You may remember a few weeks ago I took out a Youtube advert. The advert was one of those ‘In Stream’ adverts that you can skip after 5 seconds. The advert linked to a full version of this song.

You’ll be shocked to learn it was about as much use as an ashtray on a motor cycle and no one clicked through to the full song. Although there was one strange insight where the advert had a higher than normal view retention of 50% (benchmark is 15-40% apparently). I surmised either the Google Ads algorithm is counting views incorrectly or folks hung around for the song but didn’t click through because there was no clear call to action on the advert.

Second Youtube advert
I wanted to run another version of the advert. This time with a button signposting users to the full version of the song on my Bandcamp page.

I also found a way to target the advert a bit better. Google Ads allows you to be a bit more clever with who sees the advert but it does take a bit of digging around the system:

So what happened?
I found out that viewer retention went up and now about 60% were watching 30 seconds or more of my advert:

But how many users actually clicked through to the full song on Bandcamp?

2 people. Out of the 411 views, mostly watching more than 30 seconds of the advert (allegedly), 2 people clicked through to the full song on Bandcamp. I sad.

Key finding
Users really don’t want to be distracted from the video they want to watch so in-stream adverts did not help me promote my track on Youtube although it may have helped raise brand awareness given the good view retention scores.

Third Youtube advert
Armed with another huge marketing budget of around $15 and some more clue of how Google Ads worked I set off on another epic adventure to promote my soon-to-be-number-one international synthwave EDM hit.

This time I was going to choose a different song and used a different type of advert which was a discovery ad. These show up in the related content bar or search results when you’re watching your favourite EDM hyper banger:

One observation was Google Ads forces you to choose a title and subtitle for your advert, I guess to lure users into clicking it. I think this requires lots of experimentation to get just so, but I chose whatever best came to mind; “New synthwave tune” and “Brand new single”. With hindsight this was probably uber lame.

So what happened?
I would say this approach didn’t not work. I only got around 90 folks clicking through to my song over the 5 days it ran. This represented a click through rate of about 1% which is standard. Of these 90 views clicking the advert and going to my synthwave tune, I got one new Youtube subscriber and 3 likes. This is a small victory.

But it turns out Google Ads behaves in unexpected ways which might seriously trip you up if you try to take out an ad.

The first, is that for reasons between Google Ads and the Dark Lord Satan, it decided to target users unevenly:

I don’t know about you, but music played out of tablets or mobile is sucky. So I had to go away and change that so these devices were excluded.

It also skewed the keywords to synthpop music:

View retention on the third advert
I found out view retention was lame when users did finally click through to the tune:

There are two lines here: the orange line is users coming in from the advert and the blue line is organic users who found the song naturally (I think). The users coming in from the advert lost interest very quickly compared to organic users who tended to hang around (which is nice).

Key finding
Even if your advert successfully drives traffic to your full song, you still only have a few seconds to capture their attention. Forget promoting your 11 minute EDM banger with a five minute white noise intro.

Summary
Here’s my key take aways:

• Google Ads is fiddly and hard to learn
• In stream ads don’t work
• Discovery ads kind of work
• The track you link to must be catchy

But what do you think? Used Google Ads? Like synthwave music? Hook me up in the comments and let me know.