In the world of modern advertising, on-demand reporting and real-time data have changed the way advertisers analyze and strategize their marketing campaigns.
In the ’90s, advertising campaigns were more holistic in the sense of evaluating the results on a lifetime basis, however, the exposure to the real-time data changed the way advertisers analyze and optimize their campaigns daily. Since nowadays, all advertising platforms are self-serving and show real-time results, advertisers analyze results they are getting by the minute, by the hour and by the day.
If you’re selling online and actively advertising or have been advertising in the past, you might have noticed an interesting phenomenon. It normally occurs shortly after you start running your paid ad campaigns and when your ad starts getting delivery.
It begins with a little engagement on your ads, followed by some clicks to your website, then things start to heat up with some people heading to the checkout page or you have sign-ups and then finally the event that you have been waiting for: an actual purchase. The holy grail conversion that anyone selling in the online space desires.
Aside from the amazing phenomenon of actually getting a purchase from someone that can be on the other side of the globe, you probably noticed something interesting that happens after someone purchases from your site: The second purchase came faster than the first one and the third purchase came even faster, then the rate of conversions that followed continued to increase until it hit a peak.
What happens behind the scene is a phenomenon that I call the “Snowball Effect” in which, like a real snowball that is rolling down a snowy mountain increases in size, you find that the bigger the snowball gets, the more powerful it becomes. So it is with your advertising.
Under the hood
In order to understand how the snowball effect happens in your ads and how to use it to your advantage, it’s important to understand the logic of how advertising platforms work.
Most large advertising platforms like Facebook, Youtube, Google, Snap Chat, Tiktok, and more are algorithmic and pixel-based. This means that they use machine learning to automate your ads using the data that they receive from those ads and pixel as feedback for performance.
Essentially the logic is simple: Instead of telling the platforms how you want to run your ads, you tell them what you want to achieve and they will optimize your ads towards achieving those results at the best cost possible.
In other words, instead of setting up a set structure for your ads and defining your specific audience, simply tell the algorithm what you want to achieve and it will use machine learning to automate your targeting based on finding users who already completed the same action in the past. Simple as that. The more data the algorithm has, the better the results.
That’s where the Snowball Effect comes in to play. Think of the difference between a campaign with 2 conversions (results) vs a campaign with 1 conversion. The campaign with 2 results has 2X the amount of data to work with. Now think of the difference between a campaign with 100 conversions vs a campaign with 10 conversions. The more conversions a campaign has, the better it can predict opportunities and the more cost-effective it will be.
Optimization happens on a campaign level, and learning happens on an ad set level and is combined with data-learning from the website’s pixel. You will notice when you start increasing your spend on your ad accounts, spend on the actual campaigns becomes more horizontal and less vertical. This means that even if you have 20 campaigns running with large budgets, you will notice that the spend across different campaigns is not equal. Machine learning will predict, from all your campaigns, which ones have the highest likelihood to meet your goal and will alter the reach towards those campaigns.
That’s when the Snowball Effect will take place and your best performing campaigns will become bigger and generate more and more conversions. The more conversions they have, the faster they will generate future conversions. Once the peak has been reached and all cost-effective conversion opportunities have been maxed out, the snowball will start to slow down (the steep downhill ends) and eventually will stop. This happens naturally because there are not infinite opportunities in the market and there are only a specific amount of conversion opportunities available at the cost you can afford.
Successful advertisers understand this process and ensure that by the time that one campaign starts underperforming, a new and improved campaign is being formed and will eventually replace it to tap in on the remaining conversion opportunities. Understanding this logic will help you better prepare and structure your campaigns to take advantage of the snowball effect. Also, understand that the more data each campaign has, the better value you are able to generate from each impression that you purchase on the platforms.
Comprehending the Snowball Effect is important, but it’s more important that you understand how to use it to your advantage so that you can maximize your results. Let’s show how to use it to your advantage.
Planning the right structure.
If large campaigns outperform smaller campaigns because they are loaded up with conversion data, your goal is to have the highest ratio of campaigns that have completed the learning phase and have generated a lot of conversions. The more conversions each campaign has the more efficient the audience optimization will be and the more precise the targeting will be. If your spending is more vertical across many of your campaigns, the less data each campaign will have and the less effective the campaigns will be.
This means that you should plan to have a limited amount of campaigns running at once to ensure that the conversion opportunities and ad delivery is narrowed down to fewer campaigns.
Once your campaigns start generating conversions, remove any barriers by rapidly increasing the budgets to ensure that they can generate as many conversions as they possibly can, as long as they meet your ROAS and CPA goals. In addition, start pausing campaigns that are underperforming whilst keeping the top campaigns running. The goal is to have as many data-rich campaigns as possible to ensure that every ad impression will be targeted toward the most relevant audience.
Accelerate the Snowball Effect
The information above might seem common sense and you may have already understood this process in different terms, but let’s talk about how you can accelerate the growth and scale of your campaigns to maximize results.
If two ads in two different campaigns are performing well and are gradually growing into becoming your top campaigns why not combine them? If two ads target the same audience why should they have a separate learning system?
One common practice among advertisers is to leverage the learning system of advertising platforms by combining data and learning to accelerate performance. Simply take your top ads and combine them by duplicating them into the same campaign to accelerate their individual performance as well as their overall performance. If you combine your top ads into a single learning system, they will share the data and achieve conversions faster thus giving the campaign a clear advantage over others and delivering you the best results.
Learning how to combine data and most effectively leveraging how advertising platforms work will allow you to create higher levels of efficiency and get the most out of every dollar spent on ads. As your campaigns start to gather more data and learning, the level of conversions begins to grow and gain momentum like a snowball rolling down a hill. Eventually, you will find the momentum slowing as your campaign reaches its peak effectiveness. Once that happens, look at refreshing your campaign.