Relationship Mapping and Monitoring with Yammer

Yammer&MicrosoftYammer is a leading social networking platform for use inside organisations. Its recent acquisition by Microsoft is not only good for Yammer, but for the many Microsoft Enterprise clients who have been struggling to ‘connect’ via Sharepoint. What is most exciting for us is that the combination of Microsoft’s Active Directory with Yammer’s conversational platform now provides a real opportunity to implement the ‘Real-time Social Business Dashboard‘ which will enable enterprises to move beyond their current process monitoring to see how people are really collaborating (or not) to meet organisational objectives. 

So what’s wrong with the analytics that Yammer currently provide? Well nothing really, other than the fact that its only using a proportion of the intelligence available from its tracking data. Like most of its competitors, the analytics are what we would call “ego-centric”. In other words they track the activity patterns of individuals and then aggregate the data at team, department, company level to assess the level of engagement (read usage). The knowledge management community learnt a long time ago that activity doesn’t always map to productivity. Setting performance measures against ‘documents submitted’ resulted in lots of poor quality documents being uploaded just before performance appraisal time. But ego-centric measures will still reward this. In the social media space numerous postings in forums or voluminous tweets provide little indication of effectiveness unless they provoke a response. Social network measures focus on the relationship. Relationships are jointly owned. A direct response to a post creates a relationship. A heavy interchange of messages infers a stronger relationship (not necessarily friendly, but still a more established one than  where no interchange has occurred). A ‘like’ is also a connection. Counting ‘likes’ can be good for the ego, but even better when we know who is doing the ‘liking’. 

The Social Business Dashboard has ‘relationships’ at its core. That is not to say that current ego-centric measures would not be included. For example the volume of posts is clearly a useful indicator of activity. However social business is about collaboration. ‘Connected activity’ is what we are looking for, as we know that this form of activity is what leads to high productivity. The key component for a Social Business Dashboard is the Social Network Map. The map makes visible the network connections exposed through Yammer. By analysing the map one can see the flow of knowledge and information across the organisation. Accompanying analytics can identify who your key talents are, not by their CV, but who actually seeks them out for advice. We can identify the level of reliance on key players, the level of cross department collaboration, the areas where there may be bottlenecks impacting on customer service, order to cash cycles, ideas to innovation cycles and/or prospect to client conversions.

We have written previously about how network analytics can predict higher levels of efficiency, effectiveness and innovation, how social business drives ROI and what we are calling Social Analytics 2.0. We think this new relationship between Microsoft and Yammer will pay dividends by bringing ‘Social’ into mainstream enterprises, flagging a maturity in the market that we have long waited for.

Example Yammer Interactive Social Network Map

Below we provide an example Social Network map derived from a Yammer installation. The context was an “open innovation jam” where participants were drawn from across businesses to explore new energy and sustainability ideas and opportunities. ‘Connections’ are drawn from the Yammer discussion forum data. Fictitious organisation names have been added to provide an illustration as to how Active Directory profiling information would be included in the map.

The Optimice Webmapper utility enables one to interactively explore a social network map. The ‘flyout’ menu (use the orange triangle to control) allow you to select what attributes you want to colour (Organisation or Explore-Exploit) and/or filter the nodes by. You can also choose what relationships (Relationship Strength or Time) you want to explore.

The initial scenario shows the ‘Relationship Strength‘ map. The strength is determined by the number of posts made between pairs of participants. The size of the node relates to the number of posts received; a possible indicator of influence. Move the strength slider from left to right to expose only the strongest connections. The red links identify reciprocated postings. We like to think of these as another indicator of relationship strength. You can use the explore-exploit selectors to show only the explorers, or only the exploiter organisations. Clicking on any individual node will expose the network for just that individual. Select ‘Show All‘ to restore the map. If you like you can use the flyout menu to change the filter to ‘Organisation‘. You have to press ‘Update Map‘ anytime you change something on this menu. You can now turn on and off organisations.

Now try and change the ‘Relationship Strength‘ to ‘Time‘ and the ‘Strength Type‘ to ‘Strength >‘. Press ‘Update Map‘ to expose the new map. The map shown is the final state map but move the strength slider from left to right and the map will show how the connections built over time. You can now imagine how a dashboard might show this map evolve in real-time, while allowing analysts to ‘replay the past’ to diagnose impending issues and/or opportunities.

We are currently looking for organisations that have successfully integrated Yammer into their Microsoft Enterprise environment and are interested in pursuing a ‘Social Business Dashboard’ strategy, as described above.

Contact or

Human Capital Flows in Australian Banking – Courtesy of Linkedin

One of the lesser known features of Linkedin is the company insights feature that identifies where a company’s current staff have been employed in the past (just search for a company and then select ‘Insights’ to see where employees have come from). Now its not hard to take this data from selected firms in a given sector to create a network map based on people flows between these firms, if you like, a  visual market representation based on people flows.

So what insights can we gain from this type of visual market? For a start people flows are the closest we can get to identifying critical knowledge flows. Despite the emphasis that is put on protecting intellectual capital assets through the use of patents and secured document libraries, the real competitive insights are often held tacitly within the heads of the employees. While patents may protect old and existing ideas, more often it is the value of the ‘next idea’ that drives much of the headhunting activities undertaken by the recruitment industry. Of course its not uncommon to see a move by a high profile employee accompanied by a longer list of loyal followers to their new company. The high tech industries in particular are rife with poaching and headhunting activities, looking to leverage the unique knowledge of a successful product creator. 

Another aspect that can be investigated is the diversity of the source of employees. For example, its not uncommon for consultants and sales people in the IT/consulting industry to move between the top 4 to 5 firms in the sector. Ultimately this can lead to a ‘sameness’ between the key players in the industry, resulting in less innovation and less diversity of offerings.

To explore some of the insights that a people flow focussed visual market can provide, we have selected to investigate the top 4 Australian banks; ANZ, Westac, NAB and the Commonwealth. The Australian banking industry has been somewhat protected by anti-merging laws that govern these top four banks. This has led to a polarisation of the market around the ‘big four’, making it difficult for the smaller banks to grow to compete on  level terms. We used the Linkedin Company insights feature to mine the people flows around these four major banks. This feature provides the top 5 sources only. Using the snowball sampling approach we identified another 24 companies to form the ecosystem for the visual market map shown below:

Note: This map works best with Chrome, Firefox, IE9 and Safari 6 and above.

Some of the functions/features that you can explore with this map:

  1. There is a ‘fly out’ menu (select the orange triangle) where you can choose what you want to colour and/or filter the nodes by. The start up map uses the Australian head Office location for both the colouring and filtering.
  2. You can manipulate the ‘zoom’, ‘font’ and ‘expand’ sliders to explore the map visually
  3. You can slide the ‘relationship strength’ slider to expose the strongest flows
  4. The thickness of the lines and direction of the arrows identify the size of the flows.
  5. The red lines indicate a bi-directional flow.
  6. The size of the circle reflects the number different firms the company draws from (note this is limited to a maximum of 5 by Linkedin)
  7. The different colours represent the attribute selected. The other options are ‘industry’, ‘country’ and ‘vertex (company name)’.

Some Potential Insights

Here are some of our insights from manipulating the map. We’d be pleased to hear about yours:

  1. By moving the relationship strength slider to the right we expose the Sydney/Melbourne Head Office split. ANZ and NAB are headquartered in Melbourne, with Westpac and Commonwealth headquartered in Sydney. Staff appear to be happy to move banks but not as much cities. The balance of flows between these same city banks did not appear to favour one bank over the other.
  2. Click on IBM to find who they provide staff to and you can see that IBM is a dominant  resource for each of the major banks as well as Australia’s leading telecommunications company, Telstra. Much of this could be attributed to IBM’s dominant position in outsourcing and mainframe computing which is central to big banking.
  3. We can see that ANZ and to a lesser extent Macquarie Bank are ‘brokering’ the most international expertise, mostly from HSBC and Citibank. You can simply click on ANZ or Macquarie bank to expose their ecosystem. Macquarie bank appears to ‘feed’ Commonweatlh bank. Select ‘show all’ to get back to the full map.
  4. Select Westpac and carefully move the relationship strength slider. You may note that their ‘acquisitions’ in BT and St. George Bank are being drained of resources to the ‘mother ship’. Not as much going the other way. In fact St.George Bank also has a significant flow into Commonwealth Bank. Again the Sydney head offices would have something to do with this.
  5. Under the ‘flyout’ menu select ‘industry’ as the filter and then select ‘update map’. Now you will see the filter selections are industry sector. Now uncheck banking and finance. You should now see that Telstra is gaining significant people flows from other providers like IBM, Optus and Ericsson.
  6. Go to the fly out menu again and select ‘country’ for both the attribute colour and filter. Remember to select ‘update map’. Now uncheck Australia and the international network is exposed. Citi Bank and HSBC appear most dominant.
  7. Select Australia and the USA and uncheck the rest and you can see the role Citibank plays in connecting USA banking resources with the Australian banking industry via ANZ and Macquarie Bank.

We are not banking experts, so its not possible for us to know whether the above are really true insights or something everyone in the industry knows. If you work in the industry we’d value your comments and feedback.

CeBIT – Oz Day 1 Twitter Storm

Day 1 CeBIT Australia is now over and the twitter storm was intense. Some 1,244 tweets against the #CeBITAus tag. But what about the TwitChat? Only some 250 messages were actual interactions i.e. a mention/retweet or a direct reply. Storms are something to duck for cover from. However, amongst the tweet storm there are some gems that people think are important enough to mention to their followers….just 14% to be precise. We can learn a lot from this 14%. For example who is the most influential Tweeter based on the interactions that they provoke. We used the NodeXL software to import data from twitter to visualise who talks to who:

CeBIT TwitChatThe twitter profile pics are sized by the number of mentions or replies the tweeter has received. So the bigger the picture, the larger the impact their tweet/s have had on others. Here is our ‘league table’ of ‘most noticed’ tweeters for CeBIT day 1:
Twitter Name
Number of Mentions
Now we can learn a little more from the patterns of interactions from the maps. For example lets look at Lisa_Cornish’s direct interactions:

Cebit-CornishWe can see that 22 people have each mentioned Lisa’s tweet, so clearly it was the most engaging tweet of the day!

Now lets pick out someone else from the map and look at their pattern of interactions:

Cebit KcarruthersWe can see that ‘kcarruthers’ has fewer mentions than Lisa, but a bit of a ‘community’ developing around her. Lisa may have tweeted something very noteworthy, but I suspect that kcarruthers may have longer lasting influence through the community surrounding her. 

So those of you into CRM and Social CRM take note. Its not just the number of interactions, its also the nature of them!

Feel like exploring a little more? We have taken the NodeXL data and created an interactive version using our WebMapper technology. Be warned, it works best with Chrome, Firefox and the latest versions only of IE and Safari. Also only the newest IPads (IOS 6)

There is a lot you can do with the interactive map. Firstly you should play with the ‘expand’, ‘zoom’ and ‘font size’ sliders to explore how to visualise the map. The next thing you will notice is the different colouring of the nodes into some 20 groups. These were determined by NodeXL’s clustering algorithms which try to cluster nodes based on common connections. You can explore these by checking and unchecking them in the tick boxes. You can also check and uncheck the ‘mention’ and ‘reply-to’ links. Finally you can select a single node to see the network that surrounds just them (like we did with lisa_cornish and kcurruthers). Mousing over the node will expose their twitter handle. You can search for a twitter ID in the search box. ‘Show all’ will restore the original map.

Enjoy your explorations!

How Social Network Analysis (SNA) Combats the Tyranny of Top Down

We are prompted to write this post by Gary Hamel’s excellent exposition on “Leaders Everywhere”, together with some recent SNA work we have been conducting that speaks directly to the thesis around developing leaders at all levels of the organisation. This also builds on our previous paper ‘Tyranny of Top Down‘.

Much of the early thinking around inverted hierarchies can be attributed to J.B. Quinn and his writings on the “Intelligent Enterprise”. Prompted by the slow demise of the manufacturing led industrial sector and the burgeoning growth in the services sector, Quinn proposed the radical idea of inverting the hierarchy and placing customer facing staff at the top level, supported, rather than being directed by their respective line managers and executive staff. 
Inverted HierarchyAs Hamel eloquently puts it, many organisations that are locked into industrial age business models are now trying to become more client focused, adaptive, and have accountability for results at all levels. We think the first step in this journey is to understand what the current status quo is. We are using a recent case study of ours related to an organisation that has a long industrial heritage. Global competition and the drive to radically reduce their cost of doing business are forcing them to consider the sorts of interventions that Hamel is suggesting. But before launching into such interventions we suggest that it is always a better practice to ‘take the x-ray’, to understand the detail of the current status quo. In large organisations it is rare that all parts of the company behave in precisely the same way. Using the insight of the x-ray, one can provide the treatments to areas of most need, while highlighting those areas already operating in accordance with the desired state.

Using the facility of Social Network Analysis (SNA), we can take that x-ray to see what the relationship network in an organisation really looks like:

Full X-Ray

This social network map shows how the staff of this organisation are connected by ‘dependency’ i.e. staff were surveyed and asked who they were most critically dependent on in doing their job well. The red links show where this dependency is reciprocated. We like to think of these as ‘trusted links’, where knowledge and information is likely to flow more easily. The colours represent the different formal business units. We can already see some ‘siloing’ around business units, something that is common in top down hierarchies. The circles are people. The size of the circle is related to how many dependency nominations that person has received. A large circle suggests that this person is important in the organisation. In very hierarchical organisations, we would expect to see the larger circles identified with the line management roles.

Now because this organisation was interested in who the ‘natural leaders’ are in the organisation i.e. those staff who attract dependency nominations, not based on any formal role authority, but simply through the value they provide to their peer worker. Through the ‘magic’ of SNA we can explore this scenario by simply removing all the line managers from the map:

No Managers X-Ray

Now the circles are re-scaled to reflect just those nominations that are received from their peer shop floor workers. It is evident from this map that the line managers are playing a linking role between business units as the silos are even more defined; something that we would expect from a top down hierarchy. But it is also evident, especially within the red and yellow business units, that work can still proceed without a strong reliance on their line managers. This is encouraging as perhaps some of the larger circles in this line-manager-less world could be the natural leaders that the organisation is looking for and who’s value adding behaviours they could promote and replicate.

Lets now look at some of Hamel’s recommendations and how we can use SNA to achieve them:

Hamel Recommendation

How SNA can be used to help

Develop smaller, more accountable teams.

We can see from our maps where the natural teams lie and therefore these can be leveraged immediately.

Attach compensation to value provided.

The first step here is to check the compensation for the higher peer nominated staff in the ‘no managers’ map. We should then visibly compensate those voted as adding most value.

Open up the strategic conversations and information flows from the top down.

The SNA map identifies where the information flows may be constrained and who may or may not be playing important information brokering roles. SNA results can be used to maximise efficient information diffusion.

Collect peer to peer feedback

This is what SNA fundamentally does.

Provide fast feedback to facilitate adaptability at the customer interface.

With the help of enterprise social networking tools, it will be possible to provide peer feedback in real time, through mapping conversational patterns on social platforms. See Social Analytics 2.0 – Its Time

Train and equip leaders with skills, information and contacts to be fully accountable for the work they are performing.

SNA maps work to the granularity of the individual. Using SNA tools we can drill down to an individual leader to assess what their current network looks like and then assess what can be done to improve this. We have blogged previously about this in Is your Personal Social Capital Helping or Hindering Your Leadership Aspirations? 

To conclude this article on a leadership development note, here is a quick quiz:

Here are three personal network maps for leaders drawn from an overall SNA X-ray. Pick out who you think looks  like a Co-ordinator/general manager; a domain specialist; or a potential innovation broker? (Note that the colours represent different business units).

Leader ProfilesKeep watching this space for announcements about our new ‘Personal Networking Diagnostic’ for emerging leaders.

The wilting initiative – Are you at risk?

Dying InitiativeA new corporate initiative has been established. Business sponsors are hand picked, a program office is established to manage it and the change team is driving the initiatives across the organization. But then a key person leaves and suddenly the initiative loses momentum and starts wilting.

Think about the string of initiatives that runs through organisations today like Six Sigma, Lean, Social Enterprise, Cloud, Innovation, etc. To get the organisation focused, and to get these initiatives launched, they need sponsorship and a core set of people who drive them. But all too often there is an ongoing critical dependency on the key people driving these initiatives. We only find out that sustainable change has not been achieved until the core team disbands. Initiatives are really easy to get going, but hard to embed into ‘business a usual’ and so the organisations end up being dependent on a few passionate souls to keep the momentum going.

We’d like to suggest a way to determine exactly the extent to which your initiative is reliant on a selective few, and what you can do about it.

Case Study

EngineeringdrawingOne of our clients, a relatively young organisation responsible for delivering capital projects has a project portfolio worth $3 billions. At the point of our engagement there were 13 project teams delivering large construction projects across a geographical area of almost 1,000,000 square kilometres.

The organisation knew that Organisational Change and Communication were critical elements that had not been addressed well in the past. Staff were not happy when they could read about a local project in the media before any announcements had been made internally. They knew that this needed to be addressed.

Selected staff members had been appointed to drive the adoption of a change and communications approach across all of the project teams to create a step change in planning and execution. These staff members were available to assist the project teams by creating reusable templates and act as ‘coaches’ to the project teams.

Eighteen months after the initial creation of the change management and communications program, the central team were unsure exactly how well the project teams were working with each other, and how well the central team were connecting with the project teams. A recent escalation in workload suggested that the local project teams were not drawing on each other but still heavily reliant on the central team.

We were engaged to help uncover the collaboration patterns within and between project teams and their interactions with the central team. We surveyed 4 key project roles across the 13 teams as well as core members of the central team, and in the map below you’ll see what we found.

Each of the people surveyed are circles and we have drawn lines between the people based on the question ‘who are the most important people you draw on to deliver this project’. Arrows indicate nominations. In other words, many incoming arrows indicate a really important person.

(interactive version)

Program With Central TeamTwo of the key members of the central team is right in the middle of the map. Below we have removed the central team. 

Programwithoutcentral-teamWhat is left is a number of quite weakly connected project team members and some of them are now working isolated from any other project teams.

Armed with these insights the central team started planning towards the next phase of the ‘maturity’ of the program. That is, to enable the teams to work with and learn from each other and become less reliant on the central team.

If you’re in a similar situation, here are a couple of ideas about how you may lessen the (over-) reliance on a few individuals:

  1. Create regular knowledge sharing sessions where the central/program team don’t act as ‘experts’, but as brokers
  2. Rotate the role of ‘chair’ for meetings so more people take ownership
  3. Make ‘site visits’ a part of the induction process to foster sideways collaboration
  4. Short exchanges of team members. Working along side someone else creates a stronger connection
  5. Peer reviews of major deliverables before ‘top down’ approvals

Organisations should be acutely aware of the risk of relying on a few passionate souls. The first step to mitigate this risk is to map out exactly how reliant you are and on whom. Once you understand the exact gap you can put in place very targeted interventions to minimise the risk. Finally, map the relationship patterns at regular intervals to check the effectiveness of your interventions.