Social media is rapidly enabling a shift in how brands understand and engage their customers. Advice abounds on the value of a data driven approach to your social media efforts, and whether heeded or not, thanks to contributions by the likes of @mich8elwu, @briansolis and @kdpaine, social marketers are increasingly aware of the importance of measurement.
However even if you know what to measure on social media, there is much less guidance on how to interpret what you find once you are crunching through the data – and how to benchmark a good or bad result.
Wow, I’ve got a million fans!
When measuring the level of engagement on a Facebook fan page – a metric such as “average interactions per fan” can confuse or mislead unless it is understood in the context of how other relevant pages of similar sizes are performing. This is because many measures of engagement are strongly correlated with the size of a community (number of fans in the case of Facebook). To get a reasonable benchmark of performance, you ideally want to understand where you sit on the curve relative to other pages of a similar size. On its own, a high engagement rate is likely to tell you more about the fact you have a smaller fan base than your competitors – rather than how your engagement meaningfully compares.
How many friends did you say you had?
As we start to connect with our customers through Facebook social logins – and look to build a deep understanding of them based on their Facebook profiles – context again is critical when interpreting what we find.
Basic demographics such as age and gender, and the number of Facebook friends and likes a user has are now readily available through Facebook permissions and can be used to inform marketing decisions. For instance, the number of friends can provide a reasonable proxy for the social reach of a user - but how many friends is many? According to Facebook the average Facebook user has almost 200 friends, but this is highly skewed with some users having up to 5,000 (Facebook’s imposed limit).
Many of our behaviours and characteristics are highly correlated with age (income, risk etc), and not surprisingly, the same applies to many Facebook characteristics such as the number of friends we have or the number of pages we like. When faced with making sense of a set of Facebook profile data, plotting variations by age is almost always a good place to start, just as plotting metrics against number of fans is a great place to start when looking at macro Facebook page performance.
Chart 1.a) shows the average number of Facebook friends for a sample of 20,000 Facebook users collected at social login. The averages (mean and median) peak for users in their early 20s and then rapidly drop off. So if you are trying to prioritise which customers to engage based on their number of friends, the thresholds you set will differ depending on whether you are looking for the users with the most friends overall (younger) or a balanced representation that includes older demographics. Focusing exclusively on the average number of friends for the group overall (184 in this case) can lead to unintended results.
(See graphs in Gallery section)
The number of Facebook likes a user has exhibits a similar yet different pattern to friends. Chart 1.b) highlights the strong correlation between the number of likes a user has and their age, and a much larger variation in the number of likes within users in their teens and early twenties. Understanding the number of likes one of your customers has can provide a proxy to the exclusivity of their like that you have earned (or bought) and another valuable input in an outreach programme. Here again, context is key in creating your benchmarks.
Know your audience
Whilst the overall shape of the distributions seen in the charts above exhibit strong similarities with other Facebook audiences we have analysed – whether 10,000 users or 1,000,000 users – the precise height and averages can vary considerably between them. Age is one driver, but social and cultural factors strongly influence the number of friends and likes you have – and can shift the average curves up or down. Understanding the shape of your audience is critical to developing relevant benchmarks as part of a data driven social strategy.
Of course context matters whatever you are measuring – you don’t need to be a data scientist to realise that averages can easily deceive, and only tell part of the story. Knowing my 10 month old daughter’s exact height and weight is fascinating but doesn’t really tell me anything about how she is progressing – but a plot on an age-growth chart with a percentile score for her age helps me make sense of why none of her clothes seem to fit her.
You will be sent a verification email. Click on the link in the email to post your comment.
Opinion, blogs and columnists - call them what you like - this is the section where people have something to say. You might agree or you might not - whatever opinion you have make your views known in comments. Views of writers are not necessarily those of The Drum. If you would like to contribute a comment piece, email your idea to firstname.lastname@example.org.