Facebook Ads Launch Strategy
You started your store from scratch and now let’s see how you are going to marketing it on Facebook. I will introduce you to some general steps you need to take that are proven to work well. First, there are some terms you should understand before we start.
These are the regular campaigns that are used for testing different creatives/variations. You definitely want to test everything through these campaigns first and then give all in, let’s say, “real campaigns” in the future.
This is for you to collect data and don’t worry about optimization – to generate your audience and create the best ad that will help you to profit as much as possible. And you do all of this without losing a lot of money on “bad” marketing that won’t work for you.
After you collect all the valuable data, you are ready to move the PROVEN audience(s) and creative(s) to scaling campaigns. That’s how you separate Sandbox and Scaling campaigns and save money.
We are now going to go step by step to understand this diagram above:
Initial Creative Test
This first step is optional, but you will get clear results and definitely be more prepared for Initial Data Collection. So, it’s optional but recommended. We want to do this for three reasons:
- To know what’s the best creative with the most clicks
- To collect data
- To have engagement and social proof on our ads
When you skip this test, you can just create one ad and jump to Initial Data Collection.
Even if I wrote WC(VC) or PPE for the campaign objective, I definitely recommend going with WC(VC). Practice showed that PPE can give you a lot of junk data because even if people engage with the post, they sometimes don’t bother loading the page (if on the slow network connection, for example).
When it comes to budget, if you’re testing three ad sets (but you can test as many as you like), you should have a budget of $30-60. After you create the first ad set, duplicate it to ad set 2 and ad set 3 and just change the ads inside them. You don’t need to have a large budget if you think these will work better. You want to collect data and save the budget for later.
As for the targeting, choose TOP 4 eCom countries – United States, United Kingdom, Canada, and Australia.
For everything others, you can follow the diagram. Create different ads for each ad set, but with the same headline and description. You want to change only one thing per ad, otherwise, you wouldn’t know what caused the different results.
Important tip: Set the automation rule for every ad set to pause once it reaches 1,000 impressions. Only then, you can compare the data.
Initial Data Collection
You want to grab that winning ad (only one) from the first test and now test 10 different single interests by using the audience sandbox campaign.
Also, you want to test 95% video viewers’ LLAs as soon as you collect some data to do so. In other words, you will need two different campaigns in this phase.
This graph shows the first campaign you are going to need. As I said, we need to test 10 different interests, therefore we create 10 different ad sets with THE SAME winning creative. If you have two, then it’s ok to have two ads, also. Target countries remain the same.
Remember to put half of your budget per ad set – if your total budget is $100, each ad set should have a budget of $5.
CPP should be your main metric. Also, you need to closely monitor all of your ad sets. If one of them isn’t working, for example, PUR = $0 and your ad spend is higher than the breakeven of CPP, you should definitely pause this ad set. Something isn’t quite working and in this case, we know that’s probably interest. But have in mind that this is not the rule and you can pause the ad set too soon, not to mention that the audience could be the right one, after all. So, again, it’s up to you.
This is the second campaign in this phase, and you want to launch it the moment you have over 2,000 video viewers on your best creative. Create 95% VV LLAs (1%, 1-2%, 2-3%, 3-4%, and 4-5%) and start these ad sets in the new campaign like in the previous diagram.
When it comes to this campaign, you don’t have to optimize a lot. Facebook will need some data to work with and that will happen automatically in 2-3 days. You should only pause the campaign if it spends over 2x breakeven CPC with 0 purchases.
If you have tested more than three good creatives and haven’t reached to any profitable audience yet, the problem might be your offer. If nothing else works, try switching the product and starting over. Don’t spend more money on testing the audience. Only when profitable, when there are no problems to be fixed, go to the next step.
After you have your audience, there is more data for LLAs, right? This is the place where you really focus on your target market and stop losing money. You will either come to the beginning or even start to gain some profit.
Narrow down your audience by age/gender/country and launch new interests similar to top ones. If you tested 10 interests in the previous phase and two of them did great, launch two new interests similar to them. For these eight that didn’t work, try going back and replace them with eight new interests inside your sandbox campaign. In another case, if those two give you the option to narrow your audience even more, go for it.
Your goal is to find more than five consistently profitable interest audiences. Once you get to this point, you are ready for interest scaling.
Furthermore, for new campaigns with LLAs, I recommend you have >500 events on these previous campaigns. So, if your objective was “add to cart”, you want to have more than 500 ATCs to launch new LLAs campaign. As you can see, there are a lot better LLAs than 95% VV, so combine them as you wish. The campaign setup should be the same as shown in the “Initial Data Collection” phase.
As I said, this one is optional if you are getting profit in the previous phase. But, here you can really improve your ROI. When you find what works best with your audience, what messaging or what creative gives the highest CTR, that’s when you should go for it and your ad cost in the auction should be lower.
And if something isn’t working, that’s definitely not the audience. The problem could be your ad, so try and test different variables in order for it to work properly.
This doesn’t necessarily come at the end, you can start testing your creatives right away. Test as much as possible to get your winning campaign!