South African social media stats

Yesterday, via the amazing electrical Twitter machine, Russell Martin asked for some stats on social media and internet usage in South Africa. This made me realise a few things.

1. I have a whole lot of (relatively reliable) stats on the local market that I have gleaned over the years. This is thanks partly to stalwarts like Arthur Goldstuck , Dave Duarte and the cool dudes at Quirk, and thanks partly to research I’ve either been involved with or stumbled upon (if you’ll excuse the pun). MWEB also recently released some tasty survey results, which confirmed (and informed) many of my hunches. (UPDATE: Another interesting survey by Universal McCann)

2. I share these nuggets with hardly anyone – which is a shame (not least because I may be wrong and someone may have better, harder data that can correct me)

3. I have been meaning to start my own personal blog for ages and this is as good an opportunity as any.

And so, dear fellows, a breathless summary of South African social media usage, according to the best and freshest numbers, straight from my grubby paws to you:

As far as I can tell there are only seven social media brands with any kind of mass market traction:

  • Facebook
  • YouTube
  • Twitter
  • MXit
  • Linked.in
  • Blogger.com
  • WordPress.com

We have the most reliable data on Facebook. Their sales dudes shared it with us at the recent Habari Media launch. Granted they are sales dudes, so they may have massaged the figures, but they match up so closely with our own anecdotal evidence so I believe them.

Facebook stats:
Active monthly users – 2.6 million (about 1 million of these are active every week)
MWEB survey: 82% of respondents used Facebook

Male / female: 47% / 50% (3% undeclared)

Age split
18-25: 36%
26-35: 28%
36-45: 13%
46+ 9%

Graduates: 12%

Married: 20%

(NOTE: All these demographic figures are based on people’s profile info which may be incomplete or incorrect)

We have much less specific and reliable data on YouTube, so we’re forced to rely on sources like Alexa (that I don’t trust much*). They rank YouTube as the 5th most popular site in South Africa, after the two Googles (com and coza), Yahoo and Facebook (who are second). Interestingly YouTube used to be fourth until Facebook came along. That suggests that the growth rate of most big international sites is now pretty stable, with only upstarts like Facebook able to make an impact.

Given that News24.com, with its 1.5 million readers, ranks number 10 on Alexa, and Facebook ranks 2nd, we can assume that YouTube sits comfortably between these two – around the 2 million monthly browsers mark. MWEB’s survey found that 32% of respondents used YouTube. This is so low that I suspect many respondents actually use YouTube without even realising it.

Participation rates on Twitter are still low by comparison to the biggest players, but their user base growing like topsy. Arthur puts the size of the local user base at 88,000 “based on 0.4% market share of 22m users”. Note this is USERS – not visitors. I suspect that at least ten times as many people are reading tweets as are tweeting. This is bourne out by Twitter’s 8th place Alexa ranking, which puts it in the same league as News24.com (ie a comfortable 1 to 1.5 million browsers per month). MWEB’s survey found that 28% of people “used” Twitter – which correlates nicely with both Arthur and Alexa’s numbers.

Linked.in has a naturally smaller user and audience footprint than the others, given its specialised focus on business networking. Even so, it has comfortably ranked in the top 20 on Alexa for the past year. Right now it is at 15 – putting it in the ballpark with IOL’s 750K local monthly readers. This correlates fairly closely with MWEB’s finding that 14% of respondents used Linked.in.

Blogger has a massive local audience, and many local bloggers also use their service to host their own blogs. This puts the service into the top 10 on Alexa – a space it’s held for over two years. However it’s nearly impossible to separate out local readers reading local blogs from the confusion. I strongly suspect that a large chunk of the traffic comes from local readers of popular international blogs that happen to be hosted on Blogger.

WordPress.com
is the real wildcard here. They easily make it into Alexa’s top 20 and that’s just with their commercial play. If we take into account how many prominent local bloggers are running WordPress installations, I would put WordPress as South Africa’s blogging platform of choice.

MXit is the hardest to measure out of any of these. MXit stopped publishing their numbers after they surpassed Vodacom (ie 10 million customers) Anecdotal evidence points to nearly 100% penetration in people under 18. MWEB’s survey reported that just 28% of respondents used the service, but I think that says more about the age of respondents than it does about the reach of the brand. And since MXit is almost entirely mobile, Alexa has no view on their popularity. If I had to guess I would put MXit penetration between 40 and 50% of the total cellphone using population – so roughly 12 to 15 million people. That makes it bigger than all the rest of the players put together. And even if I’m out by a factor of ten they are still larger than News24 or Twitter.

UPDATE: I’ve just received some harder numbers on MXit from the lovely Sandra who does PR and strat for them. Apparently they have more like 19 million users in SA. I’m confirming some more details with her and will post again later.

So folks, what do you think? Have you got harder or fresher stats? Please, feel free to correct me – I welcome a good debate on the subject.

* Why don’t I trust Alexa? I think there methodology is fundamentally flawed. They rely on users installing their toolbar and then using that sample data to extrapolate analytics for the entire planet. Compare them to someone like Compete and you’ll see what I mean. I still think their data is useful, but only in the most broad and general sense. I don’t trust, for instance, that WordPress has more South African readers than Linked.in just because Alexa tells me so – but I do trust that they are in the same ballpark together.

*ANOTHER UPDATE*

A great comment by Justin Hartman about Alexa, setting me straight:

“While the Alexa rankings are not authoritative they do give us a “fair” indication of website popularity. When you say that you need to install the toolbar I say humbug.

For years now Firefox has been sending Alexa anonymous data without the need for the toolbar. There is an option somewhere in FF that is selected by default that gives FF permission to send “anonymous” data to Mozilla.

In addition, the Google Toolbar also does exactly the same thing if you have selected to send Google anonymous usage data.

So, based on this, I don’t think Alexa is completely off the mark however it does preclude many many IE users who don’t have the Google Toolbar installed.”

Thanks Justin 🙂

8 Responses to “South African social media stats

  • Al, these stats are a saving grace – as you mentioned, we (even the cool dudes at Quirk) sometimes struggle to find reliable social Media stats and any client proposal or presentation is just fluff without concrete evidence. So I personally bow to your awesoness.

  • Robin Fishley
    2019 years ago

    Agree about Alexa, I remember seeing a massive dip when firefox 3 came out and the plugins did not work for a while. As a sample it is heavily biased toward the more technical user. I find more consistent/realistic data coming out of the Google trends for websites tool. Although it’s a nightmare to pull out the data. http://trends.google.com/websites?q=linkedin.com%2C+wordpress.com&geo=ZA&date=mtd&sort=1

  • Some very interesting info. Since we are looking at Social landscape would it not be important to look at local players such as #zoopy, #Blueworld, #SiliconCape (Yes, it’s a social network) etc. Yes, it might not have the Facebook, Twitter etc. traffic (yet), but we are talking social media stats in SA, surely we should include them please.

  • Thanks for this Alistair – exactly what I was looking for ! Sharing is caring…

  • Justin Hartman
    2019 years ago

    Al, great post and well done. I must just disagree on one thing and that is Alexa.While the Alexa rankings are not authoritative they do give us a "fair" indication of website popularity. When you say that you need to install the toolbar I say humbug.For years now Firefox has been sending Alexa anonymous data without the need for the toolbar. There is an option somewhere in FF that is selected by default that gives FF permission to send "anonymous" data to Mozilla.In addition, the Google Toolbar also does exactly the same thing if you have selected to send Google anonymous usage data.So, based on this, I don’t think Alexa is completely off the mark however it does preclude many many IE users who don’t have the Google Toolbar installed.

  • Hi AlistairThanks for these stats. I also enjoyed your commenton their reliability.Being an fairly new industry, reliable stats are hard to come by. These stats do provide some insight and direction to make decisions about marketing campaigns online.Thanks again

  • ladybird1005
    2019 years ago

    Hi! I would like to suggest you my favorite tool – http://www.estimix.com – a web site statistics and web analytics service. Here you can find information about every single page on the web.

  • Hi! I would like to suggest you my favorite tool – <a href=’http://www.estimix.com’>www.estimix.com</a> – a web site statistics and web analytics service. Here you can find information about every single page on the web.

Speak!

%d bloggers like this: