Check out this second blog of the series "User-Edge Operations" by Jim O'Brien, the Vice President of Server Engineering, Support and IT, CounterPath, where he shares deeper insights about the components of the User-Edge Toolkit and its benefits.
One of the most fantastic parts of my role at CounterPath is that I get to talk about what our team is working on, what we've built, and the problems we've solved. I enjoy it thoroughly. This series of blog posts on user-edge operations is another expression of that privilege.
Before we get into it, I want to take a moment to thank our team members. Internally I receive a bit of flack because nearly 100% of my emails to more than one person at CounterPath and when talking to customers start with this simple opening: "Team,...". It is important to me. I want to work with a group of people who operate as a team. This becomes even more crucial when we reach across organizational boundaries and work with partners, customers and vendors. Believing in the strength of our team is a core value at CounterPath.
Our user-edge toolkit is a perfect example of the collaborative approach that exemplifies the strength of our team. Members of our Support, Operations, Marketing, QA, SDK, Client, Product and Server groups built this toolkit. The results as you will see below are awesome. Thanks to our team!
As discussed in the previous post in this series, our User-edge toolkit currently has 5 components that we believe power up the game of any operations team looking to not only improve their response to customer issues, but also looking to get some proactive insights into the traffic on their networks and the user experience that this generates. Let's jump into the toolkit!
User-Centric Client Management
User-centric based client management is a long name. It has a simple meaning, however. We manage devices on behalf of users and correlate the data from those devices to the user. Users and their experiences are more important than their devices, but devices often drive user experience! Devices then become components of a users experience. This allows for device configuration, in particular, to drive the details of the user experience across devices for a particular user. Devices can then act in a coordinated manner and the data the devices generate can be leveraged on behalf of the user.
Breaking this down:
1) User-centric Client Management starts with provisioning clients based on the services that a user is enabled with. This is service enablement. A user may have a Voice Service a Collaboration Service and a Messaging service. Each of these means that the client must be configured with a potential variety of service enablements to allow the user to benefit from these services and use the services in any manner where they may overlap.
2) User-Centric Client Management also includes tracking device usage across and among populations of users. This is a critical step for both controlling how users leverage the service offerings that are included in their services bundle and understanding how the data generated by the devices map back to a particular user.
3) User-Centric Service Enablement finally includes individual feature enablement. This is driving configuration based on the user, their profile membership (the type of user, category, services enablement selection, etc.) and organizational group. This might include turning on UC features, Voice, Video or other tiers of more technical features like codecs or call transfer etc.
We think of the broader Device Management function in two distinct categories — Device Tracking and Management. Device tracking looks into tracking the number of devices consumed by a user, group or other organizational structure. This can be used for royalty reporting or other charging or usage mechanisms.
Device management moves more into how the users are allowed to leverage a set of devices over time. Management includes policies that allow users to self-service / replace devices over time, based on an allowed # of devices and a device aging mechanisms that allow unused devices to be removed from a users list of active devices, freeing up space for additional devices. Management also allows administrators to prune down lists of devices after certain changes, or view all the devices in use by a population and export this data into third party tools that allow for better BI (Business Insights) data analysis.
Device management can include tooling to understand device usage over time and identify trends in device usage that change user experience. For example, in one population the number of users leveraging a single device for a particular service was 57 percent in 2017. That number was reduced drastically to 26% in 2018. The users not only moved from the majority desktop only, but many of the users used their mobile devices twice as often as their desktop devices. The intelligence gained from this kind of usage patterns can be striking, much more than the well-established trend that workforces are mobilizing. In particular for service providers looking to justify the value of their service tracking the expansion of user devices from desktop to mobile and perhaps from workday to night to weekend allows for immediate access to trends that drive value creation or shift policies to adapt to evolving user habits.
As a specific example: watching user uptake of new generations of devices or operating systems is an important way for application developers to understand when they need to add support for new models, OS versions and more importantly when they can drop support older models or versions.
We have a longer name for client traces which is Enhanced Network-Centric Client Tracing. Normally 'Client Traces' gets the idea across. These are logs generated by the clients that include stack level logging information. They also include information that is quite possibly as close as you can get to network truth without running Wireshark — which is critical, because in the vast majority of real user-network scenarios, running Wireshark is either not possible (mobile LTE networks) or not deterministically practical (dedicated VLANs, Security/IT policy, root/superuser access, etc).
What I really like about Client Traces are that they provide the right level of context. You can see when DNS-SRV fails and regular DNS is used. You can see events from the user-interface and XMPP stack level messaging revealing the details of the signalling between a client and their presence server. These kinds of logs, that also include what came from and hit the wire are powerful focal points for any troubleshooting activity.
The second component that makes Client Traces powerful members of the User-Edge Toolkit is that these logs just don't languish where they are generated. Client Traces can be sent by the user in response to a problem or requested by an administrator; and are transported across the network to a collection point. This collection point is the Stretto Platform, where admins work day to day to analyze other data and configure user services. This round trip of user creation, data population, user-service fulfillment, and feedback create a powerful current for data collection that supports operations teams' workflows.
In addition to proactive and data analysis measures, sometimes we all need reactive measures in our toolkit. At CounterPath, we call this capability — Help Desk Assistant (HDA).
We've seen some in the device space offer remote assistance capabilities for their devices. In our space, we've built HDA as an in-application remote assistance capability. HDA is 100% controlled by the user with each session requiring not only user enablement (start help desk assistant) but also out-of-band confirmation that the user wants to allow access to HDA to a particular support person through transmission of a short code over the phone or chat session to the support person assisting the user.
HDA provides a deep diagnostic toolkit to the service or network operations team. At its core, HDA allows the support staff member to tunnel into the UC app and understand what is going on. The support staff member has a large set of tools they can leverage to look at the application's configuration and logs. They can also perform some low-level tests or cause the application to update or upgrade itself. In its most advanced and compelling form, HDA allows the support team member to take over operation of the UC application and place and receive calls or other user-level actions on behalf of the user. This allows that support member to conduct testing on their own, relieving the end user from the burden of participating in the troubleshooting session after the point that they enable the HDA capability.
User Experience Metrics (UEM)
User Experience Metrics (UEM) is the last tool from our user-edge toolkit. UEM highlights the importance of both wide and deep data availability. It is vitally important to collect information about all users within a population as well as collect information on all of the calls that each user makes over time. With both of these axis covered; operations staff investigating quality problems have a broad set of comparisons they can make.
Our operations teams customers leverage this information to establish a trend line and understand each user’s normal operation. The details here are important. Each user is a collection of their devices, the locations they move between, and the networks they connect through. This provides a wide variety of situations, equipment and quality range. The amazing part here is that each user and each organization build its own data-set over time. Within this data-set, there will be outliers and there will be normal trendline for each user, each device type, each user-group or location.
Then the fun begins :-) With this data operations, teams are also able to look not only at the network level quality indicators, items like jitter and packet loss; but also at softer metrics like the brand and model of headsets or information about the how their users leverage their devices.
The other obvious use of this data is in response to user reports of changes (usually negative) in call quality or a bad call. Here the operations team can zoom in on the particular call quality information as well as metrics from the device at that time. They can understand where in the infrastructure that user was connected from and compare that user's call(s) to those of other users.
For the next post in this series, I plan on talking about the benefits and best practices we've learned for leveraging user-edge toolkit techniques to enhance network operations practices. I'm also excited that we will be sharing some interesting statistics based on a small sample of the UEM data we collected from about our own team's usage. The insights we gained by looking at a single random sample of 1000 calls are quite fascinating.
If you are looking to add one or more of these tools to your kit, please review the available Stretto Platform documentation here or contact us so we can talk to you about getting started with these great tools.
I hope you're following along with me and that you'll be back for part #3. If you haven't subscribed to our blog, please take a minute to do that now (by clicking on the image below) so you don't miss the next instalment: