3. Page 3
BACKGROUND & PURPOSEPAGE 3
Robotic Process Automation (RPA) is robotic software or AI which operates on the user interface (UI). Unlike more traditional types of
automation which would operate on subsurface Application Programming Interfaces (APIs), RPA replicates the actions of a human being and
can be applied to any algorithmic task.
UK mobile communications provider
Telefónica O2 deploy more than 160 robots to
process between 400,000 and 500,000
transactions each month, yielding a three-
year return on investment of over 650%.
More astounding, Telefónica O2 reached this
scale by training only four people. Another
case study subject, a large UK-based utility,
deployed more than 300 robots that process
three million transactions per quarter,
yielding an annual return on investment of
200 percent. Here, two humans orchestrate
300 robots that perform the work of 600
people.
• Related investments have grown exponentially
from $1.7 billion in 2015 to $23 billion in 2016
• 21 suppliers on preferred list of major
consultancies
• 80% cost reduction and 80-90% faster in BPO
• £1.5m savings delivered in 2013; similar amount
realised by business owners over and above this.
• Potential $1.2 trillion market - KKR
Key Dimensions:
1. Code free RPAs do not require programming knowledge and use a
system similar to MS Visio to automate tasks
2. Non-disruptive to core system by operating in the UI
3. Business user friendly – ease of use and low economic threshold
Fast Facts
Case Study
RPAs are rapidly being brought to market by start ups as well as
traditional companies and organizations seeking to maintain their market
position.
According to Harvard Business Review, most groups adopting RPA have
promised their employees automation will not result in lay offs. Instead,
humans are redeployed to do more interesting work …Ops Team??
Following the findings of the RPA industry conducted by RGP, our
understanding of practices with other competitors and recent
discussions with clients, we have put together this document to tender
points for consideration and explore areas where we may be able to
support implementation of RPA specifically regarding Financial and
Accounting operations.
4. Page 4
KEY DRIVERSPAGE 4
Shift in focus from labour arbitrage to automated service delivery
Ability to expand margins
Ability to handle niche, complex, and low-volume tasks in strategic transformation initiatives
Can reduce F&A costs by 15-30% (with labour arbitrage, total savings can be as high as 65%)
Alternative to offshoring (about 1/3 the cost of an offshore ETF)
Huge potential in record to profit
KEY DRIVERS OF RPA
ADOPTION
5. Page 5
STRATEGYPAGE 5 Developing an RPA Strategy
What?
• Assess for automation opportunities
• Which processes are good candidates for optimization?
• Which processes would be suitable to pilot?
• How should the process owners be engaged to try automation?
• What are the impacts of proceeding with the pilot?
Why?
• Build your business case
• Why does automation support your business needs?
• What are the benefits?
• What are the pain points being alleviated?
• What are the metrics to determine whether automation is valuable?
• What is the strategy for re-deploying existing resources after automation?
How?
• Determine the optimal operating model
• Which operating model works best for your organization?
• Do you have the right team to support the solution and carry out responsibilities?
• Who will manage and monitor the robot?
Who?
• Identify your automation partners(s)
• Who are the main vendors in the RPA space?
• Who are the providers who cater to your business needs the most?
• Which sourcing option do you want?
• How should you compare the pricing models in order to understand what you are paying for?
When?
• Plan the automation roadmap
• How long should your pilot be?
• What are the stages after the pilot?
• What is your strategy for scale?
• How will you ensure impacted stakeholders understand the what, why, and how of automation?
Key Dimensions in Platform
• Data: the level of sophistication in dealing
with business data (structured v.
unstructured);
• Type of tasks predominately performed:
tasks are either rules based or require
knowledge from multiple sources to
complete the process.
• Interoperability: working across multiple
applications (single application or
multiple applications and platforms), and;
• AI: the level of artificial intelligence
provided by the application ( none,
machine learning based on pattern
recognition and statistics, or emerging
true AI).
6. Page 6
AREAS OF IMPACTPAGE 6
To identify processes suitable for RPA use this checklist:
Rule based (not dependent on human judgement)
Functioning and stabile
Initiated by a digital trigger and supported by digital
data
Larger volume is better
Proof of concept, key that the system leverages key
systems in company
9. Page 9
AREAS OF IMPACTPAGE 9
BANKING
• New account entry across systems – moving data and doing multiple entries
• Account reconciliation – duplicating and moving data
• Report generation across systems
• eForm extraction – taking data from forms and making systems entries
• VAT reporting
• Accrual support – making and updating entries
• Mortgage approval – moving data place to place and making calculations
• Refinancing entries – making entries and updating records
• Notification of delinquent accounts – emails and letters to clients
• Processing of credit card orders – from web form to mainframe
• Audit support and validation
• Fixed asset amortization
• FX accounting and write-offs
• Cost accounting and purchase orders
• Bad debt and write-offs
• Month end close heavy lifting
• Journal entry accounting
• Credit note accounting entry
• Accrual booking
• Pricing reviews – performing calculations and entries
• Fraud detection – by tracking account activity
• Account cleansing – purging old known data
10. Page 10
AREAS OF IMPACTPAGE 10
CAPITAL MARKETS
• Client onboarding – documentation requirements check
• Compliance, legal, credit, and identification checks
• Operational set-up of account in trading, collateral and settlement systems
• Transparent status of onboarding process
• Consistent data mapping between source and downstream systems
• KYC/AML authentication: collection and analysis of basic identity information (CIP); name matching against lists of
known parties (such as PEP)
• Fraud detection – determination of the customer’s risk in terms of propensity to commit money laundering, terrorist
finance, or identity theft
• Reconciliation – user defined rules to generate alerts on chargeback, retrieval request and response
• Automated scheduling of reconciliation activity
• Force match option facility to manually settle un-reconciled entries
• Multi-level reconciliation that supports international and domestic networks across multiple delivery channels
• Entitlement engine – providing timely, accurate distributions of dividends and interest
• Receive/track dividend, interest, and amortized principal payment information and collect the distribution of all
dividend and interest allocations
• Simplifies/enforces the management of complex entitlement policies
11. Page 11
AREAS OF IMPACTPAGE 11
INSURANCE
• Pension auto enrolment assistance
• Product administration – update systems and produce client details for sales of new products
• Policy document data transfer from legacy system to new applications, systems, 3rd parties
• Products sales support – customer data gathering, adherence to compliance and rules of sale
• Broker CUE (motor) – compares customers reported accident claim history with actual data
• Bulk payments (motor) – updates policies following claims made through specialist claims companies
• Bulk recoveries (motor) – updates policies of vehicle recovery amounts and fees supplied by 3rd parties
• Unallocated cash report – works a report and writes off any cash discrepancy under a certain value
• Deceased notification – processing of insurance policies following notification of a death
• Insurance comparison – downloads leads from internet comparison providers and loads entries to lead management
system
• Compliance reporting automation – automated preparation of reports for the PRA
• Marketing campaign support – high speed response for back office automation to support new product launches,
product incentives
• Credit card rejections – works a report of rejected card payments and over a period sends a letter to the customer
before ultimately rejecting insurance
• Credit card unmatched – retrieves cases to work from a report and, where applicable, writes off any cash discrepancy
• Redemptions processing – cross checking multiple systems for compliance assurance
• Payment transfer – paid directly to a client or reinvested/distributed to other products across multiple systems or 3rd
party applications
• Right of set-off and debt recovery
• Funds transfer – from multiple legacy policies/system into a more modern, flexible system
• Amend standing order, direct debt, address details across multiple systems
• Accounts data mismatches across disparate systems
12. BENEFITSPAGE 12
Productivity
and Efficiency
Accuracy
Scalability
and Flexibility
Compliance
Business
Continuity
Cost Savings
RPAs have the potential to replace
humans in any task which can be
transcribed to an algorithm:
• Transaction processing
• High-volume data entry
• Repeatable, computer
centric processes
• In-system upgrades
• Concurrent or duplicate
data entry
Less expensive than off-shore full-time equivalent (FTEs);
reduces headcount, increases throughput
Can be faster than human FTEs with 24x7 functionality
Minimizes manual processes, reduces errors
Replication across geographies/business units increases
scalability/flexibility
Data and processes can stay onsite, requirements can be
embedded into automated rules
Enhanced disaster recovery/processes with ability to switch
to alternate servers
Business
Continuity
Security Reduces internal risks to data and processes
13. Page 13
PRICING MODELSPAGE 13
Most commonLicensed based
• Pay per software license for each installed robot, management server, and development tools
• Perpetual license or annual subscription
• The definition and capacity of a ‘robot’ can vary by vendor, making direct comparisons tricky
• Hardware and maintenance will add to cost
Second most commonValue based
• Pricing linked to either the FTE-equivalent savings (e.g. fixed percentage of the FTE cost that would have been occurred), or to each
completed transaction
• Can be restrictive to horizontal-scaling across the organisation, as contracts will need re-evaluating to include additional business
processes
• The vendor is encouraged to put ‘skin in the game’ and maintain a good level of service
Least commonService based
• You pay a regular subscription fee for the service, with a service agreement that defines the responsibilities of the provider
• This model is particularly attractive for AI solutions which may run on complex big data technologies that can be expensive to set up and
maintain in-house, or are needed on a spot basis
Pricing Models for RPA Adoption
According to Accenture RPA costs on average 1/3 the cost of a
globally sourced resource
14. Page 14
PRICING MODELSPAGE 14
RPA Pricing Model Overview
*Fixed cost data is from Amazon Web Services, project and robot costs are from Virttia*
• Due to intra industry competition, firms
selling RPA software are sacrificing the lion’s
share of profits for market share
• Due to intra industry competition, complex
pricing models are not disclosed.
Additionally, ‘price per bot’ is available only
upon request
15. Page 15
COMPETITIVE LANDSCAPEPAGE 15
$-
$10,000.00
$20,000.00
$30,000.00
$40,000.00
$50,000.00
$60,000.00
US UK Aus Phillipines India RPA
Per Annum Cost Comparison of Back-Office worker to
RPA in USD
Per Annum Cost Comparison of Back-
Office worker to RPA in USD
*
*RPA including only one software license and maintenance. Excluding cost of infrastructure and implementation
It is important to note that this analysis does not take into
consideration cost savings resulting from 24/7/365 operational
capability of RPA, RPA overhead cost, and back office worker benefits.
16. Page 16
TOP SUPPLIERSPAGE 16
Top Suppliers
Arago
•Outcome based
pricing: based on #
tickets processed.
Typically price of
50% of ticket cost
(self disclosed or
set by independent
industry analysis)
•Traditional
software license
(ask for details)
•No specific pricing
•Est. cost saving 30-
50%
AutomationAnywhere
•$2,500 premier
•Single user
•$5,500 integration
•Email integration,
web server
integration, OCR,
or terminal
actions
•Available here
BluePrism
•Publicly traded
•£191M Mkt cap
•1/3rd the cost of
globally sourced
agents
•Price dependent
on number of
BOTs
•Approximate cost
of $15,000 – no
specific pricing
available
Cicero
•$300 one time user
(capterra)
•20% reduction in
costs
•Traded as penny
stocks
•No specific pricing
offered by cicero
online
GoogleBrains
•£400M of company
DeepMind in 2014
•Has been applied to
android voice
recognition and
google street view
but no official rpa
product
released…yet
IBMWatson
•Offers several
different BPM
products
•IBM also offers
financing options
and special offers on
RPA (BPO) software
InStream
•No info available
online for pricing
•Inquiries available
via online
•Products
IPSoft
•Amelia and Apollo
•Amelia – full
service
•Apollo
•20% cheaper than
Amelia
•No specific pricing
available online
17. Page 17
TOP SUPPLIERSPAGE 17
Top Suppliers Cont.
Jacada
•Desktop and
server
automation
services
offered
•Pricing based
on number of
seats
purchased,
no specific
info available
via web
•Cost
reduction of
15-25%
•Services
Kryon
•LEO Suite
offers three
different RPA
solutions
•No specific
pricing
options
available
online
NICE
•Cost
reduction up
to 50%
•No specific
pricing
information
available
online
•RPA services
Openspan
•$800 per
runtime lease
(3rd party)
•Acquired by
PEGA this
spring
•Services
UiPath
•$3,990 per
year per user
•Four different
products
available
•Back office,
front office,
and
enterprise
solutions
available
WIPRO
•Cost
reduction up
to 50% (citing
case studies
in Telecoms)
•No specific
pricing listed
online
•Specific
products
offered here
18. Page 18
COMPETITIVE LANDSCAPEPAGE 18 RPA Competitive Landscape
Level 4 – Multiple decision making
•Multiple sources of data used to gain intelligence on a subject
•Statistical based learning
•Natural Language recognition & understanding of meaning
•No “personality”
Level 3 – Pattern based decisions
•Pattern recognition learning
•Works with unstructured data
•Self-learning with human aid
•Limited decision making based on information provided
Level 2 – Structured rules
•Sophisticated macro application
•Works across any Windows or web based application without
connectors
•Workflow, rules based, structured data, no decision making
Level 1 – Basic automation
•Single application macro application
•Predefined connectors into other applications
21. Page 21
COMPETITIVE LANDSCAPEPAGE 21
1. You need an army of robots to make RPA worthwhile
Fact: one of the main attractions of RPA is the ability to
automate the “long-tail” of low volume or low value
processes that would not be economical to address via
other means.
However, there is still a minimum scale required to
realize both the return on the up front investment of
establishing automation, and the ongoing overhead
costs. It is also common industry practice for vendors to
specify a minimum number of license to purchase.
Though usually not prohibitive, tools aimed at the
“enterprise automation” may not be cost effective for
smaller operations.
In sum, it is quite possible in today’s market to see
positive results with the deployment of a few bots.
22. Page 22
COMPETITIVE LANDSCAPEPAGE 22
2. Robots will take our jobs
Fact: John Maynard Keynes predicted the future held mass
unemployment as a result of automation. While this may not
have played out as quickly as he expected, recent forecasts
have seen more dire predictions; notably a 2015 report by
Gartner reporting one in three jobs will be automated by 2025.
At present, increased automation has not necessarily led to a
loss of jobs. Instead, it has resulted in workforce augmentation.
By freeing up a person’s time, employees can now focus on
more engaging tasks, and overtime organizations could see
lower turnover, higher morale, and increased internal
innovation.
In sum, the robots have not taken our jobs…yet. Technological
breakthroughs in AI have the potential to drastically reduce the
value and demand of a human worker.
23. Page 23
COMPETITIVE LANDSCAPEPAGE 23
3. Robots can now think like humans
Fact: neither RPA or AI solutions can replicate human
reasoning. RPA software bots can mimic the behaviour of
humans in the way they interact with UI, but in doing so they
are following highly methodological instructions and simple
conditional logic.
AI solutions are underpinned by cognitive technologies, which
are increasingly capable of human abilities such as
understanding natural language and recognizing images, they
can even learn by observing humans. Nonetheless, AI requires
training from humans and cannot yet replicate human
reasoning.
24. Page 24
COMPETITIVE LANDSCAPEPAGE 24
4. BPOs are doomed
Fact: RPA tools add to the traditional transformation options
rather than replacing them. In fact, BPO providers could
benefit greatly by adopting RPA solutions by lowering costs and
increasing efficiency. BPO providers could potentially become
third party leasing agents of RPA solutions; these providers also
have the potential to significantly increase their margins by
adopting RPA even if growth slows.
On the other hand, growth of RPA as a transformation option
should affect firm’s outsourcing strategies. In the context of
RPAs the firm and the BPO stand to mutually benefit.
25. Page 25
COMPETITIVE LANDSCAPEPAGE 25
5. Robots are infallible
Fact: while software robots may follow rules without deviation,
do not need sleep or vacations, and will not make typos, they
are prone to their own sources of failure. Reliability is not
100%, poor quality input data can cause exceptions, and while
the ways in which robots recognize the elements of application
UI are robust, they are not impervious to system changes –
particularly when interacting with remote environments.
In addition, robots have no “common sense,” so if a flawin your
organization’s robot management process allows an obvious
error to creep into the instructions provided to the robots, they
will still follow that error – replicating it hundreds or thousands
of times until someone stops it.
26. Page 26
SUMMARYPAGE 26
What a Robot Is
• Computer coded
software
• Programmes that
replace humans
performing
repetitive tasks
• Cross functional and
cross application
macros
What a Robot Is Not
• Walking, talking
auto-bots
• Physically existing
machines
processing paper
• Artificial
intelligence or
voice recognition
and reply software
The digital age is a the next wave
of globalisation and it is about to
hit the shore. The market for RPAs
is hyper-competitive and growing
rapidly. RPAs have the potential to
be adopted (if not already) by
most businesses served by RGP.
Specifically in F&A there are many
areas where RPAs can reduce cost
and boost productivity within the
organization.
One of the most unique and
beneficial aspects of RPAs are
their ease of use and low
economic threshold for
implementation.
27. Page 27Page 27
Page 27CONTACT
Baxter Black
HEAD OF INTERN RESEARCH
W: +44 (0) 20 7422 4173
M: +1 (206) 963 2109
Email: baxter.black@rgp.com
Eldon House, 2 Eldon Street, LONDON
EC2M 7LS
www.rgp.com