There seems to be a misconception that ‘silos’ are bad and that the widespread adoption of social network applications will eliminate these silos and transform the performance of the organisation. As far as I am aware there is no research evidence at an organisation-wide level to support the benefits of eliminating all silos, probably because no organisation has a definitive list! A couple of decades ago there was a great deal of interest in using social network analysis to understand how knowledge flowed around organisations, with Professor Rob Cross being in the vanguard of this work. Carrying out large scale SNA studies was a significant research challenge and over the last few years I have seen little mention of the approach in conference presentations about the benefits of collaboration. However it is still a powerful technique, as a recent research paper on the link between knowledge sharing connections and employee time saving demonstrates. An important outcome of SNA projects was to be able to identify employees who were the links between groups, especially in geographically diverse organisations.

My particular interest at present is in the use of employees who act as boundary spanners between groups. These boundary spanners can work in both a proactive and reactive way. In a proactive role they may be monitoring the work in a number of different groups and can recognise when a closer link between two or more of these groups would potentially be mutually helpful. As a member of a group they may also take a reactive role when the group realises that it does not have all the expertise it needs and would benefit from working with another group. Employees acting as boundary spanners need not be senior level managers. I have noticed in law firms how support staff and paralegals in a practice can create links to other practices through personal networks as well as through their formal responsibilities. Most professional services firms make a point of giving Associates an opportunity to work in a number of different groups not just to find a ‘best match’ with their skills but also to create these networks.

These boundary spanners usually have a very good awareness of corporate and group cultures, and in the case of multi-national corporates also have the linguistic skills to act as intermediaries. This does not require them to be fluent at a bilingual level but to know enough of a language to perhaps act as a facilitator of the exchange of information rather than just be a passive provider of a contact email address. From an expertise profile perspective these boundary spanners also may not be the most expert in a topic, but as a result may well get overlooked in the technological imperative to locate high-expertise staff. The resulting profiles will not reveal the cultural and linguistic skills that characterise boundary spanners, nor be able to identify replacements when one of these boundary spanners leaves or is given new responsibilities which take them away from making these connections. I suspect that silos rarely start out as silos; they become silos when the personal connections change and links to other groups are broken. Only when a problem arises does the now siloed team find that they are out of contact with groups that could help them. I must emphasis my use of the word ‘groups’. Incidentally, I remain intrigued by companies that seek to promote both the benefits of crowd-sourcing and the importance of the individual expert. Without careful consideration of the incentives and processes these two concepts do not meet in the middle.

If you want to find out more about the roles and skills of boundary managers there is an excellent case study in the September 2014 issue of the Journal of International Business Studies. The interview transcripts on which the paper is based amounted to nearly 1 million words across four quite different multi-national companies. The scale of the research and the value of the outcomes also illustrate why assessing the success of collaboration systems just in terms of levels of adoption is at best superficial.  In particular it will not lead to the types of outcome where boundary spanners are identified and encouraged, and then also used to identify structural issues with the way in which knowledge is exchanged across the organisation.

Martin White