Social listening programmes continue to fall short of expectations, but the critical issues aren’t with the technology or data itself — rather a series of mistakes during the analysis process.
Struggling to make your social data analysis initiatives work? Can’t prove that social intelligence drives business value? Finding it hard to communicate social insight across your organisation?
It would appear that you’re not alone.
Forrester reported that bad data gathering, poor sentiment tools or unreliable reporting and analytics as key complains in social listening failing to hit the mark.
But, rather than blame the technology or the data, there could be another explanation for your social insights falling sort of expectations. Ever thought about errors in the data analysis process?
Your Team Mission Doesn’t Set You Up for Success
In many organisations, social listening is set up as a reporting and analytics function. In this structure, we often find that the primary role of the department is reactive reporting.
While you could argue that this is the more traditional approach to reporting, if social listening departments are only producing reactive reports, they’ll find limited value and an inability to clearly demonstrate impact.
Social listening departments that operate as a simple analytics and reporting function can quickly find themselves in a never-ending cycle where the true power of social data analysis is never fully realised.
As Ben Ellis, Senior Research Consultant at Brandwatch explains “leaving social listening as a reactive function means that the practice and the practitioners are stuck in a box where they’re only expected to be functional”. Meaning that social listening does not have longevity and that the corporate culture restricts researchers and their work to functional rather than purposeful.
Your Team Structure Doesn’t Set You Up for Success
When establishing social listening and social intelligence functions we see two common mistakes in team structure – prioritising technological solutions and not hiring diverse skill sets.
In their Effective Social Listening whitepaper, Converseon discuss the difficult choices that need to be made when choosing technology: choose an automated analysis for speed but sacrifice reliability and precision, or, choose human coding on top of the data but lose the speed and scale that software can provide.
Fortunately, they argue that you can have both. …Not by cutting humans out of the analysis efforts but by using technologies to scale human intelligence.
As word of warning, it’s important to know what you want to analyse and use the insights for before selecting a tool otherwise you’ll be guided by their functionality rather than your own objectives.
And it’s important that social listening teams aren’t just limited to analysts. Teams need professionals with business and relationship skills — people who can help to frame the right question and communicate the results effectively.
Social data analysis is most effective when paired with human insight, and these professionals need to have strong functional domain knowledge. Eric Michelson, Social Intelligence Expert at health insurer Aetna, agrees that social intelligence requires subject matter experts who are skilled at parsing and interpreting subtle emotions, and who can correct analysis for the politics of online communities.
You’re Focusing on Data Volume Rather Than Targeting Relevancy
When it comes to data, organisations hold the mistaken notion that they should gather and add more data instead of targeting data and queries for relevancy. It’s easy to fall into the trap of thinking
“maybe there’s something hidden in there”, rather than asking “do we have the right data?”
Experts say, identifying the right data sources and writing targeted queries is critical for success.
Preriit Souda, Director of Data Science and Strategic Insights at PSA Consultants, explains that while you sometimes need to go broad and play around to find the right data of interest you must to get to a place with a well-defined area of inquiry to save time, money and processing efforts or to “stop boiling an ocean when you simply need to boil a kettle”.
You’re Focusing on Channels You Engage Customers, Rather Than Rich Insight Sources
When it comes to data priority, organisations misguidedly focus too much of their listening efforts on popular social networking sites like Facebook, Instagram and Twitter because that’s where they post content and engage with customers.
Inevitably, neglecting other data sources that can have a greater value to their business success.
Sources like Reddit, Quora, reviews, forums, blogs or even comments on newspaper articles can hold far greater customer insight than focusing too much attention on traditional social networking sites.
Eric Michelson, even goes so far to suggest that forum discussions offer the needed context through extended discussions that are easily as valuable and many times more dynamic than moderated panels.
He argues that while forum analysis is more time consuming, manual work, they are the richest source of information, and should be a go-to source as forum members are highly knowledgeable, have long experiences and are deeply engaged.
“They often know more about products, services and brand relationships from some of the top corporate insiders from their perspective outside the bubble”.
However, Rob Sullivan, Social Analyst at the agency Social Chain adds that choosing the right social data source is more industry dependent:
“if you are Converse and want to assess a product launch, Twitter is ideal because people will gossip all day about the Miley or Tyler release. But, if you’re a B2B software provider, maybe not! Twitter is still lacking in capturing the silent majority…”
Which also leads to using social data in isolation.
You’re Not Integrating Social Data with Other Data or Research Sources
Organisations are failing to capitalise on social listening initiatives because they are looking at social media data as a data silo on its own.
The power of social media intelligence is in the immediacy and relevancy of the analysis, but it is ultimately a single data source.
And, a data source where direct questions cannot be asked directly to consumers. Analysts are beholden to collecting data that is organically generated, and located in the public domain, which inevitably leads to ‘context holes’ in the data.
Kelly McKnight, Head of Culture and Trends at the research agency Join the Dots believes that:
“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”.
Be that the integration of surveys, research interviews, website analytics data, sales data and other internal and external sources. It’s about finding out what the social data doesn’t tell you and finding other solutions to fill those context holes.
You’re Reporting the Wrong Metrics or Insights
It’s common for less mature social listening teams to unknowingly report the wrong metrics and insights to different teams and departments.
But, “social metrics and insights in one form don’t meet the needs of everyone across the organisation”.
Teams have failed to realise that business units have unique needs and need different metrics and varying degrees of intelligence and insight reported.
For example, the metrics used to report content success for marketing teams are very different to those needed to convey customer insight to research teams or the insights that are needed to create new products or reduce customer service costs.
Failing to tailor reporting leads to insights that do not convey action where reports sit in executive inboxes unused and unwanted. There’s no demonstrable business value if there is no action that can be taken.
You Don’t Realise That You Don’t Need to Report Everything
Analysing social data properly isn’t easy, and it’s time consuming. One of the biggest social listening myths going that you can click a button to get insight.
Because so much work went into analysing and interpreting the data, there is a tendency to analyse and report everything instead of simply answering the original question set.
To me, it feels a bit like solving a math problem at school, you need to show how you got to your answer. But, unfortunately, you don’t get more marks for showing your working when it comes to giving a busy marketer the content ideas that they should be working with or the best fit idea for a new innovation. They don’t have time to read a 50-page report to get a simple answer!
At last year’s Social Intelligence World, this topic came up in conversation at the Getting Buy In When No One Is Listening fireside chat with Derek Lumb, the Global Social Media Lead at E.ON.
The discussions highlighted that there was a tendency for everyone reporting social insights to be driven to demonstrate their value. But in a shock turn of events, the way in which value was being demonstrated was having exactly the opposite effect. For instance:
- – There was too much analysis being reported and not enough insight that can be actioned.
- – There was insight in there that wasn’t asked for in the original brief which is great insight but not for that department and it never reaches the right people.
- – There were 50, 60, 100+ page reports that said absolutely everything about what was being discussed online but it was hard to find the answers to the original question set of the research.
Understanding the objective and how the insight is going to be used are critical first steps in every social intelligence project – they ultimately guide everything that needs to be completed for the research and communicated in the research findings.
As the social intelligence industry matures and more organisations embrace social data across the enterprise, getting data quality and relevancy will increase in importance. Maybe it’s time to explore the impact and quality of your social insights process to truly exceed expectations.
Over the last 10 years, I have been documenting and testing the most effective process for social data analysis – download my guide below.