Well, we are three weeks into the spring semester and I still haven’t reflected on the previous semester. I thought I would put together some thoughts on the transitions I have seen in putting together “Measuring Social” over the past few years. Measuring Social is a class I teach at Carnegie Mellon University designed to throw students into the turbulent waters of analyzing social interactions. It is designed as an experiential learning environment where student’s work on real world issues not simulated or canned problem statements. These are topics that clients, big brand name companies, are dealing with now. Students are expected to learn project management skills, interact with their project sponsor, as well as understand and use both social listening as well as web analytics technologies. They have a semester (4 months) to get up to speed and add value and they need to do it all while juggling a full course load…
To understand diversity of approaches, students are not constrained to a specific structure for their deliverable. I do like to see an understanding of market analysis, best practices, and unique issues specific for the industry. Depending on the project scope, some teams develop social applications while others develop algorithms and/or experimental methodologies. So if a client come in and asks if social can be used to help mitigate online piracy that’s a much different problem and project than understanding how different communities can be leverage for increasing brand awareness. This semester we are excited to be working with Autodesk, NPR, United Way, PNC, Walmart, Expedia and Sony Pictures. At this point in the semester students are in a state of shock trying to figure out how they can accomplish everything over the next 12 weeks. They always do and I’m often surprised and happy with what they develop.
So here are my thoughts on how companies are approaching and trying to Measure Social. Please feel free to contribute…
Social is a team effort not just relegated to the marketing department. There are legal issues (i.e. privacy) associated with collecting and sharing this data. There are large complex data science needs to effectively analyze large pools of unstructured data. IT is required to create and manage a data warehouse and applications that multiple departments can access, etc. etc.
Social data can be a cruel mistress – Not all data is beneficial. It can occasionally lead you down a rat hole of false promises. Evidence based management principles are critical to identify appropriate data based on the questions asked. Strong filters are required to remove duplicative data that tend to skew results. Lastly context is needed and context is hard. We often discuss the 3 Vs – Velocity, Volume and Variety but the 4th V Veracity is (I think) the most challenging.
Data has a shelf life – data diminishes in value the longer you hold onto it without using it. Yes you can do historical analysis but as you move away from when the data was generated you lose context incrementally.
Want to get good at social – Practice, Practice, Practice. Companies that understand and live socially internally tend to do it very well externally. Understanding how to be good at social online is different from being social in a physical setting. There are different norms of behavior, different challenges (attention scarcity, security breaches) and different motivations/incentives.
Data Hoarding – Lets face it when we spend time and effort to collect and organize data, there is a tendency to want to keep it and use it. After all we spent time and effort, so it must have some utility right? Keeping date (hoarding) is one thing; saying that it must have value is another (denial). As we acquire more and more data, the second can cripple an organization.
Metrics vary across departments, campaigns, companies, industries, etc. So be careful about what you chose to analyze otherwise you might spend time and effort to get a metric that means nothing to the group you are reporting to
Most importantly – Communities are much more powerful than groups, divisions or departments. Cultural dynamics and how/why different communities develop lead to an understanding of interaction at a granular level. Companies developing strategies and processes with this understanding use social as a means to engage employees anyway they can are creating more rewarding workplaces.
Care to add to the list?