Four Ways “Join the Dots” Overcomes Limitations of Social Listening by Overlaying Thick Data to Deliver Deeper Insights

If you’re trying to use social data to do more than develop a social campaign, or track a hashtag or wider marketing campaign, then maybe you’ve found it difficult to get across the contextual interpretation that you were looking for.

Don’t worry, you’re not alone.

There’s a real, legitimate fear for researchers that the data they analyse could be contextualised in the wrong way.

The good news is that by reframing your approach and addressing the limitations of social listening you can find a wealth of insights that do more than state the obvious.  This was a challenge that Kelly McKnight, Head of Culture and Trends at Join the Dots, had to take head-on when she was given responsibility for social listening at the insight agency.  

A couple of weeks ago we shared Kelly’s story and heard how she went from loathing to loving social intelligence.  In this article, we dig a little deeper into how Kelly and her team overcome the many limitations of social listening – that’s four hacks to help you get more from your social insights.  


Social Listening Limitations

If you’ve ever run a social listening project there’s no doubt you’ll have come across a whole lot of barriers to work through. Some of these may have included:

  •      How to write the social media search query
  •      How to contextually interpret the data
  •      How to handle visual analysis
  •      How to be confident that you’ve got all the data
  •      How to ask questions when you can’t ask questions

We also have to address a real elephant in the room: social tools don’t provide as much insight as they’d lead you to believe.  

To get the kind of insight you need to answer more complex questions you need a clear strategy to identify and address the limitations.  Kelly explained:

“we need to think about social listening as part of a wider solution – a set of tools for understanding people, rather than a solution itself”.

Now, if you’re a marketer reading this you might be thinking ‘that’s not what I was made to believe’. But, here at The SI Lab we urge you to remember that the tools don’t provide all the insight and they are most definitely fallible.  

Here we shed light on Kelly’s concept of Thick Data and how thinking outside of the box helped her to overcome a series of Social Listening limitations.


Thick Data

When Kelly came across the concept of Thick Data it changed her perception of what she was trying to do with social listening.

After listening to the TEDx talk by technology ethnographer Tricia Wang, she began to understand that it’s the way in which we try to use big data that makes interpretation problematic.

Taking the advice “it’s not big data’s fault, it’s how we use it” to heart, Kelly sought to find new ways of overcoming the limitations of social listening by integrating ethnographic qualitative approaches. This allowed her to reveal the contexts and emotions of the people under study.  


Four Social Listening Limitations and Approaches to Overcome Social Listening

We’ve already come to terms with the fact that there are real limitations with Social Listening. But what can we do to about them?

Here are Kelly’s creative solutions to help you overcome five inherent issues with social listening.

Limitation #1 Lack of Cultural Context

The lack of cultural context in social data impacts on two areas of analysis:

  •      Your ability to write context driven social media search queries.
  •      Your ability to interpret the cultural and contextual findings.

Anyone who has ever run a social listening research project will tell you that writing the search query is the most important part – garbage in, garbage out – right?!

The trouble is, if you haven’t properly prepared then you end up sitting staring at that flashing cursor and start writing all manner of keywords and phrases just to fill the space.  You’ll pull through the data you’ve lacklusterly retrieved and what you’re left with is a mess of unimaginable noise and irrelevant mentions.

It’s the same when the time comes to interpret the data – if you have the wrong cultural references or contextual understanding, you end up with widely inaccurate findings or struggle to make sense of what you see.

Kelly’s answer to this?  It’s beautifully simple.

The team at Join the Dots work with their network of Illume Guides, that is, real people in local markets who can help them set up searches, layer on context and understand the findings.  The Join the Dots Illume Guides help the team to go beyond taking the data at face value.

In a recent study with a UK-based global skincare brand, Kelly and her team analysed skincare innovations in eight markets to understand if the brand was behind the curve.  The team interviewed local market guides with an interest in skincare and used the insights to set up a culturally relevant keyword search. They then used the guides to help them interpret the findings.

Kelly tells us that this enabled them to get beyond the relative size of the buzz and move into opportunity areas informed by cultural truths.  This included the importance of ‘skin’ in each market, access to dermatologists and the relative importance of factors such as safety, convenience and tech adoption.  

Limitation #2 Missing Data Because of Dark Social

A big concern for many people working with social data – and, indeed, their clients – is “dark social”. It comforts a lot of us in this industry to know that we have the ability to gain access to data. After all, without this data our jobs would be practically impossible! However, with a reported 77% of mobile content shared in inaccessible dark social channels (RadiumOne, 2016), you can see why this poses a real challenge when it comes to producing contextually relevant research.

Fortunately, Kelly has developed a creative solution to the dark social problem.  It isn’t necessarily for the faint hearted and it will require more work, but, if dark social is something you want to get to grips with, it’s the best solution we’ve come across.


Develop relationships with key dark social groups. This tactic is essential in order for these groups to feel comfortable sharing the images and conversations they readily share with each other in private. With access to previously unobtainable data to help give context, it makes it much easier to understand underground online behaviour.

On a recent project with a global drinks manufacturer, Kelly put this strategy into action.  Using her Illume Guides, she was able to gain insight into their behaviour across wider digital channels.  She wanted to know everything “to a creepy level of detail” about how the Illume Guides engaged online, in order to know how to influence them around drink choices.

In addition to this, the team actually went to Barcelona and Berlin to hang out in bars with the types of “cool people” who were clearly influencing others.  By building relationships with these “social influencers” they in turn feel at ease including the team in discussions they were having with each other.

Kelly and her team used this data to compare it against what they were seeing in public channels. They also used a survey with mainstream customers in their community to piece together the full picture.  The upshot of all this work? A deeper understanding of what influences mass behaviour.

Limitation #3 Social Listening Doesn’t Decode Visuals

Visual social listening has been around for a while but it’s only now starting to be discussed seriously. The big limitation of visual listening is that it’s not enough to pull the data with a brand logo on it, you need to be able to decode the meaning behind the images.

Most social listening platforms will scrape and code images around basic attributes, but to truly get value from visual data you need more than this.

The solution that Kelly created?  The adoption of semiotics into the analysis of visual communication.

She tells us that on a recent study with a UK meat supplier, her team was tasked with analysing social images to identify consumer-led areas for product innovation and packaging.

Using keyword searches based upon meat-based keywords, her team was able to identify the nature of conversation and the types of images being shared most frequently, and by whom.  They used these images to generate semiotic codes, layering on additional research as a foundation.

Kelly tells us that “visibility and engagement ranking gave very different perspectives around the shareability of different codes”.  This has become increasingly important in the food world and provided Join the Dots with insights that were better than if they had been searching through social media themselves.  

The emergent codes from the semiotic analysis not only identified what people were sharing, why it was resonating and why it was driving consumer trends, but also gave insight into how to design for consumption.

Limitation #4 Social Assumes You Already Know the Answer

If you’ve ever been stuck trying to find insight from the noise of social data and ended up creating insights that were all too obvious in the first place it’s probably because social listening assumes you already know the answer.  You have to code and analyse the data in ways that you already know.

With social data you can’t ask a question directly to your customers – or can you?

Kelly and her team have been pioneering a method of ‘social asking’.  Combining both social listening and social asking, Kelly and her team can ask better questions in social listening that are more spontaneous, originating from the consumer world and using the appropriate language.

All this is achievable because they use their insight communities to gather additional data quickly and in a social way.  Kelly can’t show us a real live project, but provided us with a mock-up of an example community to illustrate the point.

Kelly told us that “using our communities to further our knowledge and test our hypotheses really pushes forward our understanding”.  

Quick Actionable Take Aways

No one ever said social listening was easy or will provide the “full picture”.  Kelly firmly believes that social listening should never be the full solution when it comes to research. It’s essential to join the dots by incorporating other methodologies and data sources.

We know that Kelly’s approaches take extra time, experience, resources and budget, therefore may not always be possible to apply.  However, we have also found that a period of secondary research into consumer habits, purchase behaviours and culture can help with creating queries and interpreting results.  

If you want to take the rigour even further, here is a summary of Kelly’s creative solutions to overcoming the limitations of social listening:

Want to share your own experiences in social listening?  You can create a FREE writer’s account to submit your article.  

Write For Us
Jillian Ney
Jillian Ney
I'm the founder of The Social Intelligence Lab. I champion the growth of the social intelligence industry by helping the professionals and businesses working in it to access best practice, accredited training and peer networking. After working in the industry for 12 years I believe social intelligence should become a recognised discipline - and, I'm working towards making that a reality. More content by

Want to write for us?

Interested in writing content for us? We'd love to hear from you.

Write for us