Dark social isn't a black box without a key – here's why

Brands’ concerns are genuine, and shouldn’t come as a surprise. Long pressured to prove commercial bang for their buck, marketers recognise that most dark social traffic (e.g. link sharing in private messaging apps, texting and email) limits analytical insight or even awareness.

Who comes to our sites and assets? Where did they come from? How did they hear about us? And crucially, how can we justify as much as 90% of social marketing going towards public platforms, when we know that 84% of social sharing happens behind closed doors, in dark social?

Many creative agencies have long relied on social listening and fan/follower demographics from social media platforms like Facebook and Twitter, to prove audience behaviours, attitudes and interests. Strategic and creative decks use screenshots of tweets or earned conversations to validate strategic decisions. This isn’t necessarily wrong, but knowing that so much social conversation goes unobserved does raise warning flags. What I choose to say on Twitter (public ‘me’) can be very different from how I communicate on WhatsApp (‘private’ me). Which ‘me’ should brands target? Which ‘me’ is most likely to respond?

Part of the solution is to go beyond public social, to include the potentially rich array of qualitative data from surveys, focus groups and interviews. For example, We Are Social was born in social, but we now go far beyond likes, shares and comments, recognising that insights based on those metrics alone may not uncover the interpersonal human truths we base our work on.

New developments allow us to gain a stronger understanding of what’s going on. Instead of being a black box without a key, dark social has become (slightly more) integrated in digital marketing tactics. News websites, for example, now frequently include share buttons for WhatsApp, rather than just Facebook and Twitter. Also, brands like Domino’s are experimenting with chatbots to gain new custom from audiences on Facebook Messenger. And Universal Music has gained more insight from its Po.st URL links, which track who’s sharing an artist’s content, and with whom.

Of course, this remains relatively restricted, compared to public social, and it’s worth stressing there’s a good explanation for this. Sharing is often private for a reason: I don’t intend to open my digital conversations with my mother or children to brands hunting for proof or insight.

In one recent focus group of Muslim millennials I moderated, several participants said they shared news about and reactions to Islamophobia in WhatsApp groups, rather than on Facebook or Twitter. They felt that discussions on the latter could spiral out of control and become hateful. Such privacy should remain within their - and everyone else’s - rights.

In general, we’ve seen the rise of privacy concerns and developments, such as ad-blocking and the right to be forgotten; we would do well to remain conscious of these.

However, there’s nothing to stop creative agencies from asking people for their opinion: creative agencies are challenged and sometimes surprised by the opinions of real people. Working on a recent Snapchat campaign for an electronics brand, we asked users a range of questions in a brand lift study in the app itself. In essence, the survey allows us to combine performance with impact and gain a more nuanced understanding of its success.

In general, qualitative investigations in dark social could go a long way to answer strategic and creative questions, even if there might be respondent self-censorship (but how different is that from traditional qualitative research?).

Also, while often lacking precise demographic data from dark social sharers, we’re still able to observe the most frequently shared stories or creative executions. Recent research, for example, has shown that education, religion, and style and fashion are the three most shared types of content through mobile dark social globally. What’s fascinating, however, is the flip side of this data: the top three most clicked on types of shared content are food and drink, health and fitness, and style and fashion.

In other words, a lot of what’s shared isn't actually clicked on by recipients. Content about religion, to take an example from this data, is more often shared than engaged with (a case of “did you watch that video I sent?”).

Ultimately, a large part of dark social is bound to stay invisible to brands. If we found a way in, users would be likely to migrate to other private areas. I know I wouldn’t like a world of openness with everyone, everywhere.

And yet, we can always ask people what they think, feel and do in or about dark social (they don’t have to answer). And we can develop dark social sharing evaluation by content type - looking at sharers and recipients - to understand which cultural trends spike, fall and stay consistent, in relation to which events, influencers and campaigns.

That’s a solid starting point.

André van Loon is Research & Insight Director at We Are Social. He tweets at @dre_vanloon