Statistics, Forecasting, and Social Media

Welcome to The Social Lab! This blog will not only help you navigate your personal social media, but also investigate marketing trends and examples of great social media work. Do you have something we should talk about? Make sure to send us an email or contact us on social media.

Analytics is on the rise when it comes to social media. Companies everywhere are asking why the number of followers their business has means anything when it comes to true KPI's such as leads, sales, and subscriptions. So how can you make sure that your clients understand the importance of social media? Furthermore, how can you ensure that your personal brand is succeeding and creating the amount of revenue, brand awareness, and thought-leadership you desire?

Since social media analytics can get really confusing, it's important to understand the definition of the different KPIs that exist. For example, link clicks is pretty self explanatory, but the difference between reach and impressions is a tad more confusing. Reach is the number of individuals that received a post. On the other hand, impressions is the number of times a post was displayed. Basically, impressions will always be higher because it isn't the number of individuals, it is the number of times it was displayed. We all know from experience that you will many times see a post multiple times, not just once. There are many other KPIs that you will run into, so we recommend heading over here to study up on what all of them mean.

When it comes to us, in order to truly understand all of the metrics we have access too on all the different social media platforms, we turned to statistical theories. Using Google Sheets, we created multiple tabs with all of the metrics we can find from each platform.

The image above represents one of the tabs on our Google Sheets. For Twitter, we have a breakdown of the metrics you can find on Twitter Analytics and even export. This is just one snippet of the metrics we take into account, but the point is that for each month, you keep track of all of the analytics. With that in mind, you can then begin to measure how each metric changes over time. If your client cares about impressions, then you can see how you have improved that metric since working with their account. Once you have all the metrics you want in one place, you can begin to really analyze what all of them mean. For example, what if you want to give your client or your personal social media an overall grade?

The image above is for Facebook and represents the end of the Google Sheet. The teal and beige highlighted cells represent "weights". The reason the beige is different from the teal is that it represents Facebook Ads, rather than organic. Weighting is a beautiful way to create an overall grade while also taking into account the fact that some metrics are more important than others. When you were in high school or college, homework was not as impactful as final exams, our stats represent the same philosophy. So let's say on Facebook you are tracking 50 different metric categories. You can then weigh each of them based on importance and have them equal 100 in the end. Link clicks are more important than likes, so you can have it weigh more. After this is complete, you can then create a grade, or what we call an Index Score. How do you create this number?

=SUMPRODUCT(D10:Y10,Z10:AU10)/SUM(Z10:AU10)

In order to take into account multiple arrays, you have to compare the original metrics with the weights assigned, then you can generate an appropriate grade. It's important to keep in mind that the Index Score is an arbitrary number that is great for comparing within your own business or comparing different platforms, but it's hard to compare with competitors, simply because you don't have their metrics. Of course, the Index Score is just a number, so how can we analyze it even further to determine whether or not it's good? Just like with any statistics, it's important to have a lot of numbers in order to determine significance. We took the average of all of the Index Scores over time and then determined whether or not that particular month was better or worse than the average.

=(MINUS(AV8,AW8)/AW8)*100

If you know the percentage increase or decrease, you can then say that you were 15% better than what the average would be. You could even give yourself an A, B, C, D, or F.

This all may seem incredibly confusing and not worth your time, but just remember that how you represent your work to your clients will have a major impact on retention. If you can impress people with your ability to show the importance of social media, than they are more likely to keep you on board.

What other interesting things can you do with social media metrics? Besides creating an overall grade, you can also compare two different variables by creating a Pearson Study. Let's say you want to determine whether or not there was a significant correlation between the number of posts and Index Score. You could say whether or not the number of posts you make determines the success of your overall grade. You can also use linear regressions and forecasting. What if you want to know how many followers you will have at years end? Create a linear regression model with all of your stats, and see where it ends up at the end of the year.

Many of these concepts may be difficult for individuals without a background in mathematics, so if you would like assistance with any social media statistics...send us an email or comment below. How do you go about reporting on your personal brand or the social media for your clients? Let us know so we can all learn how to report better and improve as social media gurus.