Hopefully you’ve read the first post in this series and are ready to start thinking about how to pull your social engagement strategy together. Here’s the method that I have drawn up based on my experience. You might choose to use it wholesale, or adapt certain elements. Remember with this sort of thing that ‘done’ is generally more important than ‘perfect’!
This post will focus on data – what you need to look into and pull together in order to start writing a document up. You can broadly use this approach for any of your social channels, though I am focusing here more on Twitter and Facebook, as they are widely used and have fantastic built-in analytics.
Let’s get analytical
Firstly you need to answer the question: what are our audiences responding to? You will need to do this separately for each channel – the nuts and bolts might be slightly different but the process will be the same.
- Choose a timeframe. I would recommend six months as a solid starting point. Shorter periods might be less useful, particularly if you have had a specific campaign running that might have skewed your content output over that time.
- Export the post data. I do this within the platform, as sometimes third party platforms cause variations in the data that I don’t quite understand. That said – if you prefer to use a third party platform, that’s fine too – just so long as you continue to use it when reviewing ongoing performance, so you are comparing like-for-like. Note that you might also need to do multiple exports, as some platforms have a limit (eg three months) for the date range you can export.
- Determine what to measure. I tend to look at visibility and engagement.
Twitter: Exporting the data from Twitter gives you “impressions” and “engagement rate”. Reach is generally a more useful measure for engagement, as it tells you how many people have seen your content in their feed; impressions can be misleading because you can’t break down how often the same people have seen your content. That said, for the purposes of this work, impressions is fine so long as you continue to use the same measure when evaluating ongoing results.
Facebook: Add likes, comments and shares, and divide this by the reach to get an engagement rate. Facebook gives you lots of other types of engagements, but these are the key ones. The important thing is to be consistent in how you approach any given channel so that future measurements against your initial benchmark are meaningful.
Instagram: You cannot export data from Instagram in the same way, so to measure engagement you will need to create a manual version – for each post, add likes and comments and divide by the number of followers.
Because the process is different for each channel, remember results between channels are not directly comparable to each other.
This is #2 in a set of 4 articles, check out instalment #1, instalment #3 and instalment #4 to find out more!
Want more info on what to measure? Check out what our Social Media Network had to say.
When setting up your spreadsheet to analyse the data look at average reach / impressions (whichever you are using) and engagement rate, by month, for each channel. With Twitter, remove any direct replies (ie ones that start with @ as the first character) they are not visible to most of your audience so they are not relevant when you are assessing your performance. For your average engagement rate make sure you do NOT simply take an average of the engagement rate in your data – this does not give an accurate average. You need to add the sum of all the engagements, and divide by the sum of all the impressions / reach.
I also track the highest reach / impressions and engagement rate each month. You end up with a table like this for each channel:
Average engagement rate |
Average impressions |
Top engagement rate |
Top impressions |
|
September 2018 |
0.89% |
9501 |
11.11% |
32508 |
October 2018 |
0.94% |
21182 |
6.30% |
64015 |
November 2018 |
1.77% |
6361 |
3.00% |
38773 |
December 2018 |
2.63% |
7801 |
11.45% |
32500 |
January 2019 |
2.48% |
28501 |
8.86% |
318354 |
February 2019 |
1.59% |
16774 |
2.91% |
44447 |
And this can be used to create graphs like this – which, as you grow the data over time will allow you to spot patterns:
That’s the driest part done so well done for sticking with it and I promise from here it gets more fun!
Understanding what is working
Now you have laid the groundwork for your onward monitoring, let’s start thinking about content:
- Identify what has performed best and worst in the time period. Doing this month by month will help you build up a picture and again removes any bias because of seasonality or key campaigns.
- Look at the top five and bottom five posts of each month and start to identify key drivers of success.
Screengrab the posts and copy all of the top ones into a document and all of the bottom ones into another, and start looking at what factors unite them. You might spot themes emerging – for example, I was doing some research recently looking at a heritage organisation, and spotted that posts relating to suffrage and feminism were performing particularly well for their audience. This was really interesting to me. The organisation is largely concerned with the physical history of England – buildings and monuments – and most of their posts reflect that. But this shows that the audience are also extremely engaged with social history – and their strategy can make use of that information.
When looking at your posts, ask yourself questions for both sets of posts. You can tailor these according to your cause and your brand but as a starting point:
- What is our tone of voice like? Are we warm, authoritative, disengaged, informal?
- What emotion does the post prompt? Is it hopeful, or grave?
- Are we using lots of jargon, or is our language accessible?
- What do the images look like, and what do they show? Are they good quality, are they bold and clear? Are they recent or historical?
- Do they contain direct calls to action – and if so what type?
You should start to spot factors that connect your good posts – and the bad ones. And once you do, you can start to build up a “do”s and “don’t”s list to help plan future content.
Check out the next post for tips on developing strong content.