Service Fulfilment — much more than just provisioning!
‘Welcome to the Future’ is my 4-year-old daughter’s favorite punchline nowadays (I think she picked it up from Peppa pig 🐷 ). However, it does remind me daily that the future we all have been waiting for is the present time we are all are living in today 😊.
In this short blog post, I will talk about what exactly Service Fulfillment is in relation to telecom operators (referred to as telco henceforth), its practical use cases and also the impact on the surrounding ecosystem.
The standard definition of service fulfillment as defined by eTom, TAM and various other industry models is to provision and fulfill the services ordered by the end-customer. However, there is much more to this than just provisioning. A traditional telco operator does not make any money from all the heavy investments made in the network (2G,3G, 4G, and soon 5G) until they successfully can deliver what is requested by the customer and start billing them. Therefore, the role of service fulfillment is much more than all the fundamental processes in OSS (Operational Support System layer). Timely activation of service, provisioning as per the technical and the business rules defined in numerous northbound BSS (Business support systems) with minimum cost is the goal of Service fulfillment.
In the past, most of the telco’s offered three to four types of services to residential customers and a couple more to business customers. These were provided over selected one or two core networks which were accessed by customers on the phones provided by telco itself. Today, however, there are new services being defined and introduced by telco’s to keep up with the competition and satisfy changing customer and market needs. Plus, these services need to be provisioned to a multitude of networks (Radio & Core) which can be used by an ever-evolving number of devices.
This could be challenging since a multitude of different layers (see figure 1 for some of the basic building blocks of telco architecture) need to work in synchronization and make sure all the building blocks work seamlessly. This is required in order to deliver the right service at the right time and to deliver a great customer experience throughout the service’s lifetime.
Service provisioning & activation
Service provisioning (SP) is one of the main functions of the service fulfillment layer. It is responsible for designing the needs of the service order in terms of identifying and provisioning all the network elements to create a complete service order before the billing can be started. Service provisioning needs to work in synchronization with different subsystems of OSS such as the Service Catalog, Service Orchestration, Service Management, Activation.
Since the number of services being offered by telcos is increasing at an unprecedented rate and with the introduction of triple play, Quad play, activation/update to a customer’s service involves modification to many network elements. It also involves orchestration of the order via multiple systems. (See Figure2. for more details.) The activation of service must be done in the correct sequence to stitch together different elements so that the customer gets a seamless service.
Many network functions are being virtualized into private or public clouds, hence service fulfillment applications must be able to provision to hybrid networks over different core networks as well. Automate service fulfillment, which includes customer self-service, with no human intervention, so that the users can activate/update service anywhere, with the least amount of hassle, is another aspect that can help telco’s increase efficiency, decrease OPEX costs.
Latest enablers like Network Functions Virtualization (NFV), Network slicing, Software Defined Networking (SDN) of network elements are new technologies that highly impact how the service provisioning can be achieved. Throw into the mix, 5G core, network slicing and IoT, it will impact how the legacy provisioning systems work with new network elements as there are many black holes yet to be solved on how to achieve this.
Some Practical use cases:
· VAS & third-party service activation: Apart from provisioning traditional services to network elements, a telco also needs to activate third-party OTT services such as Viaplay, HBO, Security services etc. which help customers avoid ordering via different vendors and have a convergent view/invoice of all the services. These need to be activated in vendors' infrastructure.
· Immediate provisioning to satisfy our customers: In current times no one wants to wait to get the services provisioned, we strive to have most of our services provisioned within milliseconds (from receiving the order till completion of the same).
· Bottoms up display of services which customers already have: This includes the presentation of services, Settings, balances etc. E.g. If customers would like to check how much data (GB’s) they have left in their subscription.
Customer Experience is everything
Sweden is a very small market to play around, with just over 10 million population. It is much smaller as compared to other bigger markets where telcos have a big potential customer base to play around and hence bigger risk potential. With 3 other major players in the market, it is very important to avoid a negative customer experience and BSS/OSS plays a major role to avoid the same. BSS and OSS layers in the architecture landscape must work in harmony to not only deliver services but predict how the customers use will change with time.
The Future of OSS/Service fulfillment is already here.
There have been umpteen number of use cases for AI, ML, Big data etc. in the telecom industry. Machine learning for e.g. had a significant success in BSS layer where it can be used to find special customized products for customers based on past behavior. It also helps Customer care departments to analyze services, issues due to outages much before customers contact them via various channels. This is partially since data sources like CDR’s, payment history etc. are well structured/vectorized.
However, data in the case of OSS is highly complex, interdependent and have varying formats. There have been some success stories as well:
Data Traffic forecasting based on the past high volumes due to holidays, TV events or security/fraud events. Machine learning can help in responding network capacity demands in a cost-effective way to meet the changes in capacity demands. It can also help in spotting pattern of intrusions related to fraud risks.
Combined application of ML at both BSS & OSS can have great synergies. Imagine, if any fault happens in the radio/core network, an automated Root cause analysis process is triggered and fed into ML algorithms which not only advise how to mitigate the issue but also predict future similar possibilities of this fault. Not only this, but it also identifies the impact of specific customer/services based on the above analysis and a customized discount, bonus etc. is presented on the customer care screens before customer contacts via any of the web channels or AI Bots. An awesome customer experience indeed 😊.
Peace,
Chakshu Arora.
