Who, what, when? Analysing social media traffic and demographics
If you’re a digital marketer and you want to start a social media campaign, it’s likely that your default channel will be Facebook. But have you actually considered what audience uses this social networking site? And have you thought about whether the social media you choose will drive traffic to your site and increase revenue?
This analysis aims to profile social media ‘clickers’, those users that actively engage with social networks, or as some might say, the ‘money makers’ of the social media world.
Marketing managers often have preconceived notions of the demographics of social media users: teens and millennials who enjoy sharing cat pictures, for example. So, we thought we’d crunch the numbers for a range of global data sources and discover the truth. By comparing each user base to the overall population, we can see the differing characteristics of users who are using the world’s major social media services, and whilst we’re at it provide some insight into which is the best platform to host marketing activity.
The data below is free for you to use for your own purposes – we simply ask that you provide attribution (licensed under CC BY 4.0.)
As with any analysis, there are caveats. If you would like to read ours, please jump down to the Analysis Notes section at the end of the article.
What social media services drive the most traffic?
Facebook users drive more traffic to external sites than any other social network. This may not be particularly surprising when you consider that Facebook has 1.94 billion monthly users, more than any other social media site.
What is surprising is the second and third social networks on the active users list: YouTube is second, with over 1 billion users, and Google+ is third with 540 million users. In terms of driving traffic, Youtube is seventh on our list, and Google+ isn’t even in the top 10. In other words, according to our data, Twitter, Instagram and other social channels are better at driving traffic, even though they don’t have the most users.
However, if we look at social media market share and compare that to the percentages above, an interesting picture emerges:
Facebook has 36% market share and is driving almost 70% of social media traffic. But what stands out is Twitter; with almost 8% market share has driven 22% of social traffic.
This means that although Facebook drives more traffic, Twitter users are far more likely to engage with content. This potentially means that your media will be more effective posted on Twitter than other social networks because it will have a higher click-through rate.
LinkedIn is another social media to consider as it has 2.4% market share and 2% of traffic.
However, the other social media platforms shown above should not be dismissed as the audience is huge. The above metrics could be the results of two things:
- The social media is poor at driving traffic to a site.
- The Paid Ad element of the medium is underutilised.
- The medium is more for brand awareness than to drive traffic.
These are not all negative reasons – understanding how the paid channels work for these mediums, or understanding the power of how these mediums can influence brand awareness, can provide positive results. The only way to find out is to trial a campaign – you might receive some unexpected results.
Who? Social media user demographics
Does your target audience use Facebook, or are Twitter users a better match? What are the market conditions and are they changing? When should a marketing message be delivered? Is there a preferred time or day? Understanding these questions can help you to identify opportunities and implement a successful social media marketing campaign.
The article uses the top three traffic drivers, which have the most user demographic data, to understand the defining characteristic of the channels’ users. The data below has been analysed using Indexing, which means that each data set shows how the users of each search engine differ from the overall population:
- If the value is greater than zero, it means that specific trait is more common than the ‘average’.
- If the value is less than zero, the converse is true; that specific trait is under-represented on that social media vs the average online audience.
By analysing the data in this way, we can show the differences in demographics between social media platforms, while avoiding being misled by the general demographics of online users.
Gender is usually an inconclusive characteristic as statistical significance is very rarely achieved. However, when it comes to social media there is a clear gender divide.
Twitter and LinkedIn users are more likely to be male than the general population, whereas Facebook users are more likely to be female.
Age is a great demographic as it forms the basic DNA of a customer persona. Knowing the approximate age of your audience gives a good indication of what life stage they are at. At a very high level, someone under 18 is likely to be studying or will have just started their first job. Those in their 30s are likely to have young kids and those over 60 are likely to be starting their retirement.
Twitter and Facebook users appear to a similar age range, around 35 to 55. However, LinkedIn is more likely to have a younger audience than this, about 25 to 44.
What? Social media users by device
Understanding the device that your audience uses is important, as it will help you to shape the content that you’re delivering.
The above graph makes it clear that social media users are mobile and tablet advocates. This is a clear message that any social media marketing campaign that is designed to drive traffic must have a mobile ready landing page – otherwise you may be throwing money away!
When? Social media users as first time or returning visitors
New or returning users
What is your campaign designed to do: generate new customers or engage with existing customers? This an important question that is quite often missed. Looking at the user type data for the social media platforms will help choose the correct platform to host content.
The above shows that Facebook and Twitter are great at attracting new users to a site, whereas LinkedIn is good at engaging existing customers.
Days since last session
For those customers that return to your site via social media channels, it’s good to know the average amount of days between sessions.
The chart above shows that site users are more likely to make a return within 10 days on LinkedIn and 4 to 25 days on Facebook. Twitter users take the longest to return a gap of 11 to 60 days or over 100 days. If you’re delivering content via ads at regular intervals, then understanding user behaviour will help you plan when to deliver your marketing message. For example, you could set up a remarketing ad that targets users on Facebook for 25 days following their last visited your site.
Time of day
Understanding the most likely time your audience is on social networks is vital to improving conversion rates.
From the above chart, it’s clear that social media users are very unlikely to be online before 8am.
For Twitter and Facebook, most users will be online in the evening, whereas LinkedIn users will be online during the working day.
Facebook has the most users of any social media platform and therefore has a massive audience that you can market to and engage – but it may not necessarily be the most cost-effective. Platforms like Twitter have what appears to be users that are more likely to click through to a website from Tweets and ads. This means that you may get a better return on investment by opting for Twitter ads over Facebook ads.
Other social media platforms may not drive traffic to a site but are serious contenders for brand awareness due to their large user bases. These additional platforms may provide you with a new audience, and trialling them may produce interesting results.
Focusing on some of the main social platforms we can see that users have a wide range of characteristics, the most significant being gender, age, and user type. Understanding these differences, and using them as part of your digital marketing strategy, could prove beneficial.
Remember to be mindful that this analysis was a generalised approach and that the outcome could be very different for your specific website, approach and audience. A ‘test and learn’ approach should always be taken when doing any new marketing activity as otherwise you risk not achieving the best performance possible. If a statistically significant improvement on a test happens, and you have complete confidence in the data, only then should a campaign be rolled out.
If you haven’t already, check out our article on profiling search engine demographics.
1. The data used in this analysis is website social media referral traffic aggregated from Google Analytics accounts with demographic and interest’s data enabled. Several websites with a global reach and a mass-market audience were analysed, with over 10 million data points. However, by their nature the websites analysed had a Western bias, and this is reflected in the data. We’re confident that the bias in the aggregate data is minimal towards any specific age group, gender or location.
2. The vast majority of the traffic analysed was of a B2C nature, with very limited B2B audience. The data will be biased accordingly.
3. Univariate analysis, as used here, only looks at one variable at a time and does not consider other external factors. It has the advantage of being quick and easy to understand, and can give you instant insight for each variable in a database. However, it does not allow insight from one variable to be combined with other variables. For example, if the data shows that male users were more responsive to a marketing campaign than women, and we also noticed that rainy days have more campaign responses, we cannot conclude that male users on a rainy day will be more responsive.
4. This analysis is based on social media traffic for 2016 and, where specified, compares this to 2015.
5. This is based on direct traffic that followed links from the social media site. This analysis does not take into account brand awareness ie where a user sees an ad and remembers the brand, but doesn’t engage with the post.
6. This is a generalised analysis and is not specific to a company or campaign. It is designed to offer insights and guidance only – campaign strategies should always be formed on company-specific data.