By Tom O’Brien (CMO) & David Rabjohns (CEO); MotiveQuest LLC
Have you ever tried to fix your own car?
I have, because I have the tools. Even with the tools it ended with innumerable parts spread across the garage floor and a call for help to my uncle the mechanic. Social listening research is much more complex than fixing your own car – but this doesn’t stop people from amassing a toolkit and embarking on a self-guided listening project.
Thankfully over the last 18 months there has been a mainstreaming of social media ideas, marketing, monitoring and research among the big brand owners of the Fortune 500. Along with this mainstreaming comes increased understanding of the different kinds of tools and expertise available for hire.
In particular, as this market matures, companies are beginning to understand the distinction between “social monitoring” and “social research”. Increasingly clients are setting up a dashboard for daily activities and partnering with an expert for more in depth analysis.
Right now, we are asked almost daily by new and existing clients to discuss the distinction between social monitoring and social research.
Social Monitoring: Tracking online brand mentions on a daily basis for PR, brand protection, operations and customer service outreach & engagement.
Social Research: Analyzing naturally occurring online categories of conversation to better understand why people do what they do, the role of brands in their lives and the product, branding and communications implications for brand owners.
Social monitoring is the clipping service, 800-number and suggestion box of the 21st century all rolled into one. Powerful software tools to track online mentions of your brand score, then for influence, relevance, sentiment and necessary action and feed all of that into a workflow for PR, Operations and Customer Service. We believe that every major brand should be using these tools aggressively. (Marshall Sponder over at WebMetricsGuru and Nathan Gilliatt over at the The Net-Savvy Executive are doing a great job of covering the brand monitoring/dashboard space. Minor brands should be monitoring too, but perhaps by using the free services– more on that at the Net savvy Executive http://net-savvy.com/executive/tools/monitoring-social-media-before-you-have-a-bud.html .)
Monitoring Commentary: The best companies in the social monitoring business have put together very powerful tools for monitoring, scoring, outreach and reporting. As with most powerful tools, they are non-trivial to implement. There are a couple of dimensions of complexity to consider including:
- Brand Mentions: Including the data you want and excluding the data you don’t want ranges from easy (simple, unambiguous name like MotiveQuest) to extremely hard. Consider the rental car category. The leading brands are Dollar, National, Budget, Enterprise and Hertz. 4 of the 5 are very challenging terms to disambiguate.
- Context: Sometimes the brand mentions are a very small part of the overall conversation. Food brand mentions tend to be less than 5% of the conversation. In cars it is about 40%. In both cases you lose half or more of all the relevant category conversation by just collecting brand mentions.
- Linguistic Coding: We did a project for Visa looking at the Beijing Olympics. We had to build a linguistic string with more than 250 arguments to include Visa the credit card brand and exclude Visa the travel document, both of which were highly co-related with the Beijing Olympics.
- Data Sources: There appears to be significant discrepancies in data coverage for the different tools. Also, you need to consider the data source to know what to do. Twitter is different from a blog which is also different from a forum.
- Scoring: There also seems to be significant discrepancies in sentiment scoring across the tools.
- Implementation: If you are going to maximize the benefit of implementing a social monitoring tool, you are going to need to have cross-departmental involvement from your marketing, PR, customer service, operations, legal and product teams. This is a very significant challenge.
Social research: Is purely observational. Gather and analyze millions of online conversations from the web to build a model of how the world works. Dig well beyond buzz, topics & sentiment and look for the underlying human issues, motivations, needs and drivers within a category. Then analyze the competitive dynamics within the category among competing brands. The goal is to help clients understand what to do next by having a deeper understanding of what matters most within their category and how their brand can relate to or own that.
This type of research is hard to do without a broad sophisticated toolkit and a team of experienced strategists. One of the fundamental challenges in analyzing large amounts of unstructured text data is that you simply can’t make sense of it in any sort of manual fashion. Sophisticated software tools are the answer, but how those tools are developed and deployed makes a difference.
- Language is fluid over time and across categories. Tools must be too. In our opinion, tools need to be parameter driven allowing the strategist to adjust the linguistic model and other parameters for the category and project at hand.
- Simple measures (buzz, topics & sentiment) are not useful for understanding why people do what they do.
- Understanding requires more sophisticated tools. For example, it is helpful to be able to identify what are the biggest, naturally occurring peaks of passion in a category, explore the natural word associations in the dialog, measure the motivations beneath the dialog and understand which conversation drives recommendations in the real world.
- There are many tools available, but the tools used and the order of use is very project dependent.
- The tools are just tools and the work is only as good as the person using the tools. The team doing the work, just as in any other field of research, gets better the more they explore the universe of data they are working with. They also need to connect the business needs to the data in order to solve real business problems.
- The output must have specific actionable ideas, insights and recommendations for the clients against the business problem. These tend to be in the area of branding, communication, innovation and measurement.
So what is next?
As the world of listening research matures, we believe that it will begin to more and more closely resemble the world of traditional research. There will, of course, still be listening dashboards (that complement the traditional clipping services and quick online research tools), there will be social listening specialists that provide full service answers to tough business questions, and there will be sophisticated listening segmentation companies that are a mirror to the old world asking segmentation companies.
Agencies and clients will continue to try to do a lot of the listening themselves but will increasingly realize, just as with more traditional research, a specialized area with specialized tools and expertise, listening research is often better outsourced to experts than DIY.
OK has anybody got a wrench?