Social Media Analytics tips

2019 is well underway and like any self-respecting marketer, you’ve probably spent some long hours picking apart your 2018 data. Amongst all the data to analyse, measuring the success of your social network activities is certainly the touchiest part to handle. 


Why? These are platforms (that don’t belong to you) to whom you entrust your content, which complicates performance measurement: you will use both data provided by these social platforms and data from your digital analytics tool. The other difficulty, unfortunately very present on everyone’s minds at the moment, is being able to fully trust the data provided by Facebook, Twitter and the like. 2018 was particularly rich in scandals and other security breaches, notably with the Cambridge Analytica affair, Facebook’s difficulties in providing reliable data, and Twitter’s purge of fake accounts. Vigilance is therefore in order! On this blog, we often refer to platform independence issues, notably the issue of platforms being both “judge and jury”: the tools in which you invest your ad budgets (search, display, remarketing…) are perhaps not the best placed to provide data related to their performance. Conflict of interest, anyone?


Here are a few tips to help you with your social media analyses. While it might be a little late for your 2018 reporting, there’s still time to start out on the right foot for 2019!  

Focus first on social network data

Each platform has its own analytics section which offers certain exportable indicators: 

  • Facebook has the largest number of metrics, but be careful to take it all with a grain of salt. The social network has often had to issue a “mea culpa” due to providing companies with only approximate numbers (for example, views of sponsored videos, into which many advertisers have invested in recent months). 
  • Twitter offers the main metrics, but they are not as easy to leverage: incomplete data upon export, difficulties in getting a file that can be used in Excel, analysis periods that are too short…  
  • LinkedIn has come from quite far on the analytics part, and is getting better day by day since Microsoft acquired it: more and more data is available and can be leveraged outside the social network.  

Each platform has its own API, meaning you can export your data wherever you wish. Analytics Suite users can easily integrate this data into a dashboard, thanks to the import of external data. If you’re getting started with APIs and you want to save time, consider using third-party tools like Hootsuite or Buffer which will generate ready-to-use dashboards for you, based on the aforementioned APIs. Finally, remember to export this data regularly because depending on the metrics and social networks, certain data just won’t be available after some time.

Here are a few useful analyses that can apply to all industries (beyond that, it’s up to you to complement these metrics according to your business objectives and your audience!):  

  • Net number of new followers (avoid the absolute number of followers – it’s a vanity metric which only gives an idea of your audience size and looks nice in reports, as theoretically this figure always keeps growing…). 
  • Engagements and engagement rate: likes, retweets, shares, comments, clicks, mentions… these are the ultimate signs telling you that a user has interacted with your post.  A safe bet here.  
  • Impressions: be sure to distinguish between organic impressions and paid impressions. Pro tip: Facebook also provides information about viral impressions, generated by a user’s engagement with your publication. 
  • Video views: the beginner’s error would be to take the total number of views provided by the platforms, which includes all views without distinguishing by duration of the view. Opt for more qualitative performance indicators, like the number of views that watched at least 50% or for 15 seconds, for example, and base your CPV (cost per view) analyses on this new number of views. Your ROI may take a hit, but your analysis will be more accurate! 
How do you measure ROI of social media ad
 Measuring engagement is a good starting point for your analysis, as shown in this study, State of Social, published by Buffer in January 2019. But don’t stop there! 


Verify and refine your social media analysis in your web analytics tool

Your ultimate objective in being present on social networks surely goes far beyond just racking up likes and comments, right? This is where your digital analytics solution can now step in. 

First of all, don’t attribute too much importance to clicks on your publications. They do not reflect the reality of a visit to your site or a session on your mobile app. Test this yourself: compare the number of clicks on a publication with the number of visits coming from this publication in your analytics tool. We generally observe a 20% discrepancy, in the best of cases. Of course, this is related to the nature of the metric: a click is indeed different from a visit! 

BudgetClicks (LinkedIn source)Visits (Analytics Suite source)Cost per clickCost per visit
300
79423.8
7.1

A real example taken from a LinkedIn sponsored publication from December 2018. The visit costs indeed much more than the click.

Are you used to looking at the number of visits and conversions on your ads directly on Facebook Business Manager, Twitter Ads or LinkedIn Ads? Are your audiences based on the pixels you’ve installed on your site? Watch out – it’s possible that part of your audience uses a tool like Adblock or Ghostery to block social network trackers (but your favourite digital analytics provider offers a solution!). The result: your data is incomplete and your analysis is biased. Get in the habit of using a third-party analytics solution to verify clicks and conversions on your social media campaigns. In practice, a simple xtor campaign tag will do the trick.  


Isolate organic traffic from paid traffic in your analytics tool

You’ve surely read and observed it, the reach of publications on different social networks has clearly declined in recent years. More and more companies have turned to sponsored posts to give their publications more visibility, if only to get in front of an audience that seems to be reachable (no, not all your followers will see your latest post!). The advantage is, of course, to be visible to your personas and extend your audience beyond that of your current followers. 

It is therefore necessary to distinguish organic traffic from paid traffic in your analyses in order to calculate your return on investment. The solution for Analytics Suite users? Use a custom organic source for your organic traffic and declare one or several ad campaigns for your paid traffic. You’ll be able to analyse the relevance of each source of traffic, and see how each source contributes to your conversion efforts.

Obtain trafic sources with Custom Metrics

Pro tip: Use Custom Metrics in the Analytics Suite to obtain the share of organic traffic vs. paid traffic in your dashboards! 

Though it is typically under-leveraged, social media reporting remains a very interesting and useful challenge for digital marketers. It’s an exercise that combines the leveraging of both internal and third-party data, the mastery of several platforms with their specific metrics, and a good understanding of your business objectives to give it true meaning.



Credits:

Elena Koycheva, Unsplash

Author

Alexis joined AT Internet in early 2014 after obtaining his master’s degree in digital communication and community management. Passionate about social networks and online advertising, he is also in charge of marketing automation at AT Internet. A proper geek, he strives to stay on the cutting edge of all innovations related to social networks and new technologies.

Comments are closed.