Expert Insights

AI and
automation
pay big
dividends
for telecom
Cost savings could
transform the industry
by allowing new investment
in customer-friendly technology
Experts on this topic

Warwick Hill Warwick Hill is an IBM Global Business Services

VP with more than 25 years’ experience in the
Global Lead, Telco Media, Enter- telecommunication and media industry. He
prise Transformation currently leads the development of enterprise
IBM Global Business Services transformation offerings to support sales and
linkedin.com/in/warwickhillibm/ delivery engagements globally, including
[email protected] automation and AI solutions.

Martin Guillaume At IBM, Martin leads Lead-to-Cash Cognitive

Process Services, focused on driving digital
Global leader for Sales Order business transformation in telecommunications
Management and media organizations globally. His technical and
IBM Global Business Services business experiences are vast, including work with
linkedin.com/in/martinguillaume Ericsson, Atos, Goldman Sachs and Sony.
[email protected]

Jim Thornhill James Thornhill is a telecommunication industry

subject matter expert with extensive experience in
Associate Partner shaping and delivering transformation programs
IBM Global Business Services that address growth, digital, customer experience
linkedin.com/in/james-thornhill- and cost challenges. James has a particular focus
82b935 in bringing AI and automation enabled solutions to
[email protected] clients globally.
Transform telecom processes,
driving more predictable business
outcomes, greater productivity
improvements, faster order cycles,
quicker speed to value and
profound cost reduction.

Radical change ahead

Talking points Today’s telecom operators are squeezed to grow
network capacity to meet amplified customer demands,
We’re at a key inflection point, moving yet pressed by shareholders reluctant to direct spend in
from a world of processes run by humans areas other than the network. In a hugely competitive
global industry, operators must balance ongoing
supported by technology, to processes
customer satisfaction against reducing operating costs.
run by technology supported by humans. Too often, subscale technology investments are made
A customer centric, agile design married for meager benefits, and automation is bolted on
cumbersome processes supported by decade-old
with an intelligent automation platform will
systems. Worse still, automation is often implemented
transform telecom processes. without revisiting the underlying customer experience
or evaluating what artificial intelligence (AI) could do to
improve productivity. We propose bringing together AI
Replace tedious tasks and robotic process automation (RPA) to create cost
Robotic and cognitive process automation savings, increase output and deliver more value to
used in combination can reduce excessive telecoms.

costs tied up in low value processing.

Reimagining the customer first
Deliver meaningful change in just months The traditional lead-to-cash process is a key enabler
for telecom imperatives, including a positive customer
Business benefits are amplified beyond experience at the point of sale, rapid response to change
what is achievable by over extending a and reducing costs. It’s often highly complex, internally
single technology. focused and rigid. Closely coupled with multiple legacy
systems, lack of transaction visibility and long cycle times,
the traditional process often leads to poor service and
customer dissatisfaction. Slowed by the process,
customers are disincented to stay loyal, looking for
a better offer elsewhere.

A customer-centric, agile design married with an

intelligent automation platform can transform telecom
processes, driving more predictable business outcomes,
greater productivity improvements, faster order cycles,
quicker speed to value and profound cost reduction.

The telecom industry runs rife with highly manual,

voluminous, repetitive and complex rule-based
transactions – things such as order validation, service
fulfillment, service assurance, billing, revenue
management and network management. Process
automation can transform mundane but time-consuming
and error-prone business tasks to automated ones and
streamline for simplicity. We’re at a key inflection point as

1
we move from a world of processes run by humans –– Cross-selling and up-selling through assisted sales and
supported by technology, to processes run by technology creating recommendations so that all sellers know what
supported by humans. Advancements in AI have created the best sellers do.
intelligent automation, changing the way enterprises
operate by using machine learning to optimize processes Robotic and cognitive process automation used in
and personalize the customer experience. combination can reduce excessive costs tied up in low-
value processing and drive greater value (see Figure 1).
For telecoms laden with manual processes, moving These automation and AI skills are best developed and
beyond basic robotics to intelligent interactions, or RPA, delivered through a Center of Excellence, aligned with the
can replace tedious tasks and deliver cost savings and company’s Shared Service Organization.
greater workforce productivity by:

–– Creating smart processes to automate back-end The right tools for the job
activities, such as data entry, validation, and
Entirely new user experiences can be achieved by taking
reconciliation.
an over-the-top (OTT) approach. The idea is to preserve
–– Enabling customers to self-serve so that sellers can the legacy system’s (also known as systems of record)
focus more time on complex orders. capabilities to be the custodians of the business
transactions. By interfacing with the “systems of record”
through existing APIs or microservices, one can redefine
the user experience more freely and use a combination of
capabilities ranging from business rules engine, business
Figure 1 process and management (BPM), AI, RPA and blockchain.
Exact costs with intelligent back-office automation

Knowledge vault
~20–30% of processes
Intelligence

Legacy middle
~40–50% of processes

Augmentation
Business process value

Collaboration Clerical long tail

~20–30% of processes

Automation
Robotic automation

Knowledge work Enterprise IT Clerical processing

Specialized Generalized
Degree of specialization
2
Where two or more systems work
together without being directly
connected in an IT environment is a
state of decoupling. This allows for
the creation of modern digital
experiences and justifies future
investment in transformation.

Using these tools in conjunction, and playing to the

strengths of each of them, business benefits are amplified
The model for change
beyond what is achievable by overextending a single Large business support system (BSS) and operations
technology. This enables transparent and flexible support system (OSS) transformation programs have
automation of the business process in response to a poor track record in the telecom industry. BSS
business needs. Technologies embedded with RPA can transformations can take years and cost millions;
provide autonomous decision making, enable reasoning 68 percent of telecom operators are aware of large
and remembering, and provide new insights and data transformation failures, and 40 percent have seen
discovery. For example, using AI and RPA technologies as projects fail at a cost over USD 20 million.1 Operators
part of a sales order management process can guide the have built up a complex landscape of legacy systems
seller to improve data accuracy by making and processes that no longer serve the agile world and
recommendations that improve over time (see Figure 2). that are arduous to overhaul. Overruns and doubled
A structured approach provides a cohesive process budgets, program failures and elongated processes
automation platform with documented components, plague telecom change. More than half, 70 percent,
methods, and tools in which to reengineer the experience of digital transformation projects are expected
and implement an automated lead-to-cash process. to fail.2
The cohesive process platform is loosely coupled to IT,
interacting with existing IT without having to alter or Our focus is on what works, and an alternative way to
reengineer it. The aim is to provide a uniform and drive an optimized business process. Providing short-
enhanced experience to users and leveraging full term, radical productivity gains is possible. It also creates
potential of a proven portfolio of legacy capabilities. the breathing space, defined by time, resources and
budgets, IT organizations need to make fundamental
technology investment decisions and execute their
system transformation strategy.

Figure 2
Decouple existing applications from customer experiences

Seller Partner Customer

Guided seller Customer

Dashboards Blockchain
applications self service

Cognitive process automation platform

Data standardization Work assignment Virtual agent
and validation and monitoring

Business process Single portal Robotics

rules management
.
Microservices Robotics APIs

Existing application estate

Order
Pre-sales Sales Pricing Contracting Ordering Billing
management

3
Developing technical debt Change doesn’t have to happen in one fell swoop, but
rather in incremental sprints. Shifting from traditional
multi-year IT transformations to an agile delivery
In the rush to automate, the role of robotics is
approach and Minimum Viable Products (MVPs), telecoms
being overused and overextended to cover not only
can power new experiences in small steps within a flexible
repetitive tasks, but the codification of business
and coordinated roadmap. Consequently, with technology
process and rules. This is especially true where
increasingly becoming consumable “as a service,” the
RPA is coded up by users with little to no experi-
costs to iterate ideas are coming down dramatically.
ence of business architecture and software engi-
Delivering meaningful change in just months is possible.
neering principles. Complex code-bases are being
There are many dimensions that form an MVP, including
extended with RPA often done in haste. Future
batching pain points, automation platform functionality,
agility is being sacrificed to deliver short term
process steps, products and market segments. For
efficiency.
example, an MVP might:

–– Take order requests from various structured, semi and

unstructured sources such as email text, tables, and
electronic spreadsheets or PDF attachments to
Thinking blockchain interpret contracts automatically.
–– Read and understand inbound customer requests such
in telecom as text, tables, and electronic forms, as well as email
attachments about orders, disputes, complaints or
Since many telecoms deal with billing and finan- service requests.
cial settlements globally, the ability of blockchain
–– Standardize, validate and enrich this input and perform
to handle complex transactions across multiple
a triage to help make sure that the most suitable back-
participants is a natural fit into an automated
office agent automatically processes the transaction.
process platform. Blockchains can increase trans-
action speeds by reducing clearing and settlement A process automation platform that sits “over the top” of
time, and more than one-third of CSP execu- existing IT can achieve radical cost savings, faster order
tives say they are already considering or actively fulfillment, lower operations costs and vastly improved
engaged with blockchains.3 And 35 percent said productivity. This platform interacts with IT but doesn’t
blockchains will help them simplify and automate require significant change. It lets telecoms design the
business processes, as a blockchain provides the customer experience they want, then implement the
opportunity to rationalize various aspects of their transformed process to support that experience.
operations.4
Since intensive manual intervention isn’t required, the
benefits of creating the process and experience supported
by a process platform have positive impacts on
productivity, cycle time, cost and customer satisfaction.

4
Transform processes beyond
lead-to-cash and deliver bold and
better business outcomes based
on a customer-centric design.

Potential to massively reduce 3. Create a center of competence that includes design

thinking approaches and build AI and robotic content
operational expenditure libraries and extensive industry-specific process flows
and business rules. Consider AI process platforms to
Faster order fulfilment at half the cost, more self-serve automate perceptual and judgment-based tasks
options so sellers can pay more attention to complex through the integration of capabilities such as natural
orders and intelligent lead-to-cash processes that steers language processing, machine learning and speech
the back-office towards more areas of value are key to recognition.
change for telecom (see Figure 3).
How do you get there? We propose the following areas
of focus:
1. Build a business case with the line of business or
shared services that will see the greatest gains and
Key questions to consider
align incentives to make collaboration happen. Start
» Which areas do your customers need
with relieving pain points and improving the user
experience. And start early in the business process to to address first? Cost reduction?
optimize the benefits of data quality improvements Productivity improvements? Order
downstream and reduce fallout.
fallout rate? Order cycle speed?
2. Decouple legacy from a new user experience. Don’t
Speed to value?
rebuild IT but breathe new life into interactions with
older, legacy applications to provide much-needed
budget relief. » Where could you decouple existing
applications from process management
to create completely new user
experiences?
Figure 3
Example of operator transformation business case
» Do you understand the need for a
structured approach, and simply
Q1 Q2 Q3 Q4 Q1 40%
not just diving into using robotics
alone for everything?

20%

0%
150 400 500 550 600
users users users users users

10% 50% 75% 100% 100%

volume volume volume volume volume
coverage coverage coverage coverage coverage
About Expert Insights © Copyright IBM Corporation 2019

IBM Corporation
Expert Insights represent the opinions of thought New Orchard Road
leaders on newsworthy business and related technology Armonk, NY 10504
topics. They are based upon conversations with leading Produced in the United States of America
subject matter experts from around the globe. For more January 2019
information, contact the IBM Institute for Business Value IBM, the IBM logo, ibm.com and Watson are trademarks of
at [email protected]. International Business Machines Corp., registered in many
jurisdictions worldwide. Other product and service names might
be trademarks of IBM or other companies. A current list of IBM
Notes and sources trademarks is available on the web at “Copyright and trademark
1 “BSS transformation doesn’t have to be as hard as you think information” at: ibm.com/legal/copytrade.shtml.
it is.” Openet. July 2016. https://www.openet.com/about-us/
press-room/news/bss-transformation-doesnt-have-be-hard- This document is current as of the initial date of publication and
you-think-it may be changed by IBM at any time. Not all offerings are available
in every country in which IBM operates.
2 Davies, Jamie. “70% of digital transformation projects expect- THE INFORMATION IN THIS DOCUMENT IS PROVIDED
ed to fail.” Telecoms.com. February 2017. http://telecoms. “AS IS” WITHOUT ANY WARRANTY, EXPRESS OR IMPLIED,
com/479452/70-of-digital-transformation-projects-expect- INCLUDING WITHOUT ANY WARRANTIES OF
ed-to-fail/ MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
ANY WARRANTY OR CONDITION OF NON-INFRINGEMENT. IBM
3 Alexander, Mike, Nick Gurney, Utpal Mangla, Mathews Thomas products are warranted according to the terms and conditions of
and Rob van den Dam. “Reimagining telecommunications the agreements under which they are provided.
with blockchains.” IBM Institute for Business Value. January
2018. https://public.dhe.ibm.com/common/ssi/ecm/gb/en/ This report is intended for general guidance only. It is not intended
gbe03901usen/reimagining-telecommunications-with-block- to be a substitute for detailed research or the exercise of
chains_GBE03901USEN.pdf professional judgment. IBM shall not be responsible for any loss
whatsoever sustained by any organization or person who relies on
4 Ibid. this publication.

The data used in this report may be derived from third-party

sources and IBM does not independently verify, validate or audit
such data. The results from the use of such data are provided on
an “as is” basis and IBM makes no representations or warranties,
express or implied.

64023264USEN-03