In Facebook’s Q2 earnings report last year, CEO Mark Zuckerberg said that the average US consumer spends 40 minutes on Facebook per day – and much of that time is spent on a mobile device. When it comes to mobile advertising, ad targeting is critical and Facebook is one company that does this very well.
Facebook gathers a treasure trove of information about each of its active users – nearly a billion every day. It also continues to collect more data and information on a daily basis, giving Facebook the ability to offer its advertisers information on just about any target audience on mobile or desktop.
With this many eyeballs and refined data on how to get in front of them, marketers can most certainly experience positive ROI from Facebook advertising. The caveat is that they must use Facebook correctly if they want to see success with their campaigns.
With that in mind, here are eight things not to do in your Facebook advertising:
1. Testing images, headlines, and audiences all at once
When advertising on Facebook, you’ll find that you will have an image that performs best, a headline that performs best, and an audience sector that is most receptive to your offering. If you create an ad with ten different images and three different headlines, that will produce 30 different variables. To get enough data to test all of those variables at once will require a HUGE budget. If you have a huge budget and want to do that, great. However, we’re coming at it from an ROI perspective.
Therefore, a higher value approach is to select an audience that you think would resonate with your product or service offering (e.g. new moms). From there, test the images first with that audience by rotating through them and see which image is garnering the most response. Look at conversions and click-through rate (CTR).
The next step would be to test the winning image alongside your three different headlines to evaluate which headline performs the best. Then, when it comes time to expand and scale your campaign, you’ll want to test your winning image and winning headline with additional variations of your top level target audience (e.g. new mom’s that like Dr. Seuss; new moms that live in Colorado, etc.).
2. Relying on too few audiences to carry the weight of the campaign
Say you’ve had success marketing to new moms living in Colorado. If you continue marketing to just that audience, the reality is that they’ll likely grow tired of your offer, your brand, etc. and will stop being responsive. What’s more is that within that audience, there is a limited number of users likely to be engaged. Once you’ve reached them all a few times, your performance is likely to fall.
For example, if your audience is “new moms in Colorado,” Facebook will target users that are most likely to convert for the ad, thus prioritizing those engaged new moms living in Colorado. This is problematic because you’ll eventually run through your “optimized” moms and then Facebook will start targeting other users within that audience who will be less likely to convert.
By continually testing new audiences and selecting more to target (e.g. city moms, green moms, moms in Indiana, etc.), you’re spreading your budget out over a larger area and targeting more of the optimized moms in each audience.
3. Broad messaging to audiences
Advertisers will often will set up a campaign to test the performance of different audiences, but they’ll use the same messaging. For example, if they have a product that would appeal to both moms with newborns all the way up to moms with 5 years-olds, they’ll use messaging, such as “This is a product your child will love.”
To get more useful data, post your ad with imagery and messaging that’s relevant to the audience. If you’re trying to test an ad targeting new moms, show them an image with a newborn with messaging along the lines of, “This is a product your newborn will love.” Cater your content to your audience.
4. Running the same old ad
Running the same ad over and over will lead to reduced engagement, higher click costs, and ultimately the failure of your campaign. Instead, continuously rotate your ads and test new images and messaging.
5. Relying completely on Facebook’s Behavioral or Demographic targeting
Facebook makes it possible to conduct very defined targeting. For instance, you could select “people likely to buy luxury home goods in Boston, MA.” While that’s great, it’s best to not completely rely on those defined targets due to increased competition.
Instead, layer in additional interests (which will come from your own research). In the case of luxury home goods, for example, you might layer in “people who like interior design,” or “women who like restoration hardware.”
6. Setting and forgetting
All too often an advertiser will find success with a campaign and then continue to let it run as-is. Taking this approach tends to result in higher costs and a lower performing campaign in the long run. For better ROI, use data to refine and optimize.
In the early stages of your campaign, all sorts of data starts coming in (e.g. 18 year-olds are converting better than 35 year-olds; males are converting better than females; mobile is converting better than desktop, etc.). It’s critical to use that information to tweak your campaign. This could include changing the targeting for placement (mobile vs desktop), refining your age targeting, and removing ads that are not performing as well.
7. Not testing your bidding methods
Relying on the most common bidding methods and not testing different variables is definitely not a smart Facebook advertising strategy. Doing so will likely result in you missing out on lower cost conversions. Instead, bid for clicks, conversions, and impressions and test each one.
8. Optimizing for too aggressive of a goal
Let’s say you have a product that sells for $2,000 and you launch your Facebook ad with the goal of getting your target audience to buy that $2,000 product. The issue is that Facebook (and marketers) leverages your conversion data (people that have bought the product) to optimize your campaign. With such a high price-point, you’re unlikely to get a large number of people purchasing your product, which means both you and Facebook will have limited data in which to optimize your campaign and know whether or not it’s successful.
Instead, move your conversion goal up the funnel. For example, in addition to tracking how many people are purchasing your $2,000 product from your Facebook ad, track people who added that item to their cart and start by optimizing to that goal. This will provide more data that will allow you to optimize your marketing, including creating new Facebook or email campaigns that retarget people who added the item to their cart with different messaging.