A J Johnson AJ is Vice President of Ipsos Open Thinking Exchange, an organization dedicated to ‘opening the exchange of ideas and information between marketers and their consumers’. Read the full biography here

Are you listening?

or will social media monitoring replace the need for traditional market research techniques?

By A J Johnson - 23rd July, 2010

The world of market research has changed, and as an industry we strive to adapt and shape our future. The Internet has radically altered data collection over the last decade, and the second generation of web is making even more impact on our research techniques. It is perhaps ideal timing, as participants’ patience for completing long, relentless and often poorly designed surveys is dwindling, that the social Internet enables us to listen to participants without asking questions. Listening to what people say on the Internet and making sense of it for research purposes sounds like the golden ticket. Welcome to the world of social media research!

I recently heard an anecdote about a train company who sent an engineer to fix the heating on a train before the passenger who Tweeted reached their destination

First to appear in market research was social media monitoring often called ‘buzz monitoring’, which ‘crawls’ websites, blogs, micro blogs and social networks looking for discussion around brands, products or services. Technology aggregates this buzz (the volume, location, source and timing), applies sentiment analysis and stamps an ‘influence’ index, before delivering results through an interactive web portal. Most research agencies take this basic deliverable and develop more insightful research reporting around brand health and strategy for their clients. There are now numerous technologies of this type (some free), and they have widespread use across research agencies and clients.

In Customer Loyalty, many are extending these listening technologies to trigger instant reaction to manage individual relationships. BT now has Twitter agents to respond to customer feedback, and I recently heard an anecdote about a train company who sent an engineer to fix the heating on a train before the passenger who Tweeted reached their destination. The ability to intercept negative feedback and potentially turn it into a positive is a powerful CRM tool, even if it is not always welcomed by those at the source of the content.

More recently we have seen the concept of social media research that looks to more closely match the traditional research process, using social media content as the sample source. Clear research objectives are set (although they are often more dynamic as the project unfolds), and sampling and weighting theory is implemented. A strict cleaning process is applied to all comments to ensure accuracy of sentiment and correct for duplication, bots, retweets, spam etc. Analysis of the data is a mix of qualitative coding, quantitative reporting and text analytics. Traditional research measurement strategies are even applied to enable familiar Likert scale scoring, and normative data are established to compare social media research projects.

But of course it is important to consider the issues when undertaking new methodologies. Typing Wayne Rooney into a free demo of a Twitter sentiment analysis application for example, returned analysis that was 60% negative. Not surprising perhaps given England’s dismal World Cup performance. But on looking at the results more closely it was even worse for Rooney. For example the tweet ‘“Wayne Rooney is the best footballer in the world”, said Wayne Rooney’s mum’ was incorrectly returned as a comment with positive sentiment. Technology is improving, yet still fallible to the quirks of the English language.

Just because someone makes a comment on the web, does it make it fair game to use this without their permission?

A recent report from FreshNetworks looked at a number of leading social media monitoring tools, and discovered big differences in factors such as coverage, sentiment, location, volume and data latency. The findings suggest that to make the most of these tools we must understand these differences, and have access to multiple tools to gain more reliable analysis depending on the particular project requirement.

So what do consumers think of social media monitoring and analysis? We have already seen the backlash from consumers seeing Google as ‘Big Brother’ when they launched Google Earth, and the recent revision of Facebook’s privacy settings following public pressure. Just because someone makes a comment on the web, does it make it fair game to use this without their permission? Of course they sign the website’s T&Cs, but do not necessarily expect the subsequent commercial use of their data by third-parties.

A recent Vovici survey amongst c.400 online panellists (an audience perhaps more open to online monitoring) clearly highlighted the public’s issues with privacy; 40% say they are either ‘very’ or ‘extremely’ concerned. While 85% would prefer that researchers asked their permission before reporting their comments, and 19% do not want their comments on the Internet quoted at all. A surprising finding was that only 37% thought that search engines analyse public Internet discussion, suggesting they are not yet fully aware of what currently goes on!

So will the use of social media monitoring replace the need for traditional market research techniques? Absolutely not! But it does provide yet another valuable tool in the research toolbox. Perhaps most important is to understand that these SM tools will not replace the need for the human, as technology will only go part of the way in the generation of insight. As for privacy, the industry must behave ethically and be transparent at all times, and carefully monitor evolving discussions and any legislation that could affect what we do.

A J Johnson

Comments on this article

Want to share your thoughts...?

Want to share your thoughts?

NOTE: Please note that this board is moderated, and comments are published at the discretion of the site owner.