science experiment using milk and soap to make some cool effects. With that title, my viewers understand my video title structure:

• Title - Milk + Soap = Magic

• Structure - Item 1 + Item 2 = Cool Effects

This will help create a very clear and accurate expectation of what the next video is.

Now I have a short but very effective way to describe the rest of my videos. Borrowing from the title of the video they just watched (Milk + Soap = Magic), I can now write the pitch for next video: Wool + Battery = MAGIC

So instead of being super descriptive, I'm only using the title of the video they watched to describe the next one.

Now I'll just add the link of next video to my annotation, and I'm done.

Step 5: Measure and optimize your annotations

Step 1. Check your annotation CTR.

Step 2. Check your annotation close rate.

Step 3. Tweak, wait, and optimize.

➜ [Video Guide]

You must wait long enough to get a decent sample size (400+ unique views or so) before you can judge the performance of your new annotations. Optimizley has a great tool for you to figure out how large your sample size needs to be. (Resources: Check here for unique views info; for annotation CTR, read this.)

The other metric you have to keep an eye on is your annotation close rate, which can be tracked against your CTR.

As your annotations get more views, your close rate will might go higher, but the keep your focus on keeping your CTR higher than your close rate. To bring down the close rate, try to tweak the size, placement and timing of your annotation.

If your close rate is too high possible, one of three likely culprits is at work:

a. The annotations are showing up too early, when the viewer is highly engaged with your main content.

b. The annotations are too large and distracting.

c. Their placement is taking away from the main content.

Results

Optimizing annotations does take a while to yield results, as you can see from graph below. It took a lot of trial-and-errors for me to find the sweet spot.

I hope the steps I've given you in this post give you a solid framework for improving the results you're getting from YouTube.

I know many of you have done your own experiments with YouTube, and I would enjoy trading tips with you in the comments.