Even with the boom of social media, where testing has become increasingly frictionless, creative testing has actually become more complex. The main issue marketers now face is knowing how to introduce data-driven creative in their social media accounts, which increasingly require broad machine-learning powered structures. In this article, we’ll look at some problems and some best practices related to creative testing in social media.
The most common problems with social media creative testing
Facebook Broad Targeting
Using Facebook, you can target a broad or a specific audience. Broad targeting allows you to tap into the most targeted users of Facebook for your brand. Ironic, right?
It also is an essential ingredient in minimizing platform volatility and supporting your efficiency as you scale. Unfortunately, if you like to be in control of your testing, broad targeting forces you to become dependent on Facebook’s algorithms. For most businesses, it also means a blackbox is created around one of the key tools of building an increasingly future-proofed strategy: insights.
You can still extract data from broad targeting so long as you apply rigor to your creative strategy inputs and your approach to tagging and tracking. The result is the best of both worlds where you can combine brand-level creative with increasingly refined creative inputs that are based on iterations increasingly honed into specific user segments.
Missing your target audience
If you aren’t using broad targeting, chances are your Facebook ad campaign might be failing, most likely because you are actually missing out on your target audience. The secret? Precise targeting on Facebook isn’t actually always very precise – especially compared to the platform’s machine learning driven algorithms. One of two different factors generally contribute to this.
Making your targeting too narrow on the social media network
This tends to happen when you are too specific with Facebook’s targeting features. Using too many filters will split your target audiences too much and ultimately result in low-quality users seeing your ad.
Making your targeting too broad
…without being increasing targeted in your actual creative assets you can go as broad as you are refined within your overall creative strategy and how you cascade that strategy into measurable tagging and tracking setups.
Once you know your winning creative combinations, you have to determine which type of ad formats and variations will perform best and support your efforts in scaling volume. The trick is to ensure your iteration process is driven by producing new variations from already-validated concepts. This stacks the deck in your favor. You can further validate a winning creative strategy by starting with lower-resource format types, like static Single Image Ads, as you begin to iterate. As you scale volume and have an increasingly validated strategy behind your ads you can invest deeper into high-production creative formats like animations, videos or User-Generated Content.
Best practices for successful social media creative testing
You want to see direct results from your marketing campaigns and tie your advertising costs back to those efforts. Sounds simple, right? Creative best practices should always focus on the importance of producing ads and copy that maximize the interest and engagement of a brand’s target audience based on where they are in the buyer’s journey. Full-funnel best practices will help you track specific actions and target individuals with the right type of creative at the right time.
Test a lot of creatives
Many components go into an ad, such as your logo, design, brand colors, images, text length, font style and size, and call-to-action button color. These are simply tactical choices, not even getting into the limitless options on how you can articulate a brand’s value to a given user segment. You have to accept the uncertainty and invest in your backend growth operations: the only way to maximize your outcomes is by being able to continuously produce and launch data-driven creative cycles each and every month. If you don’t prioritize this output you end up playing a game of roulette with your paid media.
Be scientific with your account structure
An optimal account structure is your foundation for being able to get maximum impact from all of the work you’ll be doing. This is one area where following the rules helps. Caveat: the rules change often so you need diligence in being always up to date on what best practice is month over month, adapting your account structures each step of the way. For example: Facebook’s system requires approximately 50 conversions per week at a minimum to have its machine learning algorithms be effective in delivering the right ad to the right people at the right time. If Facebook is predicting you’ll get less than 50 conversions, it will already begin to penalize you – one more reason why a broad targeting strategy is a winning strategy. The problem might be that your target isn’t relevant enough.
Testing for efficiency vs iterating for scale
This is a big one. Business growth is not based on just your topline volume or bottomline efficiency – it’s both, even if in a given period of time you are focused on one. Your ultimate growth journey requires your approach to marketing to be prepared to juggle both at different points in time.
Our recommendation for most businesses: split your total budget into 80% going towards Creative Iteration (new variations of concepts that already have been validated) and 20% going towards Creative Testing (new concepts which may not work, but when they do represent the breakthrough ways to improve your overall performance). It’s counterintuitive, but despite this split, the most resource-intensive and strategically challenging aspect is the development of your creative testing plans.
What’s coming next
In the next blog posts, we’ll focus on the social media platforms that Ladder’s experts work on regularly with our clients, like Facebook and Instagram. We’ll also take a look at how to structure your account on those platforms for creative testing.