I wanted to share a raw email I received from a friend of mine, Junior Reyes, who is a growth marketer but also knows a lot about YouTube. If you want to connect with him, add him on LinkedIn here, or you can also connect directly with him at rjunior1029@fgmail.com.
Below, you'll find everything you need to know about launching successful video ads including topics ranging from video SEO and new video uploads to what YouTube analytics to look for in your next YouTube campaign. I'll let Junior take it from here:
I wanted to use this chance to share some perspectives into 2020 (what a year!) and go over all of the things I learned as well as the improvements I had to make to ensure that my clients' YouTube budgets were as effective as possible. Sit back, grab your favorite drink or snack, and let's talk some YouTube...
YouTube Advertising Performance Overview
Below you can find metrics from of my clients' ad campaigns on YouTube:
- $10MM+ in YouTube Spend (that's an avg. of $27.4K/day in YouTube Spend!)
- Over 55MM+ Views, that's a ~$0.18 CPV across Home goods, F&B, and Beauty brands while averaging CPAs (sales) reductions of ~20% H2 vs H1
- Average ROAS $8.65, which is almost $90MM in Revenue
- Average Split between PROS (net new) and RT, 70%/30%, respectively
If you remember from March 2020 when the first shutdown was announced, CPMs had decreased by ~40-45% on Desktop and 30-35% on Mobile as a lot of big brands panicked and dramatically reduced spending on YouTube.
Well, those that ignored competitors, listened, and kept pushing saw massive returns by capitalizing on the CPM decreases but also the surge in online content consumption and shopping.
In March, I scaled up about 65% of my clients to capitalize on the above, since then they are experiencing a new norm in online shopping and seeing ROAS they've never experienced before. I won't bore you with all the stats on how offline shopping behavior has shifted online or how baby boomers became a demo to pay attention to in 2020 as they were forced to rapidly adapt to online shopping.
My YouTube Marketing Strategy Learnings
Scaling and maintaining YouTube as part of your marketing mix is already a challenge in itself. Now imagine doing that efficiently through all the consumer behavior changes, the political climate, as well as all of the social injustice happening in 2020. Saying it was challenging would be an understatement.
Here are my learnings from testing different video content, monitoring key metrics, and assessing overall YouTube ad performance for my clients in 2020.
Creatives
Keep an eye out for the video marketing trends, they will give you guidance in creating content (vs ads). The perfect example here is MasterClass, they saw streaming numbers skyrocket, thus adjusted their YouTube video(s) to feel more like a Livestream and then paired them with well put together custom intent audiences. The results, an insane 140% increase in clicks to their website (...and people don't click on videos) and a 70% increase in course sign-ups compared to the previous period
I will continue to repeat myself here: you DO NOT need a $200K video production. Shit, you don't even need a $25k video production to be successful on YouTube. Don't get me wrong if you have it, then more power to you, but it still needs to be done right. Set up your channel with some basic channel art, a logo, a channel description, and a playlist featuring a channel trailer. Add some CTAs and overlays to your videos to ensure that your channel looks complete and visitors know what to watch next. Then, you can begin launching ads.
If you're still struggling to create high-quality content, partner with content creators. You can find them by searching on the internet or browsing their Instagram and Twitter accounts to review their social platforms. With existing assets and a bit of editing, and following the structure below you can build a winning ad and many iterations to the A/B/C/X test.
Additionally, knowing the player, it’s a commonly overlooked aspect of the creative process. On Desktop you have a sliding banner that comes in from the bottom-left above the scrub bar and the skip button on the bottom right.
Here is a brand that does a great job at distracting users away from the skip button.
By understanding YouTube's idiosyncrasies across devices, you can begin to understand how to correctly optimize your advertising campaigns and ensure their success.
Account Structure & Optimizations
Ranking your videos organically before using them as ads helps ramp up creatives 2.5x faster than simply using an unlisted video. I tested this out several times throughout 2020. Approximately 90% of the time the rank creatives ramped up 2.5x faster and more efficiently than simply putting that creative in rotation.
To that point, one tool I highly recommend if you are not using one is TubeBuddy. You don't really need all of their features. However, the one I use the most is the keyword explorer to help improve the SEO of your videos. After all, YouTube is the second largest search engine. You should ensure that your brand account is optimized based on keyword research and YouTube SEO best practices including video tags, a relevant YouTube channel description, video ad captions, and more. With optimized YouTube content, you'll increase watch time and drive engagement on relevant videos.
Don't be fooled by YouTube's new VAC format, otherwise known as Video Action Campaign. Initially, I thought it was my best-performing unit until I took a deep dive into the data and found that it was over-crediting itself. There are many ways to check this. One of my favorites is via a conversion lift experiment (not all accounts are eligible). Additionally, you can run a PSA against the same audience. However, after about $1MM in testing budget against this ad unit, I got it to work and CPMs are ~85% lower than True-View for Action format, meaning that if you can efficiently set this format correctly against the right audience, you too can achieve CPA reductions as crazy as 30-45%. Here is how that format looks in the YouTube feed:
The YouTube algorithm is finally adopting the Facebook model of consolidating as much as possible and shifting focus to creative optimization. This is great for ease of management, but something hard to master and not advisable to all brands. It really depends on your budget.
But in short, if you had a 1:4:2 structure meaning two ad units per ad group (4) per campaign (1) and the right tCPA and budget in place, you might run into issues scaling some of the ad groups in that campaign despite their audiences having equal or larger volumes than the ad group taking up the budget, even if performance was better on those other ad groups not getting the spend.
Keep your structure simple, unless you have the budget to test multiple setups.
Lastly, beware of the Audience Expansion box (see below) when setting up audiences for your ad groups. It can cause unexpected spikes and Google will likely not give you a refund depending on your setup. I managed to get over $35K refunds due to the spikes in spend it caused but it was not easy. This tool is highly unstable in my opinion but worked great on some campaigns where I was able to tame it with the right combo of frequency caps (imps/views). For these campaigns, I was able to achieve CPA <= to overall account CPAs +10-15%.
Other Recommendations
For those that do not have a data team (or maybe you do but they are not running the right analysis on your marketing data), I highly recommend Sourcemedium (even more if you are on Shopify). They provide a Data Team as a Service for DTC Brands. Many of my clients use them to understand aggregated marketing channel data performance, run LTV and churn models as well as track HoH performance via their Slackbot. Additionally, their CEO is an amazing growth/product marketer with many success stories and also an ex-Googler.
If you are looking for a dependable partner to help you scale display ads and run a kick-ass Programmatic Prospecting/Retargeting program, hit up Acuity Ads. I've tested them many times and can attest to their ability to deliver log-level data for what they are buying for you but also ensure your ads are viewable and serving to real people. I'm a stickler for measurement and attribution, and I've put them through the ringers so you don't.
This blog post was primarily derived from the email that was originally written by Junior Reyers. Some content has been added and amended from the original source.