This session focuses on using data to understand how social networks work within the enterprise and how to optimize their activities.
[These are my quick notes, complete with (what I hope is no more than) the occasional typo and grammatical error. Please excuse those. Thanks!
From time to time, I’ll insert my own editorial comments – exercising the prerogatives of the blogger. I’ll show those in brackets. ]
- Rob Howard: This session isn’t just about collecting metrics. It’s also about superimposing a human layer over the data. The purpose is to help people find the information they need much more quickly. This creates very efficient collaborative filtering.
- Rob Howard: Currently, each system/silo has its own analytics. However, the enterprise needs to interconnect these systems and then analyze performance across the enterprise.
- Sameer Patel: The problem with metrics is that we gather information after the fact, when we can’t do anything about it. It’s better to get data in real time and be able to do course corrections before it is too late.
- Tim Young: Socialcast thinks that the future lies in understanding a company’s social graph/ informal organization NOT in understanding your official org chart.
- Social network analysis is new. Unfortunately, the old business intelligence systems can’t gather and analyze network information.
- Rob Howard: In the past, we had little or no useful data. Now we have a flood of data and organizations are struggling to find the best way to understand that data.
- Audience Q: This shift from analyzing systems to looking at individuals is a little creepy. What are the privacy implications of this?
- Leslie Fine: They set up the system so that they look at individual information only to the extent it is given voluntarily.
- Sameer Patel: Ideally, you’d have a global opt-in model. The organization analyses what you choose to do on the system.
- Rob Howard: The privacy issues are still in flux. Vendors are watching Facebook carefully, as well as European privacy law developments. There are two different types of data that can be collected:
- Explicit data – e.g. the content I contribute
- Inferred data – e.g. when do I sign in and out; where do I work, etc.
- Leslie Fine: Users need to get value out of it. The key value is that users believe that these tools now give them a better way for them to communicate with management. When front-line employees feel like they have a voice that is heard by management, that is enormously powerful.
- What’s easy and what’s hard about analytics?
- Rob Howard: It’s now very easy to gather metrics. The hard part is analyzing the collected data.
- Tim Young; We started by using simplistic success metrics (e.g., how many visits, how many activities occurred, etc.) but the more useful information is in using the social graph to measure IMPACT.
- Sameer Patel: The difficult problem is that there is a lot of bad news that we don’t know how to surface, discuss and resolve. Currently, the current approach to analyze and resolve is ad hoc – you sense a problem, form a committee and make recommendations. However, since data is so dynamic, these stodgy systems don’t respond fast enough or effectively.
- Leslie Fine: We can’t create a system that acts as “a digital gotcha.” This isn’t about focusing problems, but rather about getting to solutions faster.
- Audience Q: How do you bridge the gap between executives who are used to judging the value of a tool based on basic web analytics (e.g., page views, number of visits, etc.) and the new social analytics (e.g., who visited, what did they think about, what did they do, etc.)
- Rob Howard: The social analytics help you understand individual behaviors (especially the behaviors of influencers within the network) and then build strategies for expanding or modifying the behaviors of users.
- Tim Young: One of his media customers uses social analytics to identify what their own employees are discussing, what topics are surfacing and where. They then can help share solutions from one part of the enterprise with another.
- Audience Q: What’s the difference between collecting data this way versus the data that is collected when we use loyalty cards from grocery stores or drugstores.
- Oliver Marks: The difference lies in the nature of the implied contract between the store and the customer. In that case, the customer voluntarily provides information on their purchasing habits in exchange for promised discounts. By contrast, within the enterprise, the stakes seem much higher. Your job could be on the line!
- Leslie Fine: One frequent complaint is that enterprises have more data than insight. We need to find ways to help enterprises to achieve insight.
- Rob Howard: These analytics don’t work in just one direction. It isn’t only for the benefit of the executives. Users can benefit from the data to understand what behaviors work within the network and how to find what they need.
- Where do you start with social data?
- Tim Young: There needs to be some level of social activity of a decent period of time 9eg. 60-90 days) before you have enough data to analyze. If you don’t have an explicit strategic concern, it could be interesting to see what mysteries the data expose.
- Leslie Fine: The idea isn’t to look at the raw data and then figure out what to do. You need to start with a business problem/strategy and then mine the data for answers.
- What are the big challenges for the next 5 years?
- Sameer Patel: The big issue will be agility. This includes the ability to answer customer concerns and fluid market conditions very quickly. There needs to be more support for faster and better decisionmaking. He says that today this whole area is high stress, but soon it will be so common that it will be considered a utility.
- Leslie Fine: With more time and data, we’ll be able to solve problems faster.
- Rob Howard: A key to the future is to use analytics as a tool to unlock certain behavior patterns: information hoarding,network habits. The business value is that this will help us bring information to us rather than having to go out to find information.
- Tim Young: There two possible future paths. If you are in a B2B enterprise, you’ll tend to operate more collaboratively. If you are in a B2C enterprise with lots of customers, you need to focus on the coordinated network. You’ll fiind the most value in using traditional BI to find more quickly the answers customers need.
- Audience Q: Please give some concrete examples of companies using social analytics.
- Leslie Fine: A particular game manufacturer knows that there’s no point in marketing great games (because they sell themselves) or rotten games (because no amount of lipstick will improve a pig). However, clever marketing can improve the sales of games in the middle. This manufacturer used social analytics to identify what games really fell in that middle category and what type of marketing would be most effective.
- Tim Young: A major retailer used social analytics to obtain insights from their employees — a small group of employees in the midwest were talking about a particular book (the Twilight series) and bought the book distribution rights before Walmart and the others. They won exclusive rights for the first two quarters. This translated to a huge financial win. And this trend data came from internal social analytics, not external marketing trend data.
- Rob Howard: The national breast cancer society uses social analytics to understand the conversations within their community to target the application of their limited resources to the problems that are most pressing to their community.