As with Facebook, Google [and therefore YouTube] are making lots of significant changes that will totally change the landscape of media buying and ultimately what works on the platform.
The reasons for these changes range from Google wanting to sell more of their niche inventory (they know in-stream placements are most profitable) to wanting to remove targeting options to make the platform easier for newbies to attract more business.
There are probably a bunch of reasons beyond these discussed in board meetings we don’t even know about.
The bottom line question is: “What are YOU gonna do about these changes?” This may sound overly dramatic but we are in an evolve-or-die situation.
That may not seem to be the case right now but these changes are only the start as media buying becomes more AI-focused and control is wrestled away from the media buyer.
Below I have compiled a list of changes past and current and the solutions our Google media buying team are putting in place, there is also a bonus at the bottom but please don’t skip ahead, you are very likely to miss the meat and potatoes
Change # 1 – Removal of in-stream campaigns, replaced with responsive ads
This change was announced at the end of 2021 and implemented at the beginning of 2022. For those unaware, this change removed the ability to just show your ads only on the ‘in-stream’ placement on Youtube.
Google’s official name for in-stream campaigns was ‘true view for action in-stream campaigns’ and this is the format that is now obsolete.
Your responsive video campaigns will now show on in-stream placements but also
– Search results on YouTube
– YouTube Watch Next Page
– YouTube Hompage
– Discovery placements
To use a Facebook analogy, in-stream was ‘feeds’ only, and the responsive video campaigns they were replaced by are ‘automatic placements’.
Google realized they were missing out on revenue due to unsold and underutilized placements/inventory. They also saw that ‘automatic’ style campaigns showed an increase in conversions likely due to the different touch points or lower CPM on placements outside of in-stream.
In the last few months of 2021, your current in-stream campaigns were allowed to run but no new ones were able to be published. And in the first few months of 2022, these campaigns were slowly phased out, spending was reduced on these campaigns until they were no longer spent at all.
So there’s the boring background part done. Now let’s get into why this change still matters and what you should be doing to ensure your continued success using the responsive video campaigns format in 2022 and beyond.
With your responsive video campaigns now showing on in-stream placement but also
– Search results on YouTube
– YouTube Watch Next Page
– YouTube Homepage
– Discovery placements
Your thumbnail and also in some cases your video title will show on these placements so optimizing these is of utmost importance and something most advertisers aren’t doing nearly well enough.
For thumbnails, I recommend using a headline and sub-head and keeping it below 60 characters so it is easy to read and not too small. This is especially crucial for the Youtube homepage on mobile devices.
Also including an image of yourself works really well. You should think about the thumbnail, in the same way, Youtube creators do. You are trying to hook someone and get them to watch the video so this should be reflected in your headline and thumbnail design. If you want some inspiration, check out Alex Hormozi’s channel or Derek’s More Plates More Dates channel, they have both put a lot of time and effort into finding what works for thumbnails when it comes to composition, saturation, and subject matter. There is a lot to learn from analyzing their thumbnail strategy.
The next thing to consider is instead of using an internal naming convention for your video, use a headline as this will show in some placements.
You will obviously still need to identify the ad so I offer you 2 options, first is a headline then the internal name in brackets like this: ‘Get Traffic, Build a Huge Audience & Sell Your Course (Traffic Bundle 30s V2)’
The next option is to ditch the internal name and go with a more YouTube creator-style name like this: ‘Stop Worrying, Guessing, and Doubting Your Trades (Episode 35-B)’
You can then label each new ad as a new episode to keep track of what ad angle and variation this ad is, which is crucial for not only keeping track of your winners but identifying which hook, CTA, or thumbnail is performing best in your account so you can use it to guide further ad testing.
Change #2 – Removal of keyword targeting
This has been a big announcement from Google recently, they are removing the ability to use keyword targeting and topic targeting for YouTube campaigns. Keyword targeting in a YouTube campaign targets search terms from within YouTube. It is distinct from custom intent targeting which is based on Google search terms.
Much like the responsive video campaign changes they aren’t taking effect immediately. If you have keyword campaigns live for YouTube right now these will still run but you can’t create any new keyword-targeted campaigns.
They haven’t announced the full timeline but I imagine some time in the first few months of 2023 these campaigns will be phased out completely.
Some of your best campaigns right now might be targeting keywords so what can you do..? Well, I have a plethora of other options for you and I suggest you start testing them soon, so you aren’t left high and dry when your keyword campaigns die.
The first is probably the most potent and something we have been implementing at Hemon for a while now.
To preface this, unlike keyword targeting, the problem with custom intent targeting is that Google doesn’t show you the data for which keywords are converting the most and which are wasting money.
The way to combat this is single keyword custom intent audiences. This allows for more control and ultimate clarity on what targeting is working.
One caveat is that you need to select keywords with a good amount of search volume for this to be effective. If you go too niche your campaigns won’t spend out. After you find 4-5 winning keywords using this method I would recommend combining them into one keyword stack and testing that. Often, giving Google more people to go after will lead to better results.
The next one is URL affinity audiences. This was mis-sold by Aleric Heck when he said that it allows you to hijack your competitors’ traffic.
In reality, what Google is most likely doing is scanning your competitor’s website (or any website you choose) for keywords and then using these keywords to build an audience for that campaign.
This can be very useful. We have even had success using our client’s sales page as a custom URL affinity audience. The reason for this working is often in your headline and page copy are all the keywords and audience callouts needed for Google to target your ideal audience effectively.
Another type of website to use for this that works great is an article or blog. Again, these are keyword rich, and the good ones are optimized for SEO already.
In-Market and affinity audiences are something we regularly utilize. When it comes to higher-spend accounts and looking to scale above $5k/day, they are something you should definitely be exploring.
To break down the distinctions between the two, In-Market audiences are anybody Google perceives to be actively searching for a product or solution in a particular category, it could be Pet Food, Automotive Parts, or Business & Productivity Software. An Affinity Audience, on the other hand, is a group of potential customers that share similar interests or qualities.
You will need to do some research and find out what (if any) audiences fit your target demographic. The big advantage of these audiences is that they are normally large and if you get them to work they can scale up pretty well.
Affinity audiences are slightly different in that they aren’t made up of people recently searching or showing purchase intent. Rather, they are made up of people who have shown continued activity for a period of time and shown an affinity for certain kinds of content on Youtube, this could be Fitness Buffs, Yogis, or Cooking Enthusiasts.
These are also normally larger audiences. I would compare them more with interest targeting on Youtube. If you really know your avatar, what they are interested in, and the kind of content they consume on YouTube these audiences can prove very profitable.
Often when I audit accounts for clients or other media buyers, I see them using mainly keyword and custom intent targeting. This will prove to be a real handicap as Google continues to make changes to the way targeting is done on the platform and what really moves the needle in your account.
Change #3 – Removal of Ad rotation, optimizing for the best ad
This isn’t something that Google has announced but that doesn’t mean it’s unfounded. Speaking to a lot of YouTube media buyers, the way they run their creative tests has had to change over the last year to 6 months.
There remains an option at the ad group and campaign level to ‘prefer best performing ads’ or ‘rotate ads indefinitely’. As far as I have seen from practical experience this feature no longer works and is no longer supported.
You will probably see in your own campaigns if you have 2 or more ads that the spend is in no way split evenly between the ads in the ad group.
It still works to test ads with 3-5 ads in the ad group but it becomes hard to optimize them. When a poorly performing ad is turned off, you then need to make a note and only judge campaign performance from this date, especially if you are using Hyros.
This causes 2 problems. First of all, it results in more unnecessary work. And, secondly, it causes a longer process for testing ads. In a media buying culture that responds well to the speed of iteration and optimization, this just won’t do!
The alternative option that I and others employ is testing campaigns with a single ad. I still recommend testing this 1 ad against 2 or more of your winning audiences but this method, whilst burdening you with a higher volume campaign, will help you to have absolute clarity on how an ad is performing and how much spending it has received.
This helps you avoid wasted ad spending by turning off ads that aren’t performing sooner. It also helps you identify winners easier which will allow you to use the winning hook or thumbnail across more ads and campaigns.
The same process for optimizations is still followed. Looking at CTR and CPC before leads, calls, and sales start to come in and turning off any leading metrics outside the benchmarks you set for your account is the name of the game.
You’ll then want to let these campaigns run to 1.5X – 2X your KPI and if they don’t have a conversion you can turn them off or let them run slightly longer if leading metrics are strong.
Change #4 – Ads over 3 minutes being penalized in auctions
Going back to another topic that Google has announced, there appears to be a new initiative to penalize long videos. Indeed, videos longer than 3 minutes seem to be penalized at the auction, and thus result in worse outcomes for the brand.
This makes sense as it’s beneficial to Google. The shorter videos are, the more of them they can show. There is a little bit of nuance to the impact this has on your ads and what you should be doing with this new information from Google.
Hearing this news might have you running to create new ads shorter than 3 minutes right away… but don’t be so fast.
This isn’t actually new news, Google has always penalized longer ads, they just put a specific time on it recently.
In the past I have tested with success running a 16-minute and also a 30-minute VSL as an ad. Now obviously Google is going to penalize a video that long to a much higher degree than one that is 3 minutes and 50 seconds long.
So the question then becomes at what point should you take notice of the higher CPMs for longer ads?
I’m not saying you should not test shorter ads. With the rise of Tik Tok and Youtube shorts, you should definitely be exploring the contribution shorter style ads have for your business. But ultimately there is a marketing trump card that negates all this (most of the time).
If your messaging is on point and your ad not only hooks attention but converts that into action, the marginal difference in CPM won’t affect you much.
I realize as I’m writing this that anecdotal evidence is the weakest evidence of all on the evidence hierarchy, but I have found that some of my best-performing ads in a number of accounts have been 4-7 minutes long and have no doubt have been penalized in auction for their length.
So how did they still work?
It turns out there is no substitute for a great product-market fit and messaging that resonates. Before you worry about how long your ad is, you should spend all your time on these two factors as they are the true recipe for success.
If these two considerations are dialed in, the question won’t be whether your ad will be a winner or not, but rather to what degree.