New Killer Star

Social interactions continue to attract conversations, because they are introducing shifts the way people access information, communicate and work. This shift did happen due deregulation of the financial markets in the 80’s, driven by development of technological infrastructures (telecommunication and information systems), combined with computer power capable of the high speed algorithms processing required to handle the complexity of financial transactions, not because millennium generation and others want it and are eager to be connected.

Every time a new concept in introduced in the business community hype is created and everybody wants to follow, copy and use it. Social business is one of the most discussed themes these days. Facebook bought Instagram and the .com bubble is bringing cream full of sugar to our lips (how loved the late 80’s!). If you want to start your social journey it can be useful take a peek to this social business book list posted by Dion Hinchcliffe. Or not.

It’s useless to start you social journey if your social structure is not aligned with the process bandwidth played. But the process bandwidth should reflect the knowledge dimensions that should support the outcomes customers are willing to be achieved.

How a star can kill your social efforts

Forget for now the book list. The first thing, as a manager you need to understand is the network type that exist inside of your company. A method to carry on this assessment can be found in this previous post. Process Mining can be one of the tools you can use to find out the type of social networks that exist in reality. Organizational charts, do not reflect the way people interact.

Discovering the types of networks that exist is the first step to carry on the assessment. you need to understand the communication flows patterns. An organization is the result of time-sharing social practices that work through flows. And flows (represented in social networks by arcs between nodes that represent the actors or process participants) are the activity sequence between participants. In order words you need to find the dominant social structure.

The assessment will show two important attributes:

• The nodes – the process participants. The way the network of nodes is highly or loosely coupled or of the roles of interaction and propagation involved.
• The way people are organized around knowledge type from Tacit to Explicit in different boundaries (inside the company and outside the company). Tacit knowledge is related with judgment, intuition, reasoning, in order words with unstructured / dynamic /adaptive / ad-hoc (I put all the possibilities in order not harm beliefs), annual/industry reports ;  Explicit knowledge is related with procedures – structured processes, companies policies, industry culture.

Hence, without further explanation, it’s pretty clear that if your company plays more unstructured process instances and the discovered network is highly coupled around a hero or a star (decisions are centralized) knowledge will not be spreading and you have a very inflexible company.

Enterprise architecture is about alignment. Encompasses the entire enterprise. There is no meaning if you want to copy CISCO culture, providing autonomy, empowerment, social tools, whatever,  if your network blocks flow communication.

Network structures

John P Scott, points some basic network structures that you can use to carry the assessment. I will explore it regarding the consequences of knowledge flow and alignment regarding the process types that in theory are executed. For social networks discovery it’s necessary you define the discovery scope first (choose a process and start mining; in some organizations like healthcare providers, the “process” start when you are admitted in a hospital until you leave, thus mine for function). Other way to define the scope is mining the same process, comparing how social networks are structured across different countries or business units. This last one can be a important, as Isabel Ferreira, CEO of  Best Bank said recently

One of the major challenges is to expect the same performance in different countries. We provide the same tools, the same goals to be achieved, still there are differences in performance. The differences come from the way people work and communicate together. A manager needs to understand what is wrong regarding how people are engaged and make adjustments. Business Intelligence will tell you nothing how people are connected, just shows you if things are right or wrong, because it’s not process oriented.

Management style shapes the network configuration and the outcomes. I’ve seen some high achievers informal networks capable to deliver top results.

Network patterns

• Star networks: Characterized by being strong central and weak distributed. These kind of networks are typical of structured processes. Every decision is approved and people play a well defined role and hierarchy is imposed as formal authority. People play the same activity one after the other and the very same result is achieved on every process instance.
• Star with multiple core networks: Characterized by being strong central and strong distributed. Here, someone is responsible to orchestrating others, that have their own links to achieve an outcome. Feedback is shared, but it’s “sent to the network star”. Relationships are maintained by authority like in start networks, but communication between peripheral members is maintained and it’s improved if peripheral members maintain a close connection.
• Multiple core networks: Characterized by being weak central and strong distributed. Typical of adaptive processes. As the process progress, resources join the quest to find a solution to achieve the outcome. Feedback is shared, but is not centralized. These networks are formed based on tacit knowledge and trust. Finding people inside big organizations with plenty and the right Tacit knowledge can be difficult, thus people will stick to the ones they are acquainted (this can be a challenge if we want to involve the right people but we don’t know where to find it); trust plays an important attribute. Trust is one attribute of social interaction foundation. Looking to a broader level, companies that do not deliver a “relationship based on trust” with it’s customers will be in trouble if they want to engage with them. Today this is impossible. Going back to people domain, trust is important, because the others believe you will deliver independently the circumstances that change on run time mode.
• Loosely coupled networks: Characterized by being weak central and weak distributed. You can find this kind of networks on emergency scenarios, or if you are analyzing a company of heroes (meaning people work on a “last minute basis”).

Performance implications

Looking to the above patterns, looking the way information flows you would be tempted to choose the Multiple Core network and forget the Start type, because will be related with poor performance. But that does not hold true.

Social networks must be aligned with the process type played (structured, adaptive to emergency). And the process type played require different knowledge forms. There is an extra challenge , when the end-to-end process is blended regarding type, I mean, part is structured, part is not. Discovering social networks of this process types can be done using Process Mining filters that can shows you how people connect across the different parts of the process played according the activity sequence that defines it’s nature.

The key point is to align the network type with the process type.

• Star networks: These kind of networks are useful, contrary of what you may think, but only on highly automated, repeated processes. People play it’s role alone and report only to other that is responsible for the work is being carried and aggregate and control all the information (for example a contact center coordinator). These networks are inappropriate to knowledge sharing and dissemination, are tailored to extract the maximum of human resources available to perform. Information is analyze by the star or the starts of the network to find best practices and optimize how the work is carried.

• Star with multiple core networks: These kind of networks are useful, contrary of what you may think, but only on highly automated, repeated processes. People play it’s role alone and report only to other that is responsible for the work is being carried and aggregate and control all the information (for example a contact center coordinator). These networks are inappropriate to knowledge sharing and dissemination, are tailored to extract the maximum of human resources available to perform. Information is analyzed by the star or the starts of the network to find best practices and optimize how the work is carried.
• Multiple core networks: This is domain of knowledge intensive exchange. This is where complexity and unpredictability plays a role. Adaptation occurs from the feedback loops, and decisions are made according people expertize and experience. This is where new alternatives are explored to reach an outcome. Some will turn to be patterns (solutions) other will transform to best practices and will be lately automated on a predefined manner. On this kind of networks it can occur performance lags when network is formed if people are not acquainted (makes me remember my role in W3C and WFMC groups I’m working with) or due knowledge leveling (everyone knows the concepts meaning).
• Loosely coupled networks These networks are found in Emergency or Chaotic process types. Forget the Red books and Emergency plans. Under these circumstances people act first, measure the consequence and adjust until it’s possible to resume under more controlled or predictable circumstances, that can be played according the knowledge intensity (more intense adaptive process type, less intense structured). When the process resumes to new process type, it’s expected a shift in the network structure.

The wrong way to Hollywood:

Before you start your social journey, because you feel you are not engaged with your customers, you do not understand what they are talking about you, your process architecture is not designed to handle complains from Twitter, or internally you rely in clutter e-mail, point to point communication, or the enterprise content system does not allow project team members to reuse, learn, with other projects outcomes and your are anxious to adopt a social stack, before doing it, analyze and align your social network structures. If you are a manager and the dominant type a Start network and you need your knowledge flowing, not centralized, social technology is useless.

Knowledge types and social networks

As most of sociologists write, consultants like to draw concepts in matrix graphs like the one above. I concur with this criticism regarding it does not show the dynamics of knowledge shifts  that naturally occur inside organizations, because company mission changes or because naturally business processes do not belong to a single type, but are blended.

If you are interested how you should align network structure to the process bandwidth your company play, join me at Business Process Management Conference Europe 2012. A preview of what is going to be presented can be learned also in the webinars hosted by BP LOGIX.

References:

Polanyi, M. :The tacit dimension. New york Doubleday, ISBN-13: 978-0385069885

March, J. G. : Exploration and exploitation in organizational learning. Organization Science, 2, 71-87

Nonaka, I: A dynamic theory of organizational knowledge creation. Organizational science, 5 pages 14-37

Scott, John P : Social Network Analysis: A Handbook, Sage Publications Ltd, ISBN-13: 978-0761963394