Simplify Your Ads - Jason Burlin

Simplify Your Ads


It wasn’t always easy. Self-advertising online wasn’t always an easy task. Although Google Ads & Facebook Ads were a self-service advertising platform from nearly the beginning, many advertising platforms required a sales or an ad representative to set up and manage your ads. Luckily enough, we are beyond that era. Now, everything is designed to be self-managed and the main goal of advertising platforms is to make ad management simple. Simple enough so you won’t have to hire an expert just to walk you through and help you press the buttons so you can create your ads in the right way. 

Even 5 years ago, when I thought advertising platforms were user-friendly enough, they became even more simple, and it doesn’t stop there. You’ll often see your ads manager being updated to incorporate new changes and updates that make your life easier. The point that I am trying to make here is that since advertising platforms have simplified the way you create and manage campaigns, advertisers have been looking for complexities in other areas. Advertisers mistakenly think that a more complex advertising process translates to more effective results. Advertisers may fall into the trap of creating over the top ad setup and segmentation to a point where it makes no sense whatsoever. 

An industry emerged that of gurus, of online courses all proclaiming to have figured out how to advertise online. Baseless strategies like top to bottom funnel, mid funnel, low funnel, and many other strategies that follow the herd. Advertisers try to copy-paste strategies and are thinking that one strategy fits all. After all, if it worked for me, why wouldn’t it work for you, right? Many seem to forget the most important thing, strategies are built around your product, not the other way around.

It’s Algorithmic Targeting. 

Algorithmic targeting is a term that I like to use to refer to when you let machine learning automate the entire process of your ad delivery except for the creatives. I think the best way to understand it is to think of it in the form of a game. Let’s say that you wanted to teach a computer a game and you wanted the computer to get better at it and achieve better results. Instead of telling the computer how to win the game, you would tell the computer what are the rules of the game (how much you want to spend and what is your target results) and what players (and creatives) he can use. Obviously, if you would want to point at and roadmap each step that you think the computer should take to win the game, then it won’t have a better chance of succeeding than you would, right? But if you give it all the players and tell it to find a way to win the game based on what it learned and what it predicts, then in a matter of seconds, it will be able to make better calculations and predictions than a human.

The same logic applies with your ads, a complicated structure that tells algorithms how and who to target is created to fail. Like the example above, the idea is to tell the algorithm your objective is, what budget you are willing to spend, and what are your goals are in terms of cost per result, then feed him a bank of ad creatives based on the creative resources that you have available and let it do what it does best, get you results for the best cost possible. It’s literally that simple. 

If you’re running ads and you find yourself in one of these advertiser’s circles, then you will see that many advertisers follow very similar setups even though there is no scientific statistical significance evidence to what they are actually doing. Things like lookalike audiences, targeting based on interests and or channels, none of that has been proven to work better than the alternative, which is leaving everything untargeted to a point where you are giving the algorithm complete freedom to find the most relevant people based on the people who have already converted. 

So how does that actually work?

Let’s say that you use no targeting and you leave all your targeting options broad. There is one thing that advertising platforms will require you to select, and that’s your goal. What’s your advertising goal (leads, people adding to cart, purchases, etc.) Then, once that conversion event happens as a result of your ad, advertising platforms will see who is converting and it will start looking for similar people based on online behavior and online similarity. Back to the example of the game, think of the players, the computer was able to get one player from point A to B (from click to a conversion). What will happen next is that it will examine the path that the player took and it will look for similar audiences that are likely to complete that same event/conversion. It’s that simple. That’s why, any targeting that you use, any audience or lookalike that you use might harm performance as now you are telling the algorithm to focus on one type of audience or on a smaller audience group which might open you up to a more limited or more expensive audience. By not using any or a very minimal amount of restrictions, you are telling the algorithm,  “Here are my creatives, here are my goals, find the best and most cost-effective way to get there.”

Don’t Be Needy: Less is more.

Another common issue that I frequently see is that many advertisers tend to think that they can optimize the ads more effectively by pausing, resuming frequently altering bids, etc. 

Back to the example of the game, how do you think the computer will react in the game if there is an external factor that makes frequent changes that might impact its prediction calculation? An external factor that changes the desired outcome by increasing/decreasing the bids forcing the algorithm to recalculate again to try to hit the desired advertisements target. And of course, everybody does it. Sometimes, I am forced to do it as well. If you set your goals and budgets and you spend a lot of money with poor results and you lose money, then it’s only natural to go into the ads and make changes. I am not writing it to make you stop it, just writing so that you too, will understand the impact.

Ads are not meant to perform based on a daily basis. 

I get a lot of people who ask me why they are losing money today or two days in a row, where they set bids and budgets but the algorithm isn’t able to perform and deliver the results they want. They then are forced to lower spend and bids and then try to look for that time to increase back up. It’s important to note that computers in some cases are wrong too, and what happens the next day when the advertiser’s targets are wildly missed, the algorithm will readjust based on the recent data and will either spend less money the following day or will get better results. 

This happens because the algorithm is readjusting based on performance of the llast few days. And here is another thing to consider. One day is only 24 hours, that’s it. If someone is not directly searching for your product, then 24 hours means nothing as it relates to the bigger picture. In fact, in many events, the people optimize based on the same day and expect results based on the same day but don’t realize that some people don’t convert or purchase on the same day. I could be clicking on an ad at 9:00 PM and only convert the next day in the morning when my paycheck is deposited. It will stay less than 24 hours but will show up the next day. That’s why it’s important to allow enough time and understand that it’s a marathon, not a sprint. 

Another thing to consider is when you set a target ROAS or target cost per conversion, they are asking you for your target on a lifetime basis, which means that if you put at $15 target cost per result, it will aim to get you that target cost per result, but not every day. It doesn’t matter if one day your cost per result is $15 then another day it’s $5 and then another day it’s $25 right? What matters is the average, the amount that you are paying on a long-term basis. Understanding that you don’t need to evaluate performance on a daily basis in terms of ROAS or sales gives you more flexibility to focus on the running average and not freak out when the cost per acquisition is bad one day or really good on another. It opens up your campaigns to more freedom, which in return will allow you to generate the maximum amount of conversions and sales for the most effective cost.

In Summary

Advertising on major platforms has become increasingly simple over the years. Many advertisers foolishly overcomplicate their ad campaign in an attempt to make it more complex and effective. The social media marketing gurus pretend as if they’ve cracked the code by shilling baseless marketing strategies such as mid funneling and low funneling. What they are really doing is overcomplicating the advertising process which is programmed to be as simple and hands-off as possible.

Advertising platforms use algorithmic targeting. All that you need to do is tell the algorithm your objective, goals in terms of cost per result, and overall budget. From there, you feed the algorithm a bank of ad creatives and let the algorithm do the rest. By choosing your audience or lookalike, you are limiting the algorithm to a subset of potential customers. Similarly, pausing, resuming, and adjusting your bids disrupts the algorithm. It may be hard to admit, but the algorithm is smarter than you in most cases and will more effectively generate impressions if you give it space and time.

jason burlin

A seasoned marketer with more than a decade of experience in online paid advertising. Managed more than $150M in ad spend and worked with more than 500+ brands. He is known as the unconventional marketer.



related posts

Does Cost Per Impression (CPM) matter?

Read More - Reading Time: 9 minutes

Advertise in High Frequency

Read More - Reading Time: 7 minutes

Sign Up For Jason Burlin's blog, videos & all the latest updates.

Open chat
1
Need Help?
Hello! How can I Help Your Business?