Social Media Insights: What we learned from our joint experience with Orange
Six months ago we embarked on an experiment with Orange to help them demonstrate the value of social media as a source of insight within its organisation. It was an experiment on both parts as Orange wanted to assess the value of social media research across a number of applications and stretch the manner in which the medium was being analysed. We recently shared our joint experiences at the MR Summit to an audience of insight professionals and have summarised some of the most salient points in this blog.
What had brought us to work together in the first instance was the common belief that social media research is not about measuring or counting social media channels but about getting closer to people by understanding their attitudes, perceptions and motivations, through the unique lens that social media provides.
Orange shared the same view, that social media tools can only take you as far as the quantitative representation of posts by keyword or automated sentiment and that further reading and understanding of what is actually being said can only be done by humans, rather than machines. Indeed, social media monitoring tools are still not capable of understanding all subtleties in language and detecting a sense of irony, slang, sarcasm and tone/register of discourse.
The important point though is that within social media there are spontaneous and unbridled emotions being expressed. They can be accessed through the qualitative analysis of ‘real’ conversations but it takes rigour and effort to extract these conversations. And not every conversation is useful, and most of the time, you need to lose up to 80% of the data, which can seem wasteful. It’s not about breadth though, it’s about depth.
Researching social media is also like having access to a time machine or rewind button that takes you back to consumers’ state of mind at specific points in time. It is not bound by specific time frames as with structured surveys and it doesn't rely on recall. Its flexibility and agility are its key USPs compared to other research techniques.
We applied these principles to make sense of the emotions and perceptions generated throughout the launch of 4G in some of Orange’s core markets. As the roll out of the service was phased, starting with larger urban areas, we were able to follow consumers on their adoption journeys from anticipation to first use, to everyday access. One of the findings was the influence that media and advertising had in educating consumers about a service which they had yet to experience and how expectations were raised by service promises.
"Social data has provided a way to see how opinions about 4G formed, changed and developed over time. It showed us how the journey wasn’t uniform and consistent" said Darren Hanson, Head of Consumer Insight and MR Communications at Orange Group, our partner on this project. He added: "If you are familiar with theories of how people recall and remember experiences you will recognise these peaks and troughs – the intense times in the experience – as having an important role in how people view an overall relationship with a brand. So social data is valuable in this respect because some people share these peak moments on social media, and, if they exist we should be capturing them and using them to improve what we do in the future."
The analysis also helped add colour to Orange’s ongoing customer experience tracker, by providing additional context to the consumer voice through what customers share on social media.
Social media is also a great hypothesis testing tool that can be used to build better traditional surveys and research. It is an efficient means of testing before committing to a methodology like a pre-investigation to make the research better and more relevant. Think an exploratory qualitative phase before the quantitative one.
One thing is for sure is that one should always seek to integrate social media research with existing data and insight sources, if only to provide balance.
One of our key roles as research partner is to educate the audience, and help them confront and understand the data and its limitations. Knowing how to extract information is one thing but the ability to interpret it is another. That’s what makes it research and why it needs skill.
For organisations willing to take the plunge, there are many questions to be answered:
- Is it relevant for me? One of the key limitations is that the data is only there if people are talking about your category.
- Where should it sit within a research team (continuous vs. ad hoc)? Who should own this?
- Does it/should it replace existing methods?
And finally, there are the politics of social media, driven by the ambiguity around the definition of social media research. It's trendy and everyone wants to be associated with it. But social media research is about the ‘consumer perspective’ not the numbers and that's where the tension often lies in the organisation.
But according to Hanson: "If you are open and willing to experiment perhaps you will also find out that the real value in this data is in understanding the conversation not just in the numbers."