Hello anyone, everyone, and welcome to the week I review. My name is, uh, Amateur Analytics… Analyst, because today I want to talk about Google Ads – the other side of the equation that more popular folks on this platform use to pay their bills.
Note: This is going to be very specific to my experience using Google Ads as a way to expand the potential reach of my channel. If you’re looking for a comprehensive evaluation, this ain’t it.
Hundreds of hours of video are being uploaded to YouTube every single minute. And even though it is one of the most popular websites in the entire world, it’s oh-so easy to just get buried. And I don’t really like promoting myself all that much. I don’t even post every one – or even most – of these videos on Facebook, because I don’t want to inundate my pretend friends with the nonsense my subscribers at least actively signed up for. I tweet about them, but literally zero people click on the links I post on Twitter, so…
A few months back, I was talking with a colleague about this whole thing. He told me that I should buy fake subscribers to boost numbers and he would find me a service that would do it. I was genuinely offended by the suggestion. On both a practical and a moral level: hell no.
But around the same time, a new button appeared in the YouTube Studio: Promote. I had never really considered the possibility of advertising my videos, but the appearance of this button coincided with an email offering me $100 in ad credit if I spent $50 in real money – which seemed like a pretty good deal.
So I thought, Even though I won’t actively promote my videos, I do think they are worth people’s time, so I’ll buy a few ads and see if I can’t get people to agree. Worst case scenario: I’ll get a video out of it.
I planned to start with my review of the game GRIS, but unfortunately I made a big dumb error there that left me much too embarrassed to promote it. Instead, I went with my breakdown of a few royalty-free music subscription services, which is a pretty good video that took me a helluva lot of time to make.
And if we take a look at that video’s analytics… yep, that’s the start.
This ad was the most expensive I’ve done so far and also the worst in terms of folks actually clicking on it – both things I’ve learned useful lessons from.
On YouTube, there are two types of ads: video (the pre- or mid-roll that you might have seen on this video if I had 5 times the subscribers) or Discovery, which are the highly visible videos that show up in search as an ad.
Because I don’t think these videos in their entirety are conducive to the former – and the joke 5- and 15-second options I made were definitely not as funny in practice as they were in my head – I have gone exclusively with Discovery.
What made that initial video so expensive was a lack of understanding on my part of how the payment system works. There is a bidding system, where you determine how much you are willing to spend on a given action: either cost per view – CPV – or cost per (thousand) impression(s) – CPM. I assume on the other end, folks can do the minimum they’ll accept for an ad to be run against their video.
I have stuck with CPV, because I like the guarantee that it comes with. General awareness of the existence of this channel is less meaningful than someone actually checking it out… but it also fixes the cost. A well-run ad could have a high click rate, ultimately resulting in an overall cheaper ad. Or it could be terrible. Higher risk. Maybe higher reward. Perhaps I’ll try it out sometime.
Ad targeting is kind of fascinating for someone who has absolutely no background in advertising or marketing. The number of options you have to select potential viewers is kind of exhausting, particularly since I don’t know what my target demo actually is. So, I typically ignore that kind of stuff and focus on interests and search keywords, which is even lengthier but at least I feel like I understand. I like to get that click rate as high as possible, which doesn’t matter, per se, with CPV but gives a sense of how well I’ve aimed the ad.
Indeed, that first video had the worst of them all, with a measly 1%. And I know why it happened, because I targeted “music” as an interest, which resulted in a creator-focused video comparing licensing services being placed against, like, music videos. Now I’m more careful, but it is easier with something like my review of the very bad Netflix movie Bird Box. If I target “Movies” as a general interest and tag the keyword “Netflix,” I’m pretty much set.
I actually posted videos and ran ads on Bird Box and Black Mirror: Bandersnatch simultaneously – ya know, in attempt to be on trend. The ads resulted in decent numbers of views even though I went for an CPV of only 2 cents, vs the 6 cents default on the music subscription – 2 cents is now my go-to CPV. But the thing is, as soon as the ads ended, those views flat-lined and have barely moved since. They might be two of the highest-viewed videos on the channel, but not meaningfully.
My discussion of the competing Fyre documentaries, on the other hand, is not too far off in view count, but only 257 came from the ad. The other 1300 came organically. What does that tell me? I dunno. Maybe I shouldn’t have done the Birdbox review three weeks after the fact.
In any case, I have also stopped putting nearly as much money into ads as I did when Google’s credits were doing the heavy lifting. Now that it’s just me footing the bill, I am typically more in the $5-10 range – which is just a little boost in visibility without breaking the bank.
Plus, that little boost can be enough to trigger the psychological component to this whole thing: take the Music Subscription video. It’s not like post-ad viewership has spiked, but it has consistently grown, and I think that the fact that it had passed 1000 views is at least partly responsible. If you’re searching through videos on YouTube for something specific, would you choose the thing with 100 views or 1000? I’m typically going to go with the latter, and the ad was the differential there. So running ads that give myself a new, slightly higher base line, can make the channel seem a bit more popular to the people who care about that kind of thing.
This is, however, fundamentally different from “buying views” on some unscrupulous internet marketplace because I am really just paying for premium placement of my thumbnails. Folks who otherwise might not be subjected to my dumb face will now be forcibly confronted with it. They can then decide of their own volition if it’s the way they want to pass their time. It’s definitely a small, almost insignificant advantage, but the video is also clearly marked as an ad, so it seems like a fair one to me.
I changed some things about the way I start my videos at the beginning of 2019 in direct response to an ad that I ran. It was on the Netflix anime Aggretsuko’s Christmas special. Fewer than half of the viewers made it through the nine seconds of intro. And I’ll admit that it wasn’t my best “clickbait title,” but that was a pretty rough awakening. And so I removed that entire thing, replacing it with testing out my potential thumbnail faces.
And then I went even further by cutting the first second of the opening drum fill. The intro is now four seconds long.
And then I pushed my initial thoughts to the video title, which I think is honestly the bigger deal. Would my Fyre video have gotten the views it did without a clicky name? Not a chance! Would some of my earlier videos have done better if I had been a bit more creative in my naming? I’m sure.
My video on Salo, or the 120 Days of Sodom definitely benefitted from its title. I didn’t run an ad on it, and yet it has just this past weekend dethroned the Netflix-hating videos in terms of views, and it has long since destroyed them in terms of Watch Time.
Another metric it has beaten any other video in: subscribers gained. More than 11% of my subscribers have come from folks who, I would assume, liked what I had to say about Salo. Another 10% were from one of my first four videos, actually, which had started strong but fell off when Moviepass finally left the public consciousness. That Bandersnatch review did alright there, too, which leads to a question: because almost every view on that video was ultimately paid for… does that mean I “bought” those subscribers? If so, how much did they cost?
And here’s where we get into the question of practical value, for me, of running ads.
Views are nice, I guess – we talked about the psychological benefit of that already – but it’s actually kind of depressing on my end, because most people who click the ads don’t stick with the video, resulting in hella low watch time for a video, which also isn’t great algorithmically – and may be part of why my Netflix ads tanked after the fact.
This harsh reality could be offset by a growth in subscribers, hopefully the start of a beautiful viewership, but we’ll have to break out the calculator app on my phone and Excel on my laptop to figure out the average Cost Per Subscription across the ads I’ve run so far to find out.
To start, we need to look at the number of subscribers I received from each of the videos during the time the ad ran. I was actually a little surprised to learn that this comes to only 28 – or about 13% of my total count, and only a few more than Salo has gotten me on its own. So, we’re already starting off pretty low.
I then need to factor in that not every person who saw the video came from an ad, so we determine that percentage – number of views resulting from the ad divided by the total number of views for the duration – and multiply its count by that. Round down and we get 26. Then we divide that by the overall cost of ads thus far – $149 – which equates to… uh,
Five dollars and fifty six cents.
Even if you factor in that I got $75 in credit from Google (dunno what happened to the last $25) and so didn’t really pay for half of the ads, that’s still $2.81 for the ones I did buy.
Not a great return on investment, if that was indeed how I was thinking of it.
Particularly considering that taking my colleague up on the offer for some fake subs would have gotten me a whole lot more for a whole lot less. While writing this script, I looked up the going rate for a subscriber these days. Seems to be about 10 cents on the high end, with other options running well below a penny.
Whole lot less than what I’m doing… but oh so much grosser.
Look, my subscriber count is small enough that I get actually excited when I see it tick up. There’s a genuine satisfaction in knowing that someone decided that they wanted to see when, for example, this video was going to come out and took a couple of seconds to press a button saying so. They almost definitely weren’t expecting this video in particular, but… life is more exciting that way! For them and for me.
I get that YouTube is a job for a lot of people and many, many more would like it to be that… and that you need a certain number of subscribers and engagements and etc. in order to get that underway, but cheating one’s way to legitimacy is a bad thing done by bad people who, on a platform built largely on perceived honesty, are proving they don’t deserve success.
So, what have we figured out here? Running ads doesn’t give me a particularly good return, and also is a sometimes frustrating experience – I called technical support due to an issue and was told that there are known problems with view counts in ads and the youtube studio not matching up; and I had to dispute a block placed on the Daniel Sloss video because machines don’t understand context – but I think I’ll keep running them where it feels appropriate to do so.
For example, there is a 100% chance I’m going to run this one as an ad, because I think that’s funny. And in general, there are much worse things I could spend $5ish dollars on. To someone who’s struggling financially, this doesn’t seem to be a great way to really expand your channel… but let’s be honest: the entire foundation of The Week I Review is my ability to burn small amounts of money week after week. So, running largely useless ads fits with my MO.
Also, the genuine confusion in this comment is hilarious and something I want more people to feel.
Six Point One out of Ten