YouTube AdSense Revenue Optimization

New strategy developed by Hardly Creative

Python Processing of Large Data Sets | Optimized Bid Adjustments Daily

Get My Free Consultation

Hardly Creative has developed a process to improve AdSense revenue via scaled Google Ads campaigns.

This strategy is designed for clients looking to run 24/7 YouTube Channel campaigns over a longer time span.

Google Ads
Data Studio

Strategy Background

Based on our extensive YouTube Ads experience

We have built thousands of Google Ads campaigns for clients and record labels globally, and have been able to identify several pain points that occur when using a marketing platform to push entertainment content:

  1. Limited budget.
  2. Revenue is difficult to attribute to ad spend, making marketing budgets complicated to justify.
  3. Difficult to generate revenue when promoting music videos for new artists. Unless the artist has a robust merchant operation and tour schedule, marketing spend can lead to an abundance of unrecoverable costs.
  4. Improper targeting, which causes wasted budget and putting ads in front of users that are not likely to enjoy the content.

Over the course of 2018, 2019, and continuing into 2020, we have been developing a strategy with our clients that blends the value proposition of the entire YouTube technology stack; Google Ads, YouTube, and AdSense, in a way that benefits creators and marketers tasked with growing their entertainment brands.

All four of these pain points can be alleviated with a unique campaign build and optimization sequence that Hardly Creative has developed.

Strategy Overview

Finding the YouTube “Rabbit Hole”

The high-level goal of our strategy is to optimize Google Ads campaigns for AdSense revenue. This results in earning cash-back for the client, allowing for recuperating some or all their ad spend. This, in turn, results in YouTube Channels finding their appropriate audience and developing real followers more quickly.

We are using Google’s powerful marketing platform to target videos to users that are likely to watch additional content on the channel. We pay Google for the first view with an ad buy, but what we want is for the user to enjoy and continue watching content on the channel, what’s commonly referred to as the YouTube “rabbit hole”.

As the user racks up time watching content from the same creator, additional ads are served by Google. ~90% of the revenue earned by YouTube Creators come from skippable/non-skippable pre, mid, and post roll ads. That is not the ad type that we are buying for our strategy. We buy TrueView Discovery Display ads that appear on the user’s home page, search results, or at “position zero”, which is located under videos on their watch-page.

TrueView Discovery Ads

ROAS Improvement

Our optimization sequence runs once every 24 hours

Through careful optimization techniques and a rules-based approach which Hardly Creative has developed, our goal is to optimize volume ad spend for revenue; getting you up to 100% Return On Ad Spend (ROAS) via Google AdSense.

In addition to AdSense revenue, which our system optimizes for, scaled, targeted traffic, can also lead to increased revenue via YouTube’s merchant shelf, Superchats on live streams, and memberships (typically requires a minimum of 30k subscribers to enable).

Revenue is also generated via ticket sales with geo targeted tour information that appears to users below every video on the channel (requires Official Artist Channel designation by YouTube).

Remarketing opportunities also allow for the client to capture audiences and send them to their own website for ticket sales and other merchandise sales.

Google AdSense makes payments to content creators on about the 21st of each month for the previous month’s earnings on YouTube. This is roughly 5 months faster than the velocity of Spotify payments and other music platforms, making this scaling strategy viable for artists and entertainment brands with lower budgets that need revenue payments in a timely manner.

A Different Approach to Campaign Builds

How to think about this strategy

This new process is completely different from normal campaigns that current Hardly Creative clients are used to. These are 24/7 evergreen, channel wide campaigns. There is no set end date for the campaigns, the spend occurs on a system billed by Hardly Creative to the client on a weekly billing cycle. All costs are updated in real-time in Google Data Studio, that our clients will have access to.

Previously, campaigns were built with a total budget in mind and a start and end date. Clients wishing to switch to this new strategy will need to switch to a new, more client centric, billing cycle and understand the differences in the total approach we are taking with YouTube marketing.

Multiple videos need to be used for A/B/C/D/etc. testing to determine highly performing videos and downgrading those that are performing poorly.

The goal is to generate engagement for users. As can be seen below, this is accomplished via auto-play functionality (1), algorithmically created music playlists that appear in position (2) for all Music Channels, or via strong presence within suggested videos lists (3):

YouTube Suggested Videos

Learn about how Hardly Creative handles client billing

Billing Process

The Process

Learn about our procedure and requirements for AdSense ROAS Optimization

Hardly Creative and our client sign an agreement laying out services provided and expectations from the client in regards to billing, data access to channel analytics, permission granting Hardly Creative the right to use this data for optimization purposes, and any additional goals laid out in the agreement.

After the signing of the agreement, the process to client goal achievement begins.

Setup and Initial Campaign Build

We begin the build process based on the information discussed with the client regarding country targeting, audience targeting, and other factors.

On new channel setup we will also need to connect to the YouTube channel from the Hardly Creative Google account. At this point, that requires the channel owner making Hardly Creative a manager on the channel. We need this access in order to pull revenue data on a daily basis.

Later in 2020, Google has announced that user roles are expanding, and channel owners will be able to give viewing access only to Hardly Creative which is our preferred method of access. Until that has been engineered into the platform, this is currently the only method that gives our tools the ability to scrape revenue data and other metrics that we need from the YouTube Analytics API every 24 hours.

Review Period

After campaigns are submitted for review, it generally takes 2 business days for Google to review and approve all videos. Sometimes it can take longer and is not a process that Hardly Creative controls.

Continual Daily Optimization

As with all machine learning and rules-based systems, we need copious amounts of data so that we can adjust campaign bidding based on results. When a new channel campaign begins it will likely be 7-14 days at a minimum before we have enough data for our changes to begin making a positive impact. Adjustments are made to channel campaigns every 24 hours based on updated revenue and performance data.

YouTube is a constantly changing environment for users. The auto play up next algorithm on some days will be very generous, other days it won’t. Suggested videos lists are different for every single user based on YouTube’s algorithms.

There will never be a day when our optimization stops; it’s a continual, daily operation.

YouTube Channel Requirements

This strategy requires some important elements

YouTube Partner Program

A channel needs to be accepted into the YouTube Partner Program and have all video monetization options enabled. All YPP policies need to be followed.

Entertainment Brand

This strategy is specifically built for content creators creating videos meant to entertain and keep users on YouTube; music videos, vlogs, documentaries, travel videos, informative and entertaining content that easily garners strong watch-time and audience engagement (likes, playlist adds, subs, comments). This strategy is not for companies advertising a service or product that requires users to move off YouTube in their watch-session.

Robust Content Catalog

To effectively deploy this strategy, the more content a channel has produced the better the results will be. A channel needs to be somewhat established with data and content direction. This doesn’t mean the channel needs 100k+ subs, even 1k subs can work if the channel has a clear identity and plenty of content that has already been produced and released. Specifically, if we can identify the content that already performs well organically, we will use that content to further improve our optimization.

Wide Appeal

Content needs to have a large enough potential audience. Some channels will be too niche to run volume targeted traffic to. Most music genres from metal to jazz and classical work just fine. 

Professional Branding

With this strategy we will be driving voluminous targeted traffic to videos on the client’s YouTube channel. Video thumbnails, profile picture, cover photo, the about page, video descriptions and the videos themselves need to be cohesive and have clear branding. We recommend Music Ally’s service for conducting a total overhaul of channel optimization and SEO improvement.

Complete Content Ownership

For musicians this means ownership of both the publishing and recording rights. If performing a cover song, the song needs to be properly licensed from the publisher so that you avoid a copyright claim from the publisher which severely impacts channel monetization revenue and execution of this strategy.

Proprietary Elements

There are two components of this service that we consider to be proprietary

  1. The campaign build out structure, account settings used, targeting parameters, ad group segmentation, levers being utilized, etc.
  2. Our process of blending and optimizing data for our revenue KPI. This includes our entire rules based and machine learning system, access to our Python programming, any details pertaining to the optimization process, or our constant, iterative approach to improving our tools, logic, decision-trees, and overall results.

We have conducted hundreds of experiments as part of this strategy development and consider it a competitive advantage for our agency to retain this information internally.

Clients will have full reporting access via custom built reports in Google Data Studio from integrated data bridges; campaign overview, country reports, engagement reports, and other custom reports with data that is pulled and refreshed daily from the Google Ads and YouTube API. We will be able to build out custom reporting needs for each client with Data Studio.


Google Ads/YouTube Analytics data blending

As we calculate ROAS for each country we are running ads in, there is currently no possibility of filtering out revenue earned on YouTube derived from preexisting organic activity on the channel. For channels that are already earning substantial amounts in AdSense revenue daily, our data science approach to optimizing tens of thousands of ad group bids will be forced to use that revenue data in the total ROAS calculation.

What that means is our system is likely to bid higher for views in Google Ads based on the better ROAS from already high earning channels. The goal will still be to obtain 100%+ ROAS as we scale Google Ads volume in each country. We are looking into ways to tone down our machine learning adjustments for channels that have strong pre-existing revenue so that we don’t overpay for views in Google Ads.

One possible solution is we tweak the optimization settings to focus on 200%+ ROAS if a channel is already a high revenue producer, that, however, remains to be tested by our data science team.

As bids increase as a result of strong ROAS, that doesn’t necessarily mean our auction prices will be equal to those higher max bid ceilings. Google video ads still operate on a first price auction model, meaning we will only pay one unit higher than the next highest bidder, not the max bid our system uses.

Cost Expectations

Country data isolation

Each country we run ads in will be considered an isolated container. Country costs from Google Ads will always directly correlate to country revenue from YouTube. As we scale a channel’s ad delivery, each country will contain thousands of ad groups, each ad group needs a minimum baseline of data for our system to decide on adjustment.

In statistics we need a sample size of 10 for each variable being tested for. So, to adequately acquire the data necessary for constant daily bid improvements, we have set a minimum of $20/day budget per country being targeted.

This would mean clients should have a budget of $600/month for each country they wish to run advertising in. The more countries for which we can collect data, the better our system will optimize.

For each channel, we recommend identifying a minimum of 3 countries the content is likely to perform well in or has a history of performing well in with accompanying revenue. We also need to start off by targeting countries with large populations and that have a robust advertising market. Some countries perform much better than others, and often it’s where you’d least expect. This is a component that requires multiple countries so we can test and find the highest earners for your YouTube channel.

Revenue Expectations

What kind of ROAS should you expect?

We cannot make any guarantees or promises regarding revenue. Our system is designed to improve ROAS over time as more and more data is acquired from campaigns and YouTube, but each channel will perform differently as our system scales ad delivery.

Our goal ROAS is 100%, which means the client is running scaled campaigns for their YouTube Channel and earning back what they are investing in marketing.

What will typically happen is ROAS will be roughly 5 – 25% when starting out, and clients should generally notice an upward trend of that ROAS as our system optimizes ad delivery and bidding.

Each channel is different, each niche/music genre/video production resonates differently with different audiences. Also, the niche and audience will determine the value of the content to other advertisers within Google’s Marketing Platform. Content that is more valuable to advertisers depending on demographics, geography, income level, remarketing ability, and other factors will result in higher earning rates with Google AdSense. These are all factors our system can’t control.

Revenue is also dependent on the client properly monetizing their content within their YouTube Channel settings and having 100% ownership of their content. A cover song that is not properly licensed by the publisher will result in a copyright claim from the publisher, and revenue for that video will drop by 50%+, making revenue optimization nearly impossible.

Additional Revenue

More than just AdSense

We strongly encourage clients to take a proactive approach to building revenue with their business. Our system focuses on only AdSense revenue but there are multiple systems in place on YouTube for building additional revenue so that creators and content owners will nearly always be making a profit.

One of the largest profit sources for video creators are brand deals. This is a great article which dives into what creators should be charging brands for sponsor integration into their content.

Other sources of revenue on YouTube include:

  • The merch shelf
  • Memberships
  • Ticket sales
  • Superchats
  • Audience remarketing

Hardly Creative focuses on audience optimization for increased AdSense revenue but to truly maximize earnings so that you can not just cover your marketing costs with Google Ads, but make a profit, the entire suite of revenue generators should be leveraged.

Revenue Optimization

About Us

Hardly Creative is a U.S. based marketing agency focused exclusively on entertainment growth on YouTube. We leverage the full depth of the Google Marketing Platform coupled with custom built tools to deploy and optimize scaled Google Ads campaigns for channels within the music and entertainment industry.

Hardly Creative also develops proprietary, data driven optimization solutions with Python, enabling us to process, automate, and act daily on millions of data points generated on YouTube and Google Ads.


Find More Success Today

Get My Free Consultation