Social Media Intelligence as a Credible Research Method

06 Jul 2016 | Research & Business Knowledge

A Personal Perspective: Social Media Intelligence as a Credible Research Method

Do you remember the day you used the internet for the first time?

I still remember mine; it was a pretty unremarkable school day.  There were about 15 of us in my computer studies class, and we were all crowded around a PC’s where our teacher told us about the new ‘world wide web’.

We were told that we could search for anything on this ‘world wide web’.  I remember seeing the ‘search engine’ for the first time and looking for travel information [coming from a small town I always had wanderlust growing up].  The webpage took ages to load, but I was hooked. 

A lot has changed on the internet since that day.

For us researchers, it’s no longer a source of secondary research but an avenue for primary research and experimentation.

The Rise of Individual Voice

The changes most profound to me are a result of social media, particularly on the rise of the ‘individual voice’. 

The internet is now what seems like an inner monologue of billions that openly gets updated multiple times a day.  For me, I think about these constant updates as making the internet into a living ecosystem of opinion, experience and fascination mixed in with facts, marketing messages and sales promotions.  

It’s from my interest and analysing this ‘individual voice’ that I gained a PhD and found my role as a researcher.  Looking back, I’m sure the schoolgirl in me would approve.  

I now work as a consultant to researchers [and marketers], advising them on analysing and using social data to tap into customers’ conscious and non-conscious thinking and motivations. 

Overcoming The Challenges

There have been endless challenges around credibility, rigour and standardising methodological approaches of my research.  While there are journal articles and best practice approaches to guide the research process in other forms of research, social media intelligence has been a bit like the wild west. 

I like that social media has reduced the barriers to entry for businesses to gather market and audience insight.  However, it’s my opinion that the ‘insight in the noise’ marketing claims of social tools is degrading what an insight is perceived to be.  Social tools domination of the market, coupled with a lack of predefined analysis methods is damaging the confidence that social media intelligence can produce rich, actionable customer insight.

I’m happy to say that I’ve developed a robust framework that assists researchers to analyse social media data.  It is from this framework that I run my projects and adopt and a ‘brand emotional intelligence’ standpoint for my clients, whether we are exploring their brand, the wider market or customer behaviours.

Why Consider Social Media Intelligence

It’s these social conversations that are making or breaking brands.  It’s true that the catalyst for most of the early research with social media was around content performance, and then moving out into ‘social listening’ but we’re now moving into an area of real behavioural intelligence (when done properly of course).  

It’s not just textual data you can analyse; there has been a move into analysing emoji’s (love ’em or hate ’em) and images to conduct semiotic analysis.  As with other tools on the market, they have been aimed at marketers as end users, so for the researcher looking to go deeper, you need to ‘hack’ the tools a little to get the data you really want.

I’ve also found that social data is an excellent way to derive insights from unprompted conversations without a high level of post-rationalisation.  Of course, there are points to consider around how people project themselves online, but I’ve found that social data is effective at helping inform other research methods . 

For example, I once worked on a project that looked at sponsorship awareness on the sponsor of a major event.  After segmenting the data to the age demographic of the client, I was able to highlight a level of non-conscious sponsorship awareness.  Although conscious sponsorship awareness was low with the consumer segment, they were purchasing of the sponsors product around the event.  This insight was then used to inform focus group studies where respondents could be questioned and guided through the research process.

Quick Insights from Social Media Intelligence

I work with a lot of marketing agencies and the typical brief from them has been around measuring content performance (but they are slowly moving into using social insight to infer the creative process).  Measuring content performance has traditionally come down to reach, impressions and the subsequent engagement.  It doesn’t tell you why some content gets higher engagement than others, but if we look to adding a layer of behavioural intelligence into this, we can start to understand engagement patterns.

Let’s take these videos that were used in a festival campaign last year; both videos are about the festival and both made by the same creative team.  There is, however, a difference in how the video introduces the festival.

Video 1

Video 2

At the time of the festival, via social media, one video received on average 88% more engagement than the other.  

Which video got consistently higher engagement on social?  Why?

So, it was video 2 received higher engagement.   If we look at the psychology of festival consumption, we can begin to see why.  Consumption of festival experiences is found to be motivated by collective experience and hedonism.  Video 2 was all about the collective enjoyment of the festival, but video 1 was excluding an audience.

There is clearly a bit of post-rationalisation during the evaluation of content, but some agencies are now looking for social media intelligence at the start of the creative process.  Imagine being able to use this insight and then individual markers on the audiences’ language, emotion, experience or personality to guide the creative process… that’s where advanced analysis comes into play.

Advanced Insights from Social Media Intelligence

The volume of data and the complexity of the question asked will determine the complexity of the analysis.  In research, I tend to find that analysis will be more advanced, and the trick for these projects is in the segmentation and linguistics.  

Unless you are conducting multiple projects in one area, you will need new segmentation rules each time.  My advice would be not to use the auto-segmentation classifiers from the social tools but to develop your own – because you have research experience, I’d imagine you’ll have the skill to do this easily!  

Then it’s all about filling those segments with conversations; this can be tricky and time-consuming due to linguistics.  I once spent a whole week classifying political conversations by had because the language was too complex to be understood by the automated systems.  However, by doing this, I was able to see which politicians supported certain legislation and other wider causes.   

The types of studies that I’ve conducted with more advanced research are varied.  I’ve investigated the change in perception on hosting an international event in Scotland, through to predicting upcoming business trends, and the drivers for whisky consumption.  I’m now developing a brand tracker that will analyse social conversations about key business KPI’s, think purchase intent and value rather than share of voice and buzz.

The Future of Research and Social Media Intelligence 

There are instances when I have only used social media intelligence but the more complex research projects have required multi-methods methodologies and social intelligence has been great at informing other research approaches.

There is no exact science in social media intelligence; it’s all in the hands of the person analysing and more importantly interpreting the data.  My opinion is that this isn’t a job for marketers but someone who has a greater appreciation of behaviour.  We’re not counting likes or clicks; we are uncovering hidden customer motivations.  I see research being more aligned with this type of intelligence but I know the market itself is holding some of this back with 12-month subscriptions for social tools and a lack of context on best practices (but this is something I’m trying to fix).

Over the last couple of years, I have seen a rise in the demand for social media research, and I see this continuing to grow as more organisations go through digital transformation.  There are hidden insights in social media data; you just need to know how to find them.  

I’m going to be talking more about social media intelligence in my upcoming webinar with The ICG on 22nd of September, see you there!

Connect Deeper

Dr Jillian Ney is the UK’s first Dr of Social Media.  She consults for researchers and marketers, advising them on analysing and using social data to tap into customers’ conscious and non-conscious thinking and motivations.  Her proprietary People Science for your Digital Data frameworks guide the way to deeper data analysis, sign up here to be the first to receive her six part analysis framework.