#ExploringSocial: Facebook for FMCG Brands — The Methodology
Yesterday, we released the first in our #ExploringSocial series — focusing on how FMCG brands can make the most out of Facebook. If you haven’t seen it yet, you can find the full report here.
In this post, we‘re going to do a deep dive into our methodology for the report — how we went about creating the report, and how we made sense of all the data. Have a read on if you want to learn more about how you can manipulate the social media data you have access to, to find your own trends!
First of all, Supermetrics is the tool you need to collect data from Facebook. It integrates with Google Drive seamlessly, allowing you to pull in metrics from various sources (Google Analytics, Facebook, YouTube, Twitter, AdWords, Bing Ads, Google Webmaster Tools) and more. This tool saved us a lot of time.
After collecting all of our data, it was time to make sense of it all. We went about this by creating a list of questions (we love Socratic questioning!) that we get asked on a frequent basis by our clients — and worked out how we could answer these questions using the power of data.
Question 1: What is the typical audience for an FMCG brand on Facebook?
To answer this question, we collected data on demographics. The ‘Insights’ tab on Facebook provides a great overview of the age and gender of the fans on your page. Scroll down to ‘People’ on the left hand side of the Insights tab, and you will be presented with something like this:
We averaged out this data across our whole sample set, to provide us with our overall ‘typical audience’ for an FMCG brand on Facebook. We made sure to remove any products with age restrictions, for example, alcohol brands from our sample set. Here’s what the average FMCG audience looks like:
Question 2: What is best type of content for an FMCG brand on Facebook?
Our next question was about content. We wanted to gain a deeper understanding about which type of content works the best for an FMCG brand on Facebook — is it content that places the product itself front and centre? Content that taps into wider lifestyle themes? Competitions that incentivise engagement?
To do this, we used Supermetrics to collate every single piece of content posted by brands in our sample set over a three month period (October to December 2015). Using Excel, we then ranked this content by Engagement Rate % — and took every piece of content that received an engagement rate above average (7.8%) and reached over 300 users (to ensure that the absolute engagement levels were significant). This left us with a sample of 91 pieces of highly engaging content.
We then went through each individual piece of content and manually coded it into one of three categories — product, lifestyle, or competition. We were then get an overall view of what sort of content is the most engaging for FMCG brands, shown in the pie chart to the left.
Question 3: What makes a piece of content go viral?
This is probably the single question that we get asked the most. What’s the key to getting a piece of content to go viral?
We started this by attempting to define what viral actually means — it has become such a buzzword that it’s ‘scientific’ meaning has been slightly lost in translation! Essentially, virality is the extent to which you go beyond your current audience — usually due to people rapidly sharing the content.
Facebook define viral reach as:
The number of unique people who saw your post or page mentioned from a story published by a friend.
Supermetrics have had a stab at measuring the impact that ‘virality’ has on reach — using a metric that they call ‘Viral Reach Amplification %’. They define this as:
The percentage increase in reach that the ‘viral reach’ is accountable for.
They work this out using this formula:
((Reach incl. viral) / (Reach not incl. viral) — 1)
Essentially, the formula works out the impact that virality has had on the reach of a post. We applied this formula to every single piece of content posted by each of the brands in our sample set — and took the top 25 posts in terms of virality. Each of the posts had a Viral Reach Amplification % of over 400% — not to be sniffed at.
We then dug down deeper into each post and identified the reason that it achieved virality. It became immediately clear that each post fit into one of two camps — it has a high volume of shares, or it had a low volume of high quality shares — in other words, the post got shared by an influencer.
We manually coded each of the 25 posts into either of these two groups — and the results were very clear. Just under 80% of the posts achieved virality because they had a high quality share, rather than through sheer volume of shares. The implication from this is clear — with virality, at least initially, it is very much a case of ‘quality’ over ‘quantity’.
Question 4: Is organic reach really dying?
The next question we tackled was that of organic reach. This has been the subject of so much press over the last year or so, and so we wanted to find out once and for all if organic reach really is dying, and how much of your audience actually see what you post.
To do this, we used a sample set of thirteen FMCG brands who have been active on Facebook for at least two years. We collected data on monthly organic reach for a two year period (December 2013-December 2015) and divided this figure by the total number of posts in that month, leaving us with an ‘average organic reach per post’ measurement for each brand over the two year period. We then took this figure and divided it by the total audience size for each brand on a monthly basis, leaving us with our final figures — the average percentage of audience that was reached organically, month on month, over a two year period.
We averaged this data out across all thirteen of our sample brands, and the findings were clear — organic reach for FMCG brands has more than halved over the two year period we analysed:
We then went one step further and broke down organic reach by type of post, to see how you can maximise the little organic reach that you do receive as a brand.
To do this, we broke down the posts that we analysed by post type — videos, photos, text posts, and links. We averaged out the percentage of audience organically reached for each of these types of content over a six month period for all of the brands within our sample set, which left us with this:
The graph above speaks volumes — video is favoured by the news feed algorithm, receiving above average organic reach. (8.84% of audience reached for videos, compared to 6.4% for the average piece of content). If you have a large audience size, this is a significant boost.
Question 5: What’s the best time of day to post on Facebook?
Another question that comes up frequently — what’s the best time of day to post in order to make sure your content has maximum impact?
We were lucky in that Facebook actually offers some great data on this in the Insights tab — if you select ‘Posts’ on the sidebar, you are given this data:
We then turned this data into ‘percentage of audience online’ by dividing the figures shown above by total audience size. We calculated this across all of the FMCG brands in our sample set, and then averaged out all the data to find out when the ‘typical’ FMCG audience is online on Facebook.
We found that 9pm-10pm was the clear peak in terms of audience online — just under 50% of our ‘typical FMCG audience’ were online at this time. We therefore recommend releasing content a couple of hours before this, so that it can gain some momentum ahead of this peak.
Question 6: What is the average engagement rate on Facebook?
This comes up a lot for a very natural reason — brands want to know how they are performing against other brands. An engagement rate taken in isolation doesn’t tell much — an engagement rate compared to your competitors can be really insightful, however.
This part of our report was an attempt to do just that — to benchmark a Facebook engagement rate for FMCG brands.
To do this, we worked out the average engagement rate per post for all of the brands in our sample set. We worked this out by dividing the total reach of a post by the total engagements — and averaged this out on a monthly basis for each brand. We then took these figures, and created a combined average across all of our brands, giving us the following graph:
We found that the average engagement rate across the two year period was 8.71% — however, perhaps what is more interesting is the fact that the average engagement rate appears to be in decline. The average engagement rate for FMCG brands was 7.8% in 2015 — this is what you should benchmark yourself on. However, we are interested to see if this drops further in 2016…
That’s how we wrote the report. If you have any questions, please don’t hesitate to get in touch. And of course, if you haven’t yet read the full report, head over to SlideShare to dig in.
