Social media marketers can easily get so caught up in measurement and tracking that the impact of the information gleaned can get lost.
The data from analytics and metrics reports may lead to observations, but therein lies the problem: Observations are not actionable. We must strive for insights from data and recognize the difference between observations and insights.
Professor and author Luke Williams highlighted this distinction in an article in The Atlantic by noting, “observations are raw data, the gradual accumulation of information that you have consciously and carefully recorded—exactly the way you saw or heard it, with no interpretation. Insights are the sudden realizations in which you interpret the observations and discover patterns.”
In other words, the difference is the classic correlation versus causation—the distinction between knowing two variables are related and knowing whether one causes the other.
To ensure you’re moving past observations and turning your social media metrics into actionable insights, here are four keys to keep in mind.
1. Establish a measurement system
Insights aren’t obvious. You need a set system and reporting structure to unlock them.
I am a big proponent of a three-tiered metrics structure. The top two layers are made up of business metrics, which measure what social media marketing is contributing to business success, and marketing metrics, which measure general social media and channel-level performance.
The third layer of optimization metrics, though, is the most critical in the search for insights. Optimization metrics are those that provide insights on how to improve your social media marketing or indicate what needs to change. This section includes the rates, ratios and tests being conducted that can determine what is driving real results. By delving into the optimization-level metrics, social media marketers can assess what’s contributing to changes at the higher levels of metrics and begin to pinpoint the insights needed to make data-driven decisions that can iteratively enhance a marketing program.
For instance, you may see a decrease in the overall revenue generated from your social media marketing, but your audience is growing. Therefore, you are failing to meet a business goal, while achieving a marketing goal. Then, you may see a discrepancy in the optimization-level metrics of compound growth, audience retention and conversion rate, which may signal that despite many new followers being acquired, very few are being retained. However, those who are retained usually convert into customers.
Armed with this observation, you can identify a multitude of tests to run to find ways to increase retention and the conversion rates of those who do stay. Examples of these types of tests for retention could include comparing the performance of different messages, content types, and targeting; while conversion rate tests could include examining elements like call-to-action placement, the structure of individual posts, and image types.
Not only does such a structure allow social media marketers to pinpoint where changes are happening, but it also allows for an understanding of how those changes influence larger metrics, what type of larger scale impact they have, and limits the chances of overvaluing “vanity” social media metrics.
2. Start with questions and test hypotheses
Before they can find insights in data, social media marketers first need to know what they’re trying to find.
To do this, start with questions about elements that you can or are willing to change, such as, “if we do X, will it improve outcome Y?” From these questions, apply the scientific method to test hypotheses, such as, “if we change X, it will lead to outcome Y over Z period” to establish and guide tests.
When considering what to test, try to challenge big ideas and assumptions with an eye toward improving your audience or customer experience and providing more value for them.
For instance, an organization may be considering whether it should start using Facebook Live, despite never having previously done something similar. In this situation, a hypothesis could be, “if we use Facebook Live twice a month over the first quarter, we believe it will increase our social reach and the size of our Facebook audience by 10 percent month to month.” In this case, a hypothesis that includes a time frame and clear measures for success are defined to set a framework for the test, from which the organization can clearly weigh whether or not to proceed with Facebook Live after the first quarter. This hypothesis also sets the foundation for further testing in the second if the original one ends successfully.
That being said, tests should only focus on elements that can be changed and that fit within an organization’s appetite for change—otherwise the effort can be a waste of time, energy, and resources. And worse yet, the insights found would not be actionable.
Another trap to avoid is testing to simply improve a particular metric without context or an understanding of its overall relationship to larger goals and audience experiences. Keep in mind that numbers are representations of real-life actions, experiences and feelings. The end-goal is—and always will be—creating more value for audiences.
By developing hypotheses and testing them, social media marketers can measure the results, the impact on metrics, and come away with implementable insights.
3. Establish a cadence of testing and incorporate it regularly
When testing and measurement systems are ineffective, it’s often because testing is not being done consistently.
What “consistent” means can vary from organization to organization depending on capacity and organizational needs, but a defined cadence of testing needs to be in place and adhered to. While ad hoc reporting and testing can be beneficial and necessary, relying solely on this type of tracking does not provide the process and structure needed to prioritize and standardize measurement. In order to assess performance accurately, consistent metrics must exist along with organizational commitment to reporting, analysis and constant optimization.
Additionally, the ideal optimization and refinement of a marketing program is predicated on iterative and incremental changes that build upon themselves, which require consistent reporting, reviewing, testing and new ideas.
4. Continually ask why
Often, social media professionals believe they know what has led to a change represented in metrics or mistake observations for insights because of their own personal experience, behaviors or preferences. However, these preferences are nothing more than guesses or assumptions that need to be established as hypotheses and tested to drive substantiated insights.
The principle of “five whys”—asking why something happened five times, digging deeper each time to get to the root of the issue—can be helpful when trying to set what to test. It can be used as a means to get to the root of what should be challenged or, conversely, what caused a change and should be confirmed through testing.
As previously mentioned, actionable insights do not come from the most obvious places and the best social media marketers continually dig and ask “why” to find them.