‘Collaboration quantified – solutions to measure and analyse user’s behaviour in the Google Apps environment’
The General Audit Tool is a leading G Suite marketplace tool for auditing and securing your Google Apps environment. Over the 4 years, we have been developing the product we realized we were working on a potential goldmine of useful business intelligence. Our challenge was to extract that information for each organization.
Our goal at the General Audit Tool development group is to engineer ground breaking software and tools for the corporate Google Apps environment, which increase your security, expand your insight and greatly increase your return on investment. We will give you a view from the top that is new, insightful, beautiful, yet simple, powerful and intuitive. Audit is simply the means to our goals, come with us and we will show you the power of what we can extract.
It is generally accepted that collaboration enhances both worker experience and output. In light of this, the important questions for all organizations are ‘how do you enable collaboration?’ and if enabled ‘how do you measure collaboration?’
Our view is that in the Google Apps environment people have many of the tools they need to collaborate today. Tools such as drive sharing, email, calendar and Hangouts. If collaboration is not happening enough one reason is because it is not being enabled, either through culture or training.
The problem for management is that they can’t measure it or see it, they have no ‘top view’ of collaboration, no metrics, so they don’t know where to apply fixes or how much fix to apply. Management also can’t say if collaboration is contributing to success or not, because they have no idea how much collaboration really happens with and within successful groups. The General Audit Tool (GAT) solves these problems.
We have built a suite of 6 business intelligence applications, each designed to give unique insights into the operation of the enterprise domain and collaboration of enterprise staff with each other and with customers.
We have developed many techniques both to improve information accuracy and to reduce ‘noise’. We have worked to distinguish between communication and collaboration, both in email and document sharing. We also built damping and noise reduction filters allowing you to see the wood from the trees.
By being able to identify collaboration we can start to count collaboration in order to help us identify who the true collaborators are. We can rank individuals and groups in terms of collaboration performance allowing managers to track if there is correlation with real life performance. It also helps spot early signs of trouble if a group is not collaborating. In brief, the six initial areas covered are …
- Internal/External Collaboration
- Group Analyses
- Domains Communication
- Knowledge Research
- Social Graph
- Collaboration Measurement
In the following few pages, we will go through each of these tools and show how they can help a business in real life. Several of these features are entirely new and much of this is from the cutting edge of network research and modern graph theory. Some of the visualizations are copyrighted as they are novel and have never before been represented. Much of the research is underpinned by statistical analyses.
Before displaying the data in a graph it is possible set both the time range of collaboration you are interested in and also to focus on a specific Google group. For large Google domains, this will help focus to be achieved quickly.
You can, of course, leave the default value of ‘All groups’.
On clicking ‘Show Graphs’ you will be taken to the graphing tool and the graph will be built in front of your eyes. At first it might look like a ‘tangled ball’, read on to see how to untangle the ball and identify all sorts of collaboration relations using this incredibly powerful tool.
(Note : Chrome users to See the graphs – Under ‘Show Advanced settings’ – ‘Privacy’ – ‘Content Settings’ – ‘Cookies’ – ‘Block third party cookies and site data’ must be un-ticked. Java must be script enabled.)
This is a graphing and analysis tool that lets you look at current collaboration inside the enterprise domain and between the enterprise domain and external domains. The default look back window is the last six weeks, but this can be extended to reach back further. The tool allows for a powerful series of filtering and dampening aids, thus surfacing real intelligence, really quickly. Apart from the time window, the auditor can decide to focus on internal or external collaboration, once graphed the auditor can focus on document sharing, calendar appointments or email collaboration or all combined.
The tool itself has powerful visual aids to surface information.
Users are represented by the circles (nodes) and their collaboration is represented by lines between them (edges). Some nodes are bigger than others, this represents volume of collaboration that this user participates in, the wider edges represent more collaboration between two specific users.
Colour is also used. Some people are poorly connected (poor at collaborating with many people – low degree) and represented in blue. Others are well connected, collaborating with many people at once, we say these people have a high degree and are represented in red. This type of colouring is called ‘Colour by Intensity’. Look for large red or yellow/orange circle – these are the key collaborators in every way.
In many cases though, where we are looking at an individual or a group, we say we are ‘filtering’ that user or group into focus, so we then want to ‘Colour by Filters’
The default ‘filter’ is internal vs. external, so when you turn on ‘Colour by Filters’ for the main domain, blue NOW represents the internal users and red NOW represents the external users.
Then the auditor can select an individual, a group, an OU or domain as the target of their focus. If an individual is selected all the groups to which that individual belongs are automatically selected and presented on separate tabs. This allows for single click analyses of how that individual is collaborating inside any of their many groups.
If a group or Organization Unit is selected a statistical analysis of the groups (or OU’s) collaboration behaviour is performed.
This analysis covers both level (which looks at the total volume of collaboration across all channels) and degree (which looks at the spread of collaboration among users).
For any given combination you can turn on or off more options using the dampening buttons at the bottom of the page
Always the focus of the filter is in Green, the rest of the internal nodes in Blue and external nodes in Red. Sometimes you want to study the collaboration between a user and a group or between two groups. Non-participating internal nodes are then represented in gray. These nodes can be removed from view by dampening internal collaboration.
In fact for all selected criteria, dampening can be applied to remove data not relevant to the subject of focus. This ensures that ‘noise’ from non-relevant nodes and edges can be reduced to the minimum.
As if all that was not powerful enough a second set of targets can be chosen to allow the auditor to study the interaction between both areas of interest. For example this lets you study how sales is collaborating with pre-sales, or marketing with sales, or support with a external customer domain. In seconds you can surface the size, the spread and the direction of any collaboration relationship. At any time you can press the ‘Reset’ button to do a new analyses.
You can start to ask and answer questions like ‘do sales and marketing collaborate?’, ‘are we collaborating with supplier A?’, ‘who is the gatekeeper in that group?’, ‘does this group suffer from factions?’, ‘are there any isolated individuals in a group?.
These and many more questions like them can now be asked and answered for the first time in a meaningful way.
Sometimes you just need to see the data behind the image, in particular when the graph still has a lot of data, to look at the relationships in table form, click on the tab ‘Pairs’, to see a list ranked in order of magnitude of all the collaboration pairs on the graph.
See how internal groups collaborate with each other or with the outside world. Identify the key communicators in each group and who the external relationships are with, dampen out non-relevant connections quickly allowing you to see the tree from the woods with just a few clicks.
In addition we also do keyman analyses. Here we can mathematically study the relationships of any individual and identify, from a relationship and collaboration perspective, who is best positioned to replace an exiting individual based on degree of collaboration overlap.
For any user, select the group where you want to see who their best replacement is and look at the table underneath for the ‘Two closest users’.
We appreciate on very large domains, even with very flexible visual tools it can be hard to identify problem groups quickly. Often the fastest way to identify an issue is with mathematical analyses. Using this tool we automatically statistically analyse all your Google groups against their collaboration metrics. We can identify the top collaborating groups both by amount of collaboration and degree of collaboration. Amount measures the amount of collaboration while degree measures the spread of collaboration. In both case you are looking for high scores and low variance, indicating that all members of the group are working and collaborating with other group members in a strong and universal fashion. Presently no other tool can give you such insight into the performance of your groups with just a few clicks.
To explore a group in detail, just click on the group name. You will see the group is expanded to show you the individual detail behind the collected averages. This can be very useful to help identify the top collaborators in a group.
Using this data you can start to correlate other performance indicators with collaboration. You can start to ask and answer questions like ‘do collaborating sales teams close more deals?’, ‘do collaborating support teams have higher customer satisfaction scores?’. It will also allow you to identify unnecessary groups, perhaps allowing for them to be formally wound up, allowing for the better use of resources. You can search by group size up (eliminating smaller non-functional groups) or by group name.
Using this tool we can see how you are communicating with other Google domains across time. Our unique copyrighted chart always places your domain at the center of the graph and shows it’s relationship with every other domain in the chart in a manner that reflects the size of the relationship, the direction of the relationship, the relative strength of the relationship and the relative age of the relationship based on activity. This extremely powerful graph can be drawn based on the data from any timeframe, and can be further filtered by narrowing the time window inside the original timeframe or by selecting relationships of different strengths inwards or outwards. There is nothing comparable in any other software environment.
With this chart you can see both emerging and fading relationships. Managers can quickly identify where new relationships are starting and also where old relationships are fading. You can quite literally see at a single glance the activity horizons of your business. You have an instant view of the domains that your company are interfacing with, both through email, calendar appointments and document sharing. Communication differs from collaboration because it can be often one-way and may not form part of a collaborative event. It is also susceptible to a high level of noise. Spam like email in or out might be recorded, as might document sharing that had no follow through, thus showing intent, but not action.
It is equally vital for managers to be able to track this.
This unique copyrighted graph can give a manager a full view of the horizons of his departments relationships, identifying new and emerging relationships on the right and old or fading ones on the left. A manager can identify the key opportunities even before he hears from his staff and can easily view which relationships are being neglected. Both the time frame and the ‘density’ of the relationship can be set by the manager. Email and document sharing are covered. There is no other tool in the world like this.
Domains Communication can also be approached from both the drive and email audit tools. This will allow for the powerful filtering options available in both these audits to be applied before displaying the domain network. For example if a manager is only interested in the domain relationships of the sales team in the drive audit the manager can select that group as filter and if then if required further refine it the selection by selecting only spreadsheets say. Any time period can be selected. Clicking through to the Domains Communication will then give a full visual summary of all spreadsheets shared to and from external domains. Anomalies will jump out immediately in a way that is not possible from a tabular view.
This tool allows you to search for any Contact, email address, domain, product, part, topic, subject, etc. domain wide. The results returned are a list of individuals ranked in order of strength with a relationship to, or a knowledge of the search topic. Use this to find internal expertise or the right connecting into a company.
A click of a button can also output the results to a spreadsheet for legal hold.
The scope of the search also covers emails and contacts.
The search criteria can be set to be broad or narrow, allowing you to focus on those who have first degree knowledge of the subject or a direct connection with the domain or the individual.
There is no better way to find the internal expert on a subject or the person with the right connections to the company you are targeting.
In addition Knowledge research can also generate a spreadsheet on all the emails, documents, calendar appointments and G+ postings relevant to the search topic. this spreadsheet can then be used as the basis of a comprehensive legal hold. By it’s very nature it will be domain wide and will have identified documents and emails in the accounts of third parties that are related to the search subject. No other such comprehensive search mechanism exists in Google today.
Enter a seed contact from your domain in the form of an email address and find all the internal and external social relationships of that contact. The seed allows us to build the social map inside the company and then identify all first degree external contacts for the members of the internal social network.
The depth of search can be set to show ‘friends of colleagues’. Extra filters can be set to see who is following and who is being followed. In addition we have a filter to identify only those who have a declared organisation (eliminating most personal connections). When finished the entire relationship map can be output to a Google spreadsheet for further inspection or analyses.
Many companies talk about the social network – few understand what it means for their staff and what the connections are.
Internally this is the best way to see who the key social influencers are for your organization.
In the ‘Users’ audit section of GAT we also present information drawn for the BI analyses.
The internal collaborative events, such as emails sent and replied to, documents shared and visited and calendar appointments made and confirmed are all scored and totaled. This allows you to see and report on who the top collaborators are inside your domain.