Selecting the Right Social Data Sources Means Creativity and Questioning

We’ve been asked if following a netnographic methodology means that you might miss out on legitimate data sources?  Here’s what I have to say about selecting the right social data sources… It’s all about getting creative and questioning the data source.  

Last week our article on netnography created a bit of a stir behind the scenes at The Social Intelligence Lab HQ.  There were emails praising for sending a little scientific help to those self-taught practitioners, and there were emails saying the netnography methodology constrained the selection of data sources too much that you might miss critical mentions.

It got us all talking about selecting the right data sources.

Quality Data: Clean Data

It’s been a long time since I ran social data analysis that focused on gathering brand mentions to monitor brand health.  Simply wanting to get as much data as possible.

Instead, I usually answer specific questions related to a business need.  For instance, why are more people leaving employment early to care for elderly parents?

This requires a bit of work to get high quality and clean data.

For as long as I’ve been analysing social data, I’ve based my work on netnographic approaches.  This means I get creative and pay attention to the data sources, and question specific sources.

Selecting the Right Data Source

There are many published articles on netnography.  One of the things that I’ve noticed and probably why this conversation about data sources has come up is that it looks like in netnography you choose the data source at the start of the process.

For example, if you’re looking at financial matters, you might opt to analyse Martin Lewis’s Money Saving Expert Forum.  If you’re looking to understand diabetes, you might choose a forum that specifically deals with diabetes.

The data source selection seems very literal and doesn’t leave room for spontaneous serendipitous conversations.

People might be on Money Saving Expert as that is a passion point for them, but it doesn’t mean money, saving, debts or money saving tips are the only things they will discuss.  For all we know, there may be a whole host of diabetes conversation happening there.

Social listening has the power to bring back much of the conversations happening about diabetes from across the web.  More data equals more insight, right?

Well, that’s not always the case.  The more data you have the wider the discussion range and the more holes you might find in it as Mela and Moorman argue in their Harvard Business Review article.

We all know there is a lot of junk that can come back when setting up new social listening queries.  So, while selecting data sources up front can be to narrow, taking everything that comes back from social listening queries can be too wide.

This is why we cleanse the data.  It takes a long time to cleanse that data and refine the search query.  But we shouldn’t take the data at face value. In cleansing, we should be questioning the data source of the mention.  Is the source legitimate?

Questioning the Data Source

Let’s say you found a lot of diabetes conversations on a Harley Davidson forum.  Seems a bit odd, right?

Before using the forum, you need to check why the conversation has been gathered.  Is there really a legitimate conversation about diabetes? The chances are that the Harley Davison fans might have diabetes and are talking to their friends about their condition.

While it seems like diabetes might not fit with motorbikes, we need to check the legitimacy of the conversations.

Netnography isn’t about selecting the data sources up front, it’s about questioning the source and ensuring the legitimacy of the conversations.

In doing this, it raises another important question, should we be more transparent about the data sources we use?

Being Transparent About Data Sources

When selecting a social listening tool, the data sources and quality of the sources are really important considerations.  But, when it comes to analysing and reporting insight, there can be a big hole in transparency about the data sources used.

Should we be more transparent around where the conversations happened and where the insight came from?

I say, yes.  In being more transparent, it could also help the business better understand what information can be used for key decisions.

‍