Blogging was where we began, and how we built our company so we have preserved this archive to show how our thinking developed over a decade of developing the use of social technology inside organisations

Data-driven business improvement


Following on from my talk last week about the future of social media monitoring, and how I think we are moving towards a more holistic view of social business intelligence, I wanted to focus on the internal realm of the enterprise, and how these ideas can help drive behaviour change and business improvement, which was the focus on my keynote talk at the E2.0 Conference today:

You can watch the video of this talk here

Early adopters of internal social business or E2.0 technologies are now sitting on large flows and stores of social data that probably contain some very important insights about the organisation. In the past, some software vendors have tried to use email stores as datasets to mine for social network analysis, based on the idea that this could detect natural networks among employees and also hidden sources of expertise. It turns out this is not very reliable, which is why it has not caught on, and also carries privacy complications, because people tend to assume their email is personal. But since the advent of social networks, activity streams, content bookmarking / ‘likes’, and publish/subscribe technologies such as RSS, we now have a growing body of elective, freely volunteered network information, which is probably a lot more reliable than that inferred from people sending emails to each other. But very few companies are using this to derive insights about their business.

Social business inside the enterprise is not only about direct collaboration, but also ambient sharing, dynamic signals and various other types of indirect or weak ties between people in networks. Especially in dispersed teams and business units, these signals are vital to maintaining a sense of common purpose and connection. But these connections are also key to realising Dave Gray’s notion of the connected company, based on small ‘podular’ teams, and make up the network fabric that enables them to coordinate.

What can we learn from the use of analytics in consumer social media? First, that aggregate measures (buzz, sentiment, etc) are necessary but not sufficient. In many cases, it is more important to drill down to specific events or issues and act on them in real-time. Analytics without action is just a rear-view mirror.

Increasingly, companies are looking to the emerging field of big data analytics to enable this kind of insight. There is a lot of innovation in this space, but what matters is not how big your data is, or whether you are using Hadoop, Mapreduce, etc, but rather what you do with the insights. Dan Catt recently wrote a great piece arguing that hiring ‘Quants’ or Big Data Scientists is only useful if you are immersed in your data, can feel its rhythm, and therefore know what it really means. But, as Jeff Hammerbacher of Cloudera explains in his interview with Businessweek, there is a whole world of potential benefit to business and society in this emerging field.

We can also learn some lessons from sport, where analytics has led to rapid improvements in Forumla 1, baseball and soccer. In these fields, analytics have also learned how to refine the stats they track in order to draw the correct conclusions, as Simon Kuper writes in a fascinating piece for the FT.

So, how we do begin using Social Business Intelligence to drive business performance in the enterprise?

First, instead of creating microcosms of social data to study outside the working context, we should try to immerse ourselves in the ecosystem that includes both internal conversations and the data they produce, ideally also correlating them to find common themes and issues.

Second, one thing we need to change relating to current practice in social media monitoring is to spread the insights, rather than leave them locked up in dashboards and reports in PR or marketing.

Reports are a poor way of influencing behaviour, and suffer a time-lag that means their interpretation is always out of the context and flow of work. Instead, I think the notion of activity stream should be extended to include feeds and flows of potentially actionable insights derived from analytics, as well as automated machine data and events (‘server has gone down’, ‘factory production is on hold’, etc) and human-generated status updates, thoughts, ideas and microblogging.

This connective framework for information would then enable various people in the firm to start to make sense of the issues, actions and questions flowing around the firm, based on the idea that ‘many eyes’ and community interpretation can scale better than a single person or department reading reports. But crucially, this needs to be allied with an action framework that encourages individuals to pick items out of the flow and say “hey – we need to act on this, and I know how.” This is not only a more efficient way of interpreting and dealing with large volumes of data, it is also a great way to help everybody in the firm get closer to clients, and to encourage a culture of ownership and action.

Nothing motivates people better than objective real-time feedback about how they are doing. Open data in the enterprise can therefore play an important role in stimulating productivity and using customer insight to keep people on track.

This opens up a whole new practice area that combines technology, data and behaviour change, which we are loosely calling Social Business Intelligence. It has the potential to create the characteristics of evolutionary improvement in networked organisations, and I think over the long-term, will help us create more direct, low cost mechanisms to keep people focused on what matters in their work, rather than wait for management permission or edict.

For companies interested in taking the first steps towards this, there are a few simple steps that can begin the show its potential:

  • Apply analytics to internal social data
  • Correlate with external social listening efforts
  • Create a subscribe model for personal filtered activity streams
  • Create simple ways for people to claim actions
  • Encourage ownership and action by all

Comments are closed.