How Much Advocacy is Going On in Social?

How much does Social advocacy influence what people do? And what’s the value to brands?

Rob Blackie, Director of Social at OgilvyOne UK, has investigated the topic further. The following look at the influencing power of Social advocacy appeared on his blog on December 5, 2014, and a second post discussing the value to brands will follow.

 

The first thing to disentangle is how much Social advocacy goes on. Here I’m defining Social advocacy as any content sharing from or through Social channels, or a recommendation (e.g. ‘Innocent’s new Thai curry is great!’) that happens in Social.

Unprompted advocacy of brands is often low…

Let’s quickly dispose of the second of these. Our global brand advocacy study in 2013 found that unprompted recommendation in Social is generally relatively small. For instance hotels typically only have one positive piece of advocacy for every 100 mentions in Social, even when some of the hotels we studied had 80% customer satisfaction scores.

Anecdotally from many listening studies I’ve seen or done over the last 8 years I’ve worked in Social, it is also the case that unprompted advocacy is often low.

On the other hand prompted recommendation is now huge. Whether it’s content programmes in Social channels, or product reviews on websites like Tripadvisor, there is a lot of this content online.

How much though?

Obviously it’s very hard to track how much offline advocacy results from Social, but it is increasingly possible to get a handle on how much advocacy drives website traffic through people sharing links to it.

How significant is Social sharing?

It’s now well accepted* that standard website analytics doesn’t give a complete picture of how many people visit a website because they have been passed a link. This is because many Social shares that refer website traffic are not recognised by analytics packages. This is called ‘Dark Social’.

Chartbeat have handily summarised these challenges in their latest ‘State of Dark Social’ report.

Source: Chartbeat, 4th December 2014

The upshot is that lots of traffic data that currently shows up in analytics as Direct traffic (i.e. typed in or bookmarked by a user) is probably resulting from a Social share.

In turn this means that the proportion of website traffic resulting from Social sharing behaviours is dramatically higher than previously thought.

I asked RadiumOne if we could have their UK data on Dark Social, and they have very kindly shared it. If you don’t know RadiumOne, they are a Social CRM technology, largely used by media buyers and publishers, and their link sharing technology PO.ST (you’ll have seen it used as a URL shortener) gathers data from 100,000 websites.

Finding one: Social is three times as important as you thought

Dark Social accounts for 71% of UK Social traffic.  This is a pretty dramatic finding.

Source: RadiumOne data for the UK

In fact for some categories this is even more extreme. For one category I’m currently working on there are nine Dark Social visits to websites for each one traditionally measured Social visit.

Finding two: Social is now competing with Search as the leading source of website traffic

Shareaholic recently suggested that Social drives 31% of total website traffic. Buzzfeed, obviously a publisher optimised for sharing, so at the extreme end of the spectrum now find that Socialdrives five times as much traffic as search.

Source: Shareaholic, October 2014

Conclusion: OK Social is a lot of traffic – but what’s it worth?

Social sharing isn’t just a common behaviour now, it’s one of the biggest, if not the biggest, source of traffic for many websites. And it’s being prompted by content in Social channels.

So if there’s so much sharing going on, what’s it worth?

That’s what the next blog post will examine.

___________________________________________

*For more reading on Dark Social see Alexis Madrigal’s writings on Dark Social here in 2012, here following up in 2013 and most recently here in 2014, and BusinessInsider’s writeup of Chartbeat’s recent data.