Social Media Data Analysis: How to Build your Plan

Social Media Data Analysis: How to Build your Plan

The Devumi blog is chock full of ideas to help move your social media marketing forward. If you’ve been reading regularly you’ll know a great deal about getting more Twitter followers, increasing your YouTube views, crafting response to negative comments, and on and on.

The most successful social media business accounts will be massive, sprawling communities bursting with users, comments, and information. How you can make a bit of sense of all this is through social media data analysis.

Without good social media data analysis you may find your stumbling in the general direction of forward, rather than smoothly moving upwards with your follower numbers, mentions, and general brand reach.

Here’s how to start your social media data analysis campaign…

Create a spreadsheet

You can do this in Excel, Google Docs, or anywhere else you like to create spreadsheets. I personally use Google Docs. The basic columns that you’ll want to start recording here are:

  • graphDate of message
  • Social media platform
  • Product or sale advertised
  • Sub-product information
  • Your call to action
  • Total Impressions/views
  • Comments or replies
  • Likes or Favorites
  • Shares or Retweets
  • Clickthroughs
  • Total engagement across all categories of social data
  • Tone of message
  • The actual message sent

I like to show you behind the scenes information here at Devumi, but that can’t be the case today. Here’s a sample spreadsheet that I put together to give you an idea of what it should look like:

social media data analysis basic

You’re free to spice it up more, change category names, add a column for hashtags used, or anything else that makes it more relevant to your social media analysis needs. Those are just the basics for a Twitter data analysis campaign.

Choose your Categories and Subcategories and stick to them

The first column that needs your attention is Product/Sale category, and  the subcategories that relates to them. These are going to be the broad categories of what you tweeted, and a more specific piece of information. Examples would be:

  • Product/Sale – most opt to just write whether they’re advertising a product in the tweet, or if it’s a special/sale/promotion.
  • Subcategory – after figuring out the above, you’ll want to state exactly which type of product or sale: Was it a sweater, socks, t-shirt, etc. For the sale, was it a holiday promotion, a discount sale on old goods?

An example that would be filled out for a clothing retailer would look something like this:

social media data analysis retailer category


This is a fairly general social media data analysis spreadsheet. You’re free to get even more specific by adding another column to the right for brands being mentioned, what department (men’s, women’s, children, etc) the product came from, or anything along those lines that will help you further organize your data analysis.

Work out basic calls to action

Your actual calls to action may differ slightly within the actual messages themselves, but there are a few basic ideas behind them that ring true on every social media platform:

  • Buy Now
  • Share
  • Visit our website
  • Subscribe (to our YouTube channel, email list, etc)
  • Like/Favorite

Pretty much all of your calls to action will revolve around one of those five basic ideas. Your messages will vary what exactly they’re saying somewhat, but be sure you’re always hitting on one of those on the messages you want to track.

Get to work and fill in that spreadsheet with your data!

Time to really get to work! Here’s a brief example from a clothing retailer:

social media data analysis retailer

I didn’t fill out a fake column for ‘Messages” where the actual message would have been cut and pasted to. Make sure you track this data as you can look at your most successful messages and emulate their style – that’s the whole point of social media data analysis!

Now where do you get this data from? There are tools out there which make it easy:

  • Twitter Analytics will let you export your last 500 tweets, which is some serious data
  • Facebook Insights can be exported in blocks of 180 days
  • and Hootesuite both have click-through data collection options

All three of these tools are free to use. and Hootsuite will allow you to track data for multiple accounts on one dashboard if want to condense your workload a bit.

Now what do you do with all that social media data?

The point of your social media data analysis is to begin to see what messages are connecting with people in your social networks by simply looking at the numbers. Make sure you put a freeze on your first row, and start playing around with your numbers by looking at:

  • Sort total engagements by descending order. Which posts were the most successful? What can you continue to emulate about them going forward?
  • Looking at specific data points – comments, likes, shares – did they perform well enough? What words did you use exactly to get the best reactions? This is where the ‘Message Sent’ column comes in handy.
  • Take the time to compare impressions versus engagement across multiple platforms. This is a point where people are starting to see Twitter as the stronger social platform.

Let’s take a look at the example I’ve been playing around with, sorted for total engagement:

social media data analysis engagement

To analyse the fake data, my holiday sale received the most engagement as people looked to share it and make their friends aware of it.

My tweets relating to children’s clothing are clearly lagging behind. This is a sign to me that I need to work more on either language in them, or if I can’t get the numbers up, shift my focus on advertising them to another type of online marketing.

Every column can be sorted. Every piece of information that you get from this is valuable. Do shoes do better than sweaters in total engagement? Are people more ready to talk about sales than products? Can you get the most shares by saying ‘Please share this now,’ or does a quick ‘Please RT’ get the same results? This is the kind of data you’ll want to look at after using our Twitter retweet service. You can see what that added social proof is doing to spur on more engagement from real Twitter users.

Social media data analysis: Your best weapon

The days where you posted willy nilly, whenever you had a moment, are over when your social media accounts really start to take off. If you’re really on the ball, you’ll start your social media data analysis the same day you start your social media accounts. Lots of data = lots to learn!

As they say, knowledge is power and…

Screen Shot 2015-02-18 at 10.40.07 AM

Share This