Viralstyle Blog

How to test Facebook Ads
28 Feb 2017

Testing Facebook Ads, Explained

/
Posted By
/
Comments0

It doesn’t matter if you are selling cars, software, or custom apparel on Viralstyle – the methodology to testing ads is essentially the same. There are many facets to testing; identifying the targeting criteria, deciding how much to spend, or knowing when to scale or kill ads, it can be hard to make a decision with so many moving parts. As Luke Stasi explains in our 102 Academy video tutorials, it’s like learning how to drive a manual transmission. It requires a lot of experience and gained intuition for you to make confident decisions early in the marketing process.

Why is Testing So Difficult?

Nobody said digital advertising was easy. It requires analytical abilities and a strong understanding of the data and individual characteristics of your niche. If you are unfamiliar with the behavior of the users you are targeting, you may not even know where to begin.

Not to mention, ads during the testing phase probably aren’t generating positive ROI yet. This can toy with your emotions as a marketer or entrepreneur so it’s important to stay patient and be objective with your data.

How to test Facebook ads

In this blog I outline 4 phases to effectively test ad campaigns on Facebook and navigate your way to positive ROI.

Phase 1: Research and Planning 

The first step to an effective testing strategy is putting in the research needed to understand your niche. Knowing what to advertise is just as important as who you will be advertising to. You should not only be looking for targeting options but also characteristics of your niche. What they are passionate about? What do they find funny or amusing? What do they spend their money on? Once you have identified what you would like to advertise to them, you should begin to compile various ways of targeting potential customers. Targeting options to look for:

  • Broad interests
  • Other pages, stores and brands
  • Blogs, trade groups, magazines, forums and events
  • Public figures and celebrities
  • Behaviors and spending habits
  • Occupations, schools, or fields of study
  • Demographics and geography

Once you have compiled a list of targeting options, categorize them and begin to analyze them with the audience insights tool in Facebook or other various research tools. In doing so you should be learning characteristics such as demographics, audience size and the amount of crossover between targeting options.

Phase 2: Audience Selection 

Now that we have a pool of targeting criteria to choose from, we can start to build our ads. But how do we know which targeting options we found are the best ones? If you are going after a new niche this can be a tough decision to make. There two general approaches to selecting your test audiences. You should look to use a combination of both methods in a way that makes sense for your objective.

Method 1: Standard Audience Selection

With standard audience selection we build ad sets by selecting a wide range of targeting criteria split into individual ad sets. For example you could create 4 ad sets total; one that targets a blog or magazine, another with a broad interest, the third a public figure, and the last ad set targeting an online store that sells similar products to your niche. This strategy will allow you to target users who are involved in every area of your niche. If you find success in one area, then you can scale out your campaign to include more targeting that is similar to what is successful.

Method 2: Inner/Outer Onion Audience Selection

The Inner/Outer Onion audience selection method is a way of concentrating on your targeting based on their level of involvement or interest in your niche. This is typically achieved with flex targeting, allowing you to narrow interests or combine users who have multiple interests. For example, if your niche is cats you may build 3 ad sets that represent different levels of involvement with cats. The outermost selection of the onion is the ‘cats’ broad interest on Facebook, followed by the behavior selection of cat owners, and lastly as the 3rd most inner section of the onion you could pick a flex audience of Facebook users who are cat owners that also like a Facebook page about cat memes.

Of course there is no exact science to picking your initial test audiences and most marketers use a combination of both of the methods described above. We suggest starting out with a minimum of 3 ad sets, but in some instances you may have to try out over 10 audiences until you find something that works. Once you are more experienced with a niche you will learn which targeting options work best for your own objectives and how much testing is required.

Phase 3: Choosing a Budget 

The next phase of building an effective test is setting your budget. Setting a budget that makes sense depends greatly on your individual spending spending power and overall goals. You should also consider your advertising objective and the amount of time and money required to allow Facebook’s delivery engine to optimize. Other factors to be aware of when testing is the audience size and cost per thousand people reached. Some targeting criteria is more competitive, or may require a larger reach for the engine to learn about the users in order to optimize delivery. There are two general approaches to setting your initial test budget:

Method 1: Standard Test Budget Strategy

With standard budgeting strategy, we will decide our budget based primarily on the minimum amount needed to achieve our objective. For Website Conversion ads selling a t-shirt where our profit is $12, setting a daily budget of $10 will allow us to determine if this is an effective ad with positive ROI. If you are selling a big ticket item like a mattress, you should be willing to spend more to discover which targeting converts best. Engagement and Clicks to Website ads cost much less per action and these actions happen more frequently so ads with these objectives usually don’t require more than $5/day to get a feel for their true costs.

Method 2: Slow Burn Test Budget Strategy

The Slow Burn test strategy attempts to set daily budgets below what is required to achieve your objective. The idea behind this is that by setting a lower budget over a longer period of time, it forces the engine to learn more slowly allowing it to pick up on more information about the audience. In theory, the lower budget will force the ad’s delivery to be more selective about when and who sees it. There are a number of reasons why you would want to try this. Slow Burn can be a useful strategy if you are unfamiliar with the niche or if your niche has countless targeting options and you are having trouble narrowing it down. The slow burn method can allow you to be more conservative with your spending so that you can test more audiences. The downside to this method is that with larger audiences, lower spending means a smaller reach. If you aren’t reaching a large enough portion of the audience, the engine may not have enough data to learn how to optimize.

Phase 4: Kill or Scale?

Our goal in this phase is to trim out the ad sets that are losing money and increase spending on the ad sets that are generating positive ROI. This is possibly the most difficult decision to make when you are starting out. The budget required to test an idea depends on a lot of factors. The most important question you should ask yourself is how much are you willing to spend on the idea? How much potential do you see in it? Is there anything in your current data that shows promise?

Overtime you will develop your own methodology to systemize your testing in a way that makes sense for you. You will need to be familiar with your data to get an understanding of each ad set. Ultimately, your Cost per Sale is the most important metric but you also need to be aware of other pieces of data like conversion rates compared to cost per click, cost per thousand, profit margin and more to understand how each ad set is performing. Having a high cost per click or a high cost per thousand is okay if your conversion rates are also high. Or, you may have a really low conversion rate but your cost per click is so low that you can still generate positive ROI.

For most t-shirt campaigns we recommend allowing your ads to run for at least 2 days and you shouldn’t need to spend more than $50 overall to determine if an idea is viable. This is especially true when you are first starting out or attempting to hit a new niche. If you don’t have conversion events or purchases with $50, it’s likely not the home run you were hoping for and your time and energy at this point is best used trying a new idea. With more experience and cash flow you will start to find more confidence in your spending and know when to make adjustments or keep pushing campaigns that aren’t making a profit yet.

Now that you have made it this far, hopefully you have found a few winning ad sets that you can start to scale. Check out our case study about scaling Facebook ads for in-depth strategy and results!

Want to learn more advertising strategy for your custom apparel campaigns? Enroll in the Viralstyle Academy today.

Learn ho to sell t-shirts online

Comments

comments