TM Forum members banded together to look at how best to assure network service, and chose to take the doctors’ route, the ones who all know that prevention is better than any cure.
Specialists from some of TM Forum’s member organizations united to develop a more proactive approach for digital service providers (telcos, smart cities, OTTs) to handling network issues. By monitoring risk to prioritize network management activity, this Catalyst project, Customer-Centric Service Assurance Phase 2, sought to stop issues before they happened.
Notably, the Catalyst also centered on a way to get more of the customer’s voice in its actionable data. Now, as well as understanding network issues, telcos can match customer reactions via Twitter, NPS scores, etc. with live network data to help with this prioritization.
A little background
That was the second phase of the project; the first back in 2016 identified this need for a customer oriented methodology to service assurance and developed a best practice approach.
“The reality is, the customers aren’t happy in the telecoms industry and the NPS [see image below]scores tell that truth, telecommunications scores bottom of all the technology sectors. So we need to do something about it,” said Martin Finn, Sales and Marketing Director, Galileo Software.
The team developed visualization-based tools making complex networks, and any problems they have, easier to understand, and thus quicker to resolve. Click into the video below for a quick overview of what users can see and do.
What the team members themselves contributed to this new phase is as follows:
- GCI and Smart Liverpool set out the problem statements (extract below):
Digital services providers and enablers such as GCI and Smart Liverpool need key employees to be able to visualize and understand multiple data sources such as network issues, service impact events, sector information, social media feeds, utility feeds and engineer tracking data with customer experience metrics (such as NPS, CSAT, CES). A useful, accessible aggregation of data can help them act fast and with the customer in mind.
- Liverpool John Moores University provided guidance for APIs, metrics and the overall approach
- Cardinality determined if and how data could be collected, and decided on the best use of the data
- Galilee Software designed, created and implemented the analytics and visualization tool
Attendees at Digital Transformation World 2018 met the team responsible, who gave them a step-by-step explanation of this new phase and the strategy behind it, which I describe below.
Out with the old: CEM vs CRM
Customer relationship management (CRM) while previously helpful, is very internally focused. It looks at developing processes, systems and skills to manage the customer relationship. This approach creates a gap between the customer view of service quality and the service provider’s view, with the Service Manager trying to act as the bridge across the chasm. Customer experience management (CEM), meanwhile, looks externally to improve, through the customer’s eyes.
For example, a key quality indicators (KQIs) for service quality is availability but planned outages are not considered here, so despite regular outages, services may show as having high availability. This might help staff internally, especially in meeting KQIs and other targets, but the customer is still unclear. Measuring ‘serviceability’ could help to inform the customer better; a simple measurement of the percentage time a service is available, irrespective of outage reason.
The team used TM Forum’s business process framework (eTom) to map and analyze the project’s operational process. The framework holds an industry-agreed set of integrated business process descriptions, and was particularly relevant to the project, as these description were created with today’s customer-centric market in mind.
The diagram below shows a level 1 eTom model showing seven end-to-end vertical process groupings required to support customers and manage the business. The heart shape indicates the project’s focus on the core customer operational processes of fulfillment, assurance and billing (FAB), with assurance at the heart CEM.
The tools to succeed
To expand on what users could visualize in phase 1, phase 2 created the tools to view key customer data to compare with network and service events (all being geographically segmented) including:
CEM metrics, average revenue per user (ARPU)
CEM metrics: Net Promoter Score (NPS) – Visualize customer sentiment using geo mapping, to, for example, correlate negative service events to poor customer sentiment as a given time.
Customer sentiment: Data from customer calls to a call-centre, interactions with a chatbot and social media posts can all help to pinpoint service issues quickly and understand specific customer frustrations.
Fault and performance analysis: Network / service impact issues covering OK, at risk, critical, failed. data including: network faults, configuration errors or ‘at risk’ thresholds, metrics such as dropped calls etc.
Mapping the network: View combined network and service data in geographic segments with customer experience metrics, review this data to track changes leading up to a current issue, or to use as predictive analysis of a future issue such as a storm event, additional service added to the network, additional capacity planning etc.)
Customer experience impacts: Quality of service checks, snapshots of geospatial data tracked over time.
Service impact analysis: Using actionable insights to initiate automated workflows. Looking also at service impact events like OTT provider issues, utility data, weather etc.
Service optimization: Viewing the impact of past, present and future events/issues on customer experience, and planning or reacting to best achieve quality of experience (QoE)
Matrix Analysis of Combined CEM Metrics for each department: Technical, Sales, Marketing, Finance, Board Level, such as:
- Layers of segments – £, $ segments by value of ARPU/ average margin per user (AMPU): in customizable ranges (low, avg, high, etc.)
- Geographical plotting – Average scores within defined areas showing NPS, CSAT, CES etc, and percentage these areas in customisable ranges (low, avg, high, etc.)
- Algorithms – How to combine these metrics for insights such as the propensity to churn using social media insights, ARPU, NPS, etc.
Use case 1
The customer phones a call center about an affected service. In the video below, you can see the stages from the initial call through to the problem identification.
The service agent identifies the customer and what service he has, followed by the service ID, the customer description and what he utilizes. The blue line reaching across shows you exactly what is affecting his service and you can really drill down into the specific node details here to help with reach a resolution.
Use case 2
The service customer service agent can click on the unavailable service to see what the root cause is right. You see in the video below that the agent is highlighting the customer’s TV package and you can see exactly what is affected. This kind of information means that the service management center can prioritize what they can, and should, focus on first, while the service agent can pre-warn the customer about any existing risks based on the data say within their area or related to their specific package.
Use case 3
Customer notifications can be pushed out through apps for services that are at risk or unavailable. You can see this play out in the video below whereby the services topology portal start of all green with no problem and then all of a sudden there are problems (in red) , and risks of problems (in orange). In the next image, you can see right across the matrix of what’s going on in the network.
In with the new
The team anticipates a phase three with additional learning within the DSRM, and a closer look how the AI and analytics side of things can aid with proactivity. Watch Galeleo’s Finn discuss the project further in the video below.