A $5 YouTube Ad Experiment

YouTube $5 Ad Experiment

I just finished running my first ever YouTube ad campaign through Google AdWords. With a $5 budget, I set out to see just what I could get for the price of a Starbucks latte.

I’m sharing my experience because I know I’m not the only stingy, nerdy, web guy out there wondering about YouTube ads. If you’re considering building up an audience on YouTube, for whatever reason, you might want to consider running a few small YouTube ad campaigns.

The ad campaign was for one video: WordPress 4.4: New Features Montage.

YouTube was spending money a little faster than I expected, so I extended the campaign for 2 days, costing me $7 instead of $5. Was it worth it? Read on to find out…

Strictly By The Numbers

For all you numbers nerds out there, here’s a breakdown of the data.

Campaign Report (from AdWords)

This data was pulled directly from AdWords. It only includes what AdWords can directly relate to the ad campaign. This does not include extra promotion or attention the video may have received as an indirect cause of running the ad. More on the indirect effect below.

  • Campaign Time: 9 days (12/2-12/10)
  • My Max CPV: $0.02 (cost per view)
  • Spent: $7.01
  • Impressions: 2,045
    • 1,958 in-stream
    • 87 in-display
  • Video Views: 556
    • 554 in-stream
    • 2 in-display
  • View Rate: 27.19%
    • 28.29% in-stream
    • 2.30% in-display
  • Avg. CPV: $0.01 (cost per view)
  • Clicks: 6   CTR: 0.31%
  • Earned Views: 3
  • Watched 100% of video ad: 16.88%
  • Earned subscribers/shares/likes: 0

From YouTube Analytics

This data was not directly related to the AdWords campaign, but it represents data about the video that I ran the campaign for, during the same time in which the campaign was running. While I can’t say with certainty that these numbers were a result of the ad campaign, I have good reason to believe that the ad impacted these numbers.

Much of the activity that occurred on other videos on my channel happened on videos that I linked to from the WordPress 4.4 video. I don’t think these other videos would have gotten as much exposure if it weren’t for the traffic that the ad brought to the advertised video.

Data from 12/2-12/10

General Channel Stats

  • Subscribers: 5 gained, 0 lost
    • 2 from the video, 2 from my channel page, 1 from a different video
    • For the entire month of November, I gained 2 subscribers
  • Views: 1,754
  • Watch Time: 2,058 minutes
  • Likes/Dislikes: 5 likes, 1 dislike
  • Shares: 22
  • Card clicks: 5
  • Annotation clicks: 5
  • Estimated Earnings: $2.18

Video-Specific Stats

  • Views: 1,415
  • Watch Time: 1,368 minutes
  • Likes/Dislikes: 2 likes, 1 dislike
  • Shares: 22
  • Card clicks: 5
  • Annotation clicks: 2
  • Estimated Earnings: $1.20

Analyzing the Numbers

Analytics is far more than just numbers. We must interpret those numbers, hypothesize their meaning and test our theories.

The raw data here only tells a small part of the story. I’m not suggesting that this campaign was an enormous, overnight success, but we need to dig deeper to get a better understanding of the true impact of the campaign, and what kind of ROI it actually provided.

All of the following data looks at the 9 days during which the ad was running, compared to the previous 9 days immediately preceding it.

Watch Time & Views

For the past several years, YouTube has used Watch Time as a key metric in its search algorithm. Views can be deceptive if you get hooked in with a misleading title or thumbnail, and only watch a very small percentage of a video. But if you watch a large chunk of that video, it’s a good indication that it was a better match for your search query. Let’s take a look at both metrics.

Watch Time vs. Views

  • Watch Time went from 252 to 2,058 minutes. That’s an increase of 1,806 minutes (~800%). The advertised video accounted for 1,368 of the 2,058 (66%) total watch time.
  • Views went from 122 to 1,754. That’s an increase of 1,632 minutes (over 1,000%). The advertised video accounted for most of those, but remember, the ad stats could only directly tie 556 views to the campaign. 556 out of 1,754 views is only 31%. The other 69% of video views (1,198 views) came from other videos.
  • During the advertised time period, my #2 and #3 most-watched videos (which I did not run an ad for) each accumulated more watch time than my #1 most-watched video from the previous time period.

Watch Time vs. Views Comparison

All About the Benjamins

If you just want to look at money, my 7 dollars didn’t get me a whole lot in return. I got $2 back, for a net loss of $5. That assumes that ALL earnings during this time period are attributed to the ad, which is possible, but somewhat unlikely. About $0.75 of the earnings came from other videos (which may or may not have been watched had I not run the ad).

YouTube Ad estimated earnings graph
Estimated earnings between 12/2 – 12/10

Channel Engagement

Money matters, but it’s not everything. Especially when you’re just starting out. With only 15 videos, 30 subscribers, and about 4 weeks of consistent uploading, I didn’t expect to make a fortune. I was hoping to at least make my money back, but I went into this with low expectations. I was just curious what I could get for $5. Aside from what AdWords told me, and the $1.50 in revenue I received, courtesy of my advertised video, what else did I receive from my $7 video ad?

  • Heavily increased watch time. With this being a key metric in YouTube’s algorithm, increasing Watch Time for a few of my videos will help them show up in future searches.
  • A few (10-15) subscribers. Even though none were directly attributed to the ad, I’m confident that the 5 subscribers I received during the campaign, as well as the 8 I’ve received in the week following the ad, probably wouldn’t have happened without the ad.
  • Video embedded on a few sites. Looking through the external sites in the playback report, I noticed a few legitimate blogs that have embedded my video. They only account for a few views here and there for now, but this helps increase my reach over the long haul.
  • Brand recognition. Specifically in the WordPress community. People who watched my video saw the name & logo for WP Smackdown. If they see it in future search results on Google’s SERPs, or social media, they’ll be more likely to remember it. And hopefully care a little more about what we have to say (on YouTube & our website).

With so few card click-throughs, very little traffic was sent back to wpsmackdown.com.

YouTube Subscriber Increase

Ad Types: In-stream vs. In-display

In-stream ads are the skippable video ads that play before other videos. In-display ads appear in search results, as recommendations in the sidebar when watching other videos, etc. They consistent of your thumbnail image & some text.

I setup the campaign to run both types of ads, and optimize for the greatest exposure. Turns out, I only received 2 views from in-display ads. And they cost me 6x more than each view of my in-stream ads.

Learn more about In-stream vs. In-display video ads in YouTube’s help center.

In-stream vs. In-display ad performance
In-stream vs. In-display ad performance

What Did I Do Well?

Timing

I launched the ad campaign 6 days before WordPress 4.4 was released; but there was already significant buzz happening. Multiple blogs had already posted about upcoming features. People were curious.

I did see the most activity on the day of the release, but there was plenty of action during the days prior. Not to mention the following 2 days, as well.

I recommend using Google Trends, researching other sites & blogs, and just your general knowledge about the topic, and consider the timing of your ad campaign.

Related Content

I only received a few card & annotation clicks, but that’s in large part because I only spent $7. I did, however, make sure that the video I advertised had references to other videos in it, and I used annotations & cards to help promote it.

Use one video advertisement as an opportunity to promote your other videos. Show potential subscribers that you’re more than a one-trick pony. You’ve got other great material waiting for them.

What Would I Do Differently?

Use a video with a longer shelf-life

I chose the WordPress 4.4 New Features video for a couple of reasons:

  • It was relatively short (just over a minute).
  • It was much more catchy than a boring tutorial video.
  • It was one of the most searched for topics that I had a video about.

Looking back, however, it also has a short shelf-life. I significantly increased both views & watch time, but in a few weeks, no one will care about WordPress 4.4 anymore. The searches will die down, and all the work I did ($7 of ads) to increase watch time… won’t continue to yield any positive results.

Had I advertised a video that consistently gets 1,000+ searches/mo., and will remain a hot topic for months to come, the watch time would yield better exposure in search results, and ultimately provide more ROI.

Run only In-Stream ads

I paid 1 cent for each view of an in-stream ad, compared to 6 cents for each view of an in-display ad. I only got 2 views from in-display ads, so it’s tough to measure engagement with any level of accuracy, it’s safe to say it’s likely not even worth it. Had I spent $20, $50 or $100, I think my money would have yielded much better results if I stuck with in-stream ads.


If you’ve ever run a YouTube ad – of any type or magnitude – I’d love to hear from you in the comments. What did I do well/poorly? What advice do you have for me, and our audience, about ways to improve performance? How did your campaign do? If you’re open to sharing some numbers, I’d love to compare.

What Are Your Thoughts?

All fields are required. Your email will not be published.

You can use standard <code> and <pre> tags to post code examples, or a service like codepen.io.