I previously mentioned the role of a Chief Data Officer in the measurement of the ROI of social business initiatives. Here’s an example of how big data could be used to calculate the ROI of an enterprise social network.
Two years ago I had a very interesting discussion with Peter Reiser from Oracle. We both agreed on the fact that the value of an enterprise social network lied in information reuse to speed up task execution, problem solving, exception handling. He told me that there was a “quite simple way” to measure it.
The value of sharing lies in reuse
Let’s take the example of a shared doc. What’s its value ? Zero if no one uses it and if it has no impact on operations. The point is to track how the document is used. If metadata are attached to the document, it’s possible to know in what collaboration spaces it’s been shared, or downloaded. By making the connection between these spaces and the workflows, projects, processes they support, it’s possible to measure the difference between the average time needed for execution and its reduction due to the reuse of the document. According to him (and backed with real experimentation), the difference is often a two digits percentage.
The problem is that the information contained in an enterprise social network is mainly freeform – not linked to any formal container – and it’s nearly impossible to track how its used, contrary to a document that’s downloaded, uploaded, modified, copied but keeps its metadata. Of course it’s possible to measure the number of reshares but the practice has limitations.
1°) Because very few tools allow to “reshare” between separate tools (from ESN to CRM for example) or even link them (what would make it possible to track links) and fewer are the companies that implement such connections when possible.
2°) Because, as long as it’s about freeform information, people are more likely to say “I’ve read that…” or copy/past than link, what makes tracking impossible.
3°) Because of obvious reasons related to right management, one may be allowed to access information he can’t share with a workgroup as it is. Same for a conversation thread : the conclusion can be shared but the conversation that lead to it may be too sensitive.
We can’t track the life of freeform information but can measure its impact
In brief, in regard to freeform information, tracking how something is used by tracking the information itself is illusory. On the other hand it may be possible to correlate the fact people read such source, such content, participated in such communities with operations measurement and find what impacts what and to what extent.
This idea came back to my mind when I read “Big Data: A Revolution That Will Transform How We Live, Work, and Think“.The book mentions a very interesting case that looks quite similar to what can be done with an enterprise social network. It’s a system capable of predicting how capable a student is to answer a given question based on the forum posts he read in the context of online courses, the tests he tried or not and, reciprocally, in the case of a wrong answer, to suggest him relevant content to learn more.
Replace the forum with an enterprise social network, test by the completion of a process or tasks and you’ll have a taste of what would be possible in regards to enterprise social network ROI calculation. Measure the impact of such or such content, community, blog etc on the ability their readers will have to better perform and provide those who struggle with relevant content, information, communities and people.
The system should provide
• A tangible ROI measurement or, at least, of tangible and measurable improvements : the link is made between social activities and measured improvement in operations.
• An analysis of the impact of behaviors and content. Once again, keep in mind that making an impact does not mean being read, liked to retweeted but give others something that provides an actual operational benefit.
• Learning mechanisms that identify what people should read or whom they should contact when they struggle at performing a task.
And even if it looks impossible today I’m sure that we won’t have to wait more than two years before enterprise social network solutions integrate such functionalities.