Use A/B Testing to Inform Advertising on a New Social Channel

You’ve had some success using advertising on various social media channels and are looking to expand your reach. How do you determine where to invest in advertising on a new social channel?

This step-by-step guide will help you pursue new social media advertising on new platforms with confidence. With clear, proven approaches that can assist in the creation of effective campaigns, this guide will give you the tools to make smart, productive and profitable social media decisions to build your brand.

Know the Platform, Know the Audience

There are many options available today for social media advertising. Facebook, YouTube, Instagram, Twitter, Snapchat, Pinterest, and TikTok are among the best-known, most-used and most popular platforms. Each platform has a distinct audience and functionality. Research each of the platforms to understand more details about their demographics and popularity.

The Pew Research Center publishes an extensive list of platform demographics, including who uses social media, who uses each platform, the growth of platforms historically and how frequently Americans use each. HubSpot has a detailed list of insights on each platform’s demographics, including usage frequency, age, gender, and geography.

When you select the platforms you want to test, be sure to research what the best practices are for each in developing social ads. Most platforms have guides that can help you develop ads that are effective. In addition, there are several guides available that break down best practices for each major platform.

Before you begin testing paid social on new channels, you need a clear understanding of the audience you’ll be targeting. Most platforms have powerful tools that help you in segmenting the audience. Consider the demographics you want to test around. In addition to personal demographics, you may also be able to test against the device being used, interests or online patterns or behaviors.

Know the Goals

New social media advertising can have many different intentions. Some campaigns are designed to raise brand awareness. Others are looking to boost engagement, whether through app downloads or website visits. Some campaigns are looking to generate sales leads while others are intent on pushing consumers directly to a sales website.

For testing purposes, it’s best to identify a single goal for the campaign for ROI and a single goal for your brand. With one well-defined goal for each, you can better determine the variables, measures, and timeframe for your testing.

For example, if you’re looking to push app downloads, you can define the copy that is appropriate to that goal. You can identify the target audience, using tools embedded in each platform. And you can develop a timeframe – one week, two weeks, several months – that makes sense. You can also determine the right metrics upon which you measure your success.

Use Technology

There are multiple tools available to help you manage your social ad campaigns. Many of these tools also include testing tools that allow you to build campaigns that will test the efficacy of your platforms, including A/B testing (see below).

Given the size of social media campaigns, technology is a must. These tools allow you to aggregate, track and measure results with pre-defined and customized reporting functions that allow you to track progress and analyze results.

Run an A/B Test

A/B testing is a scientific approach to measuring the impact of your marketing work. It involves changing variables in two versions of a campaign to assess which has the best impact on your identified goal.

To begin, you split your target audience into two groups. This segmentation should be randomized to ensure there is no bias built into the test groups. Each of the two segments is given one of the two versions of the ad. One of the groups is considered the control group.

Based on the measures you’ve determined; you then assess the results.

The key to success in A/B testing is to be sure that you are only changing one variable per test. If, for example, you change the copy used in the ad, and a picture in the ad, you will not know which variable affected the results. To test multiple iterations of ideas and concepts, you’ll want to develop and launch multiple tests.

Benefits of A/B Testing

There are many benefits to conducting A/B testing for social paid advertising. Doing your own A/B tests lets you play out, in real time with quantifiable data, different ideas or concepts. Among the benefits:

• Understanding Audience Preferences. Changing the right variable can determine which tagline, call to action or sentiment your audience prefers.
• Determining Segment Differences. Do some messages work better for certain audience segments? A/B testing, repeated over time, can help you pinpoint what works for one audience segment compared to others.
• Identifying Platform Differences. While you are not always comparing apples to apples, A/B testing can help you understand differences in resonant messaging between platforms. For example, a message to 18- to 34-year-olds on Facebook may play differently on Twitter.

The key to A/B testing is to continue to test and retest, refining the messaging accordingly. The more you test, the more clarity you will gain on what works and what needs to be scrapped.

What Can be Variable in an A/B Test?

As noted above, you want to limit the variables to one per A/B test. That said, nearly every element of your social paid ads can be the variable within your test.

• Copy. The words, number of characters, punctuation, and font can all be varied during the test. You can also think about the style and tone used. Do you want to keep it whimsical? Serious? Do you want to include a question or a statement of fact or statistic?
• Visuals. Do you have different images or illustrations you want to assess? Does your ad include an emoji? Research suggests that ads with a video or image work better than those that do not. However, that is no the case for all ads and all audiences. You can test text-only ads versus those with visuals, a static image versus a GIF, different photos, and different video lengths.
• Preview Content. You can edit the ad’s headline within a linked headline to determine what resonates.
• Call to Action. How are you asking the reader of the ad to act? Different calls to action can elicit different responses from your readers. It’s a critical part of any paid social ad and it’s important to get it right.
• Ad Type. Certain ad types can work better on different platforms and for different audiences. Does it make sense to have a larger display ad or a smaller, targeted ad on LinkedIn? Are carousel ads the best for new products?
• Hashtags. Do hashtags work for your audience? Should you run one or multiple hashtags? Are there audiences that respond better to hashtags? Experimenting with hashtags can make sure you’re using them in the right places.

Learning from the Results and Repeating the Process

What did the A/B test tell you? Analyzing the results of an A/B test is essential and requires some basic use of statistical analysis principals.

Determining a winner is not always about just which ad got the most clickthroughs. First you need to determine if the test was valid.

Is the test statistically significant? Statistical significance is a determination of whether the results can be attributed to chance alone. There are various programs that can calculate statistical significance. Marketers typically look for a 90 percent significance.

You can then break down each sample more closely – by the personal demographics or other factors you used to develop the test. Looking at these subgroups can help further clarify if the results are meaningful. These observations can also lead to additional A/B testing ideas.

With A/B testing, it’s important to continue to develop and repeat the tests. Each new test allows you to delve into more detail, test different hypotheses and understand a key market segment target.

Assessing ROI Across Old and New Social Channels

With commitments to any marketing strategy, brands want to determine the ROI. The simplest formula to use is to take value minus investment divided by investment times 100. This calculation is a standard measure of social. Applying that calculation to each social media platform helps you determine which investments should be made in new social spaces. If your ROI is above 0, your business is making money off the social media investment.

The social media costs to factor into the calculation include tools and platforms used to manage social media, the budget for ad spends, the creative costs to make the content, ongoing costs for salaries and agency fees.

Value can be interpreted in different ways. It could be revenue directly attributed to the campaign. It can also be a quantified valuation of impressions, conversions, reach or other outcome measures.

Tracing these results back to your goals and comparatively across platforms helps make a broader assessment of your ad spend.

The landscape for social media is changing rapidly. As platforms continue to grow, either by increased usage or acquisitions, advertising may not be optional. To be present, brands will continue to need to evaluate how to make the most of their social ad spend budget. It’s smart for brands to invest in growth opportunities, even early on, to be active in channels even as they are emerging.