Testing and ad ideation are the keys to unlocking the full growth potential of your brand. However, testing the wrong way can be just as bad as not testing at all! In this article, our goal is to teach you how to plan and conduct a simple ad test with Preflect that can revolutionize the way you run Facebook ads.

The first step to testing Facebook ads is to set realistic goals and expectations. What are you trying to achieve by conducting a test? Are you trying to learn about your customers? Are you trying to make more sales? Are you trying to figure out which products people think look better?

Whatever your goal is, you need to have one and establish it before you begin your ad ideation because this will be your guiding light throughout the entire process. It shapes your decision making and allows you to disregard turbulence along the way that is irrelevant to your goal.

For example, if your goal is to earn more sales, then all you should be focusing on is ROAS and the metrics that go directly into the ROAS equation. If your goal was to sell more products in a certain collection, then your goal isn’t just ROAS, it’s also the sales volume of select products.

This is going to change the types of products that you put into your ads, and the landing pages you send traffic to, and it’s going to shift your view on the conversions that you’re receiving. What we’re trying to say is that your goals should help you put your tests into perspective!

The first step to planning your ad test is deciding on a budget. Our recommendation is to test 4 - 8 ads at $100/day in advertisement spend. As your daily budget increases, you will need fewer ads per dollar of ad spend, but it varies from merchant to merchant. In Preflect, there’s a notification at the top of our ads page that tells you exactly how many ads are optimal to run for your specific daily budget.

Plan on running your ads for 3 weeks uninterrupted. Only make changes after 7 full days of running if your ROAS is below 1x, your CTR scores are low, and the ads are recommended to be replaced.

Once you know how many ads you should be running, move on to ad ideation and start creating!

A solid ad ideation process helps you avoid throwing “spaghetti at the wall.” If you aren’t familiar with that phrase, it means trying just about anything until you find something that works. At Preflect, that’s just not how we do things and neither should you!

USPs

The easiest way to come up with a variety of ad concepts is to first identify your brand's USPs. What’s a USP? It stands for Unique Selling Proposition. It is a feature or characteristic of your brand or product that sets you apart from your competitors. It is what makes you, your brand, your products, or your mission unique.

USPs are explained in greater detail in our article about writing good ad text, so if you’d like to pause right now and go read that section, click here!

One thing that’s important to remember is that when listing USPs for the purposes of ad ideation, they don’t need to be extremely strong or prominent. They can be as simple as the fact that your product is premium or affordable or for women or for people who love the ocean! The point of listing them out is to have a pool of USPs to pick from when you start testing.

Once you have a list of at least three USPs, you have enough ideas to now test three ads! For each USP create at least one ad. Write unique primary text and headlines for each ad that really highlight that specific USP. They should be three uniquely distinct ads that highlight very different aspects of your product or brand. If it takes you a while to come up with three, that’s ok! There’s no rush. Having at least three is great for the testing process and it’s worth thinking about for a while.

It’s best to start with at least one ad per USP with a minimum of three ads. For example, if you had three USPs, you could test three ads, six ads, nine, etc. So long as there is always an even number of ads per USP. Or you could test five USPs and run five ads, one ad per USP. Just keep the number of ads per USP even!

Also, make sure to run the same number of primary text variants and headlines for each ad to make sure that each one has an equal opportunity to succeed. Even though the ad text will not be similar, ads that have more text variants will likely outperform ads that have fewer variants because it is able to optimize for the best one.

As stated in the planning section, plan on running ads for three weeks during initial testing. After week one, there may appear to be a clear winner, however sometimes Facebook’s AI overserves certain ads early on because their AI likes the ad, not because it is truly the best performing. Running ads for three weeks allows enough data to accumulate to get a clear picture of how each ad performed.

The most important thing that you must remember during your testing process is: do not touch anything. We cannot stress this point enough. If you’re committed to a test for a certain period of time, any adjustments that you make to your ads or ad set will disrupt the process and invalidate the results of the test. To avoid this, just take your time and launch your ads when you’re positive that you’re ready! Again, if no ads are performing well and your ROAS is below 1x, then go ahead and do a creative refresh after at least 7 days of running those ads.

So, you’ve run your ad test for three weeks and now you have some results! So what numbers should you look at specifically? Assuming that your goal is to earn more sales, you should be looking at ROAS and to a lesser extent, click-through rate.

Click Through Rate (CTR) is the percentage of people who saw the ad that clicked on it. To read all about what it is, how it’s calculated, and much more, click here.

The reason you must consider CTR is because a high ROAS ad with a CTR of 2% or lower is not scalable. This means that the ad may have a high ROAS with a low budget, but if you were to scale (increase) your budget, the ad would no longer have a high ROAS.

So, this means that the winner of every test is going to be the ad that yields the highest ROAS with the highest CTR.

Your winning ad tells you which Unique Selling Propositions resonate the best with your target audience from an advertising perspective. The reason why you have to test this in the first place is because there’s more than one way to sell your product. What makes your customers want to buy from you might be something totally unexpected about your brand (or it might be exactly what you expected!). Either way, the only way to know for sure is to test it.

Once you have a winning USP, you now have a great starting point for taking action on the results of your test. With this new knowledge of what your customers liked the best, you can now create a larger quantity of ads centered around this new, winning USP.

Sometimes this can happen. If you test three USPs and all ads have similar metrics that are all great (i.e. CTR greater than 2% and 2x+ ROAS) then congratulations, you’ve found three winning USPs. Now, if you have three USPs that are all performing poorly, you need to test more ads, more USPs, or both. You need to test until you have clearly defined winners and losers for your USPs so that you have confidence in your decision to move forward with some and not with others.

This guide to testing is a very basic one and is meant to help you start your first ad test or encourage you to run one if you haven’t in a while. For merchants that have multiple winning USPs and need to start testing ads and ad copy within those USPs, stay tuned for another article where we dive into advanced optimization strategies including advanced testing. Always feel free to message us in support if any of this information isn’t clear!