As experienced transformation experts and one of the greatest advocates of software automation over the last 10 years, T-Impact have become synonymous with RPA in a number of industries.
This guide is intended to help businesses navigate the subject of RPA, as they identify relevant applications to their organisation and build the business case for their first RPA project.
RPA is the abbreviation for Robotic Process Automation and it involves using software (the “robot”) to perform tasks at the user interface (UI) level normally carried out by humans. These tasks need to adhere to a set of rules and are typically highly repetitive.
If it can be done by a human and follows specific rules, then a robot can almost certainly do it. They can copy and paste data, log into applications, extract and process content from documents and emails, and make calculations.
• RPA can assess IT systems, extracting & updating data via screens.
• RPA can communicate with humans using email, text message, letters and work flow systems
• RPA can be configured with business rules to apply human type logic to activities, e.g. deciding when to approve/decline things, knowing who should be notified, tracking status against service levels
• RPA can be combined with AI to have natural language conversations with humans
The examples depend on the industry in question (see specific use cases at the bottom), but some common sector agnostic use cases include:
• Payroll processing
• Customer on-boarding
• New enquiry processing
• CRM updates
• Order processing
Reduction in cost
Improvements in quality
Enhanced employee engagement
Reduction in cost
Robots are able to work 24 hours a day, 365 days a year. They don’t call in sick or waste time chatting by the water machine. Put simply, for certain tasks they are far more efficient the humans, enabling more to be done at less expense.
Improvements in quality
Robots don’t make mistakes. Human error is entirely eliminated from any task that is fully automated (assuming, that is, that it’s been configured correctly!).
Greater compliance – the enhanced consistency and accuracy makes it far easier to satisfy regulatory requirements.
Enhanced employee engagement
The major fear of RPA is a resulting loss in jobs. However, while that is a rational expectation, in practice the vast majority of companies investing in RPA have not only avoided large scale redundancies, but also found more valuable and rewarding ways to use their staff, improving employee experience and engagement.
One of the challenges of many forms of enterprise technology is its inability to adapt to changing circumstances. RPA, on the other hand, can be modified rapidly, far faster than a human workforce.
So much of business strategy nowadays is fuelled by data, but most businesses struggle to leverage this value as their data is inconsistent, inaccurate or inaccessible. RPA doesn’t make mistakes or skip steps, so the data it produces can instantly be fed back into the machine to ensure ever smarter decision making.
Perhaps the greatest benefit of RPA is its ability to scale massively. Traditionally, when a company uncovers a great commercial opportunity, its ability to capitalise is limited by the speed with which it can hire, train and deploy new talent, which will invariably take months if not years. RPA shifts that paradigm entirely.
These are what we call “exceptions”. Ideally tasks being automated would always contain structured data that follows certain patterns and rules. However, even highly repetitive tasks will often contain unpredictable elements. These exceptions are highlighted by the software so that they can be manually reviewed by a human.
This depends on the task being performed but one of the attractions of RPA is that with basic training any employee in an organisation can learn how to use robots to assist in their day to day role. In fact, it is the vision of the major RPA vendors that before long every employee will have their own robot and use it for all of their dull, repetitive tasks.
Unattended automation involves those tasks that can be fully automated using RPA. This depends on the data being entirely (or almost entirely) structured, so that everything can be programmed. There may be a number of “exceptions”, but they too can be built into the rules for the robot to follow.
The trouble with some tasks is that they contain so much unstructured data that it’s virtually impossible to create all the necessary “exceptions”. Instead, the robots will need to collaborate with a human worker, so that the robot automates the structured elements, and the human deals with the rest. This attended automation is sometimes referred to as “Desktop automation”.
This of course places a cap on the speed with which the robot can operate, which is why increasingly vendors are integrating elements of Artificial Intelligence (AI) to replace the human component.
While traditional automation requires significant programming as the developer has to integrate different underlying systems, RPA simply replicates the action of the user at the UI (User Interface) level. In other words, traditional automation is a highly technical endeavour, whereas RPA is able to bypass much of that complexity and still achieve the same results.
Further to that, developers may not have access to the source code which in some instances may make the necessary customisation for traditional automation impossible, or the system may be so old that finding the necessary technical skills could be difficult. Again, RPA overcomes these hurdles.
This relative simplicity also accelerates the “time to market”, which can be vital in building the business case commercially.
Perhaps most importantly, the fact that non-developers can use RPA means that the functions being directly affected by the technology can have a strong input in its design, maximising the likelihood of it solving real business challenges, while traditional automation often becomes silo’ed within the technology department.
There is a world of difference between these terms. RPA is simply a piece of software that mimics human activity, while AI is the simulation of human intelligence by machines.
RPA in one sense is actually very rudimentary technology, which is precisely why it’s so powerful. It is simply a machine copying what a human would do, but faster, more accurately and without stopping. AI, on the other hand, has no limits, and in truth we are barely able to comprehend its potential to transform both business and the broader world. It’s this complexity that means that, for the majority of businesses, the practical value of AI is in many cases less profound than that of RPA. For now, at least.
To confuse matters further, both terms actually sit within a continuum known as IA – Intelligent Automation, in which RPA is the “doing”, while AI is the “thinking”.
Increasingly, RPA vendors are looking at how they can apply elements of AI in order to take greater value from their robots. For example, let’s say that a task being automated has a large proportion of unstructured data. RPA alone will not be able to interpret that data (unless every “exception” is programmed, which in this example we are assuming would be too large a task to be practical). One solution is “attended automation”, in which a human would complete those elements of the task that depended on instructed data. However, that human input obviously places a bottleneck on speed that may significantly reduce the value of the automation. This is where using an element of AI or machine learning can help the process to automate fully, although it will of course add a level of complexity and cost to the project.
An increasingly common phrase now being used by the large vendors is “automation first”. They have seen from implementing thousands of RPA projects around the world how those companies that derive real value from automation, build it into their strategies from the outset. They understand that every function in a business can benefit profoundly from automating repetitive tasks, and freeing up the humans in that department to focus on the things they do best – thinking creatively and adding value to the customer experience.
RPA is now moving beyond task automation to more complex processes and workflows. With RPA as the means of assessing your IT systems, it is integrating AI, machine learning, process mining, digital analytics and other emerging and disruptive technologies to go beyond the simple automation of everyday human tasks, and towards entirely new organisational capabilities that allow ambitious companies to achieve things they previously couldn’t have imagined.
Once the business case for RPA has been proved and the organisation is seeking to roll it out company-wide, there are operational questions to be asked around how this large scale adoption is managed. There are three main options:
The centralised model – the centralised approach is where the company places responsibility in the hands of one single team, usually within the technology function. The benefit of this is that the team has a really strong understanding of RPA and the likely technical hurdles. The team can also ensure certain standards and frameworks are developed. The advantage of this model is that there is a high level of standardisation and control, but the downside is that the solutions designed may not always address the most significant challenges and opportunities faced by the individual business functions.
The decentralised model – this is where power is passed to each of the individual business functions, so HR may have its own RPA team, for example. The upside of the decentralised model is that each of these business functions best understands the most powerful potential use cases for RPA, and can apply the technology accordingly. The downside is that it can result in a lack of standardisation and quality control, with every function doing things that little bit differently.
The hub and spoke model – the hub and spoke approach is the one we typically encourage, where the hub is the centralised RPA unit and the spokes are each of the separate business functions. The idea of course being to find a sensible middle ground, where certain rules are standardised whilst also offering sufficient flexibility for each function to apply the technology in the way that they know will have the greatest possible impact.
In order to embed RPA deep within the organisation, we recommend that companies set up an RPA Centre of Excellence. Ideally the CoE will include all of the following roles:
RPA Champions – the person or people, ideally at a senior level, who evangelise about the benefits of RPA and educate other key influencers within the organisation on the potential use cases.
The change manager – the person who will take the helm from a cultural and people perspective. The person who will articulate to those most directly affected why this is in their interest, to engage and consult them throughout the process and address their concerns as they arise.
Business analyst – the people who understand the processes that are to be automated and how they need to be mapped out in order to maximise business benefit.
Solution architect – the expert who defines the RPA solution required, and oversees the implementation phase ensuring all the right tools and features are integrated.
Developers – those who actually build the solution, testing throughout and collaborating with the analyst and architect to ensure alignment with their plans and objectives.
Service support – once deployed, the RPA support experts will handle queries and manage exceptions to ensure the full value of the solution is realised day to day.
Changing expectations of employees – As Guy Kirkwood, Chief Evangelist of UiPath put it, “The young are not prepared to put up with the sh** that we put up with in the last 30 years.” The point being that young people entering the workforce are so comfortable with technology, and so conditioned to turning to technology as a solution for all of their problems, that they will not tolerate companies that fail to adapt. They want to be doing interesting, rewarding jobs, and pass the boring tedious stuff onto the robots.
Economic disruption – be it the financial crisis of 2008 or the Coronavirus outbreak of 2020, we live in unpredictable times. Businesses will always exist for and because of humans, but it’s important that business resilience and continuity is not dependent on humans at every minute of every day. RPA offers a chance for businesses to be far more durable, so that when crisis strikes, humans can turn their attention the other parts of their lives that matter, without fear of losing their jobs or the business hitting the wall.
The Amazon effect – it doesn’t matter what sector you’re operating in, consumers have come to expect fast and frictionless transactions with the brands they engage with, so when you’re judging the quality of your customer experience, you need to ensure you’re not only comparing yourself to your direct competitors, but to the likes of Amazon and Netflix. What can you learn from their user journey that you could apply to your own? This attention to detail is only possible if you are using technology to ensure consistency of experience, while freeing your people up to do the things that humans do best – engage with other humans. As of today, RPA is probably the single most powerful technology in facilitating this revolution.
5 years – At the current rate of growth, RPA adoption will become saturated – meaning nearly every company will be using it in some form – in the next five years, according to Deloitte. In other words, RPA will reach near-universal adoption at some point in 2023, Deloitte predicts.
2020 – Gartner projects spending on RPA software to hit $1.3 billion this year.
$2.9 Billion – Forrester has predicted the RPA software market to total $2.9 billion in 2021.
800,000 Robots – By the end of 2020, there will be around 800,000 unattended Robotic Process Automation robots running on cloud and on-premise servers, according to the latest report from the Everest Group.
58% Increased Productivity – The increase in productivity in the manufacturing industry since 2009, largely attributed to the widespread use of RPA, from the Office of National Statistics.
26% Increased Revenues – The increase in revenue per employee seen by companies adopting RPA according to CGI Group.
10% Costs Cut – The cut in enterprise costs seen by companies adopting RPA, according to the CGI Group.
30% Automation – The percentage of all jobs that could be automated by mid-2030s according to PWC.
50% Transport Jobs Automated – The number of transport industry roles that will be available to be automated by 2030 according to PWC.
$6.7 Trillion Global Value – The global financial implications of automation and RPA technologies, according to Mckinsey & Company.
13.6% Market Share – The market share held by our RPA partner of choice and world-leading RPA provider UiPath.
72% on RPA Journeys – Deloitte estimate that by 2021, 72% of their RPA survey respondents will have embarked on their RPA journey.
Less than 12 Months Payback – Due to the efficiency of RPA, Deloitte found that investments in robotic process automation are usually profitable within a year.
Annually there’s a huge spike in the pressures on local government staff as housing associations submit rental increases. These spreadsheets must be handled quickly so that the council has accurate financial reporting and statistics to base decisions upon. Unfortunately, many councils have limited resources to devote to this anomaly.
T-Impact designed a Rental Increase Robot to automate the process with virtually no “exception” conditions. It is able to:
• Confirm that the rental increase is equal or below the approved percentage.
• Apply the new rate to the address, taking multiple service costs into consideration.
• Apply benefit discounts related to the property for eligible residents, and generate letters to confirm outcomes.
• Deal with password protected spreadsheets by checking for an associated email containing the password, or contacting the housing association if they forget to provide one.
Automating rental increases reduces the burden on council workers, eliminates the need for temporary staff and leads to greater security when it comes to financial planning. Staff are freed from repetitive administration to focus on giving taxpayers value for money.
Housing benefit applications, changes in circumstances and council house exchange are labour intensive, with data manually being replicated across numerous IT Systems.
RPA that manage this process can:
• Verify both new claims & changes, consistently applying complex rules.
• Isolate the small percentage of true exceptions for expert evaluation.
• Enter the verified data into Revenue & Benefits systems, such as Capita One & Liberate.
An untold number of work hours are reduced, freeing council workers from mundane administrative tasks and enabling them to devote their time to adding value through direct engagement with citizens. It also eliminates numerous data entry errors that often take days to isolate and resolve.
Heads of Recruitment and Talent Acquisition often have to search through wide pools for candidates, but many people who show interest in opportunities are not actually qualified or do not fit requirements for roles. To avoid wasting time speaking to candidates that are not suitable, recruiters need to streamline candidates to identify quality prospects, rather than engaging every single applicant.
Recruitment companies can use robots to search existing talent pools using predefined criteria to identify candidates that are suitable, as well as eliminating anyone who does not fit the requirements.
Allowing robotics to manage your initial candidate outreach means:
• Greater numbers of candidates can be processed.
• Recruitment teams can spend more time engaging with the streamlined list of quality prospects to find the best candidate for the role.
• Recruitment teams can also use the time that has been freed up to liaise with clients to better understand their needs and advise on their talent acquisition strategies.
When discussing potential new opportunities, candidates want to be able to engage recruitment teams at any and all times. If they cannot do this, or experience downtime between a question or submitting an application, this could reflect poorly on the hiring company and cause a candidate to have second thoughts. However, it’s impossible for a recruitment team to have the capacity to engage and respond to candidates 24/7.
RPA technology can support recruiters by formulating appropriate responses to candidate communications and questions. They can help to pre-screen candidates on key aspects like experience and education, and respond to frequently asked questions. Most importantly, they can respond to C.V. and cover letter submissions and schedule interviews or ask for more information where appropriate.
Using chatbots to support your recruitment means:
• Companies gain the ability to instantly respond to candidate questions and communications, keeping them engaged and interested in the role.
• Interview schedules are controlled by the chatbot to assist HR and senior management teams with time management.
• Real-time management of candidate experience through data collection and satisfaction surveys.
• Other administrative burdens can also be removed from your recruitment process via automation.
When the time comes to make decisions on candidates, the information contained in offer letters need to be specific to candidates and need to comply with all necessary regulations. Although recruiters can find this information in company files, it is time consuming and often stored in multiple systems and databases, increasing the likelihood of errors being made due to the manual nature of input.
RPA software can support the immediate creation and sending of offer and rejection letters. They compile all necessary information from all systems and fact-check it to ensure there are no errors. Once sent, they can also automate requests for any additional checks e.g. DBS Checks and monitor returned documents and responses.
Using RPA to help generate candidate offer letters results in:
• Faster processing for new offer letters.
• Zero percent chance of error in letter generation.
• Immediate application processing for additional checks and on-going response monitoring.
• Recruitment staff and HR staff freed up to begin facilitating the onboarding process.
Central Government systems often contain duplicate versions of the same record held in different departments. Updating these details becomes a tedious process with a high potential for human error, and accumulates hours that add little value.
Using robotics to help manage these changes allows you to update all systems at once via a central hub that works to predefined rules. Once a singular record is created or updated, the robots can ensure these changes are applied across all systems. Automating this process means:
• Zero data entry error which ensures complete legal compliance.
• Operational costs are reduced as staff no longer need to spend time completing this repetitive task.
• Efficiency will increase as robotics can be deployed 24/7.
• Staff members have more time to devote to tasks that add value to taxpayers.
Accountability is paramount in Government organisations, and senior leaders need to be able to quickly and accurately whilst viewing current performance and future projections at all times on any device. Managing this information for human staff is complex and data errors occur frequently.
Robots can be deployed to audit on performance levels at all times, taking into account operational factors and review them against predefined criteria for what is success.
This information is then fed back to senior leaders who can access it on mobile, laptop or email. This results in:
• A completely impartial review of proceedings removing the potential for bias.
• A rounded view of current activity to help for future planning and benchmarking.
• Empowered departments as they have total accountability for their actions.
• Senior leaders no longer have to chase departments for specific information as they can access it all in one place.
• Guaranteed compliance as RPA connects to legislative portals and updates automatically if anything changes.
As Governments embrace digital change, departments are quickly realising they need to amalgamate existing and obsolete systems into an updated IT infrastructure. If they are unable to do this, they either face the arduous task of manually rekeying all the relevant data, or invest in an expensive new system that will also become obsolete in the future.
As an alternative to this, Government departments can deploy RPA to help integrate legacy systems into updated IT frameworks. RPA can speed up the entire process of migrating data and upgrading databases, without needing any human resources to support. Robots can:
• Upgrade old systems that they cannot afford to replace to unlock their virtual workforce.
• Allow governments to remain in full control of their IT infrastructure, which can be scaled up or down depending on future need.
• Achieve flawless data migration with no margin for human error.
• Reduce time wasted on complex integrations to legacy systems, maximising efficiency and operational agility.
Supplier invoices often differ from goods received and agreed costs. Regular review of costs can substantially improve profitability but can be quite time intensive. These manual reviews tend to focus on the top 20% of spend, as reviews of low volume/value goods would cost more than the savings generated.
Suppliers quickly realise which products are price sensitive and make up margin on other products. Staff spending time comparing goods received to invoices or summary reports/spreadsheets could be proactively running competitive tenders to reduce costs further.
Every supplier requires a significant amount of work after they are on-boarded to ensure value for money. Although smaller suppliers may be able to offer goods at a lower price, the management of additional suppliers may require additional staff, which erodes any savings.
Robots can automate this work, making it cost effective to ensure pricing is correct for all goods received. A robot could:
• Download summaries of goods received
• Extracting supplier specific data for reconciliation against invoices
• Verifying quantity and price of goods received against supplier invoices
• Cross-reference unit costs against supplier catalogues
• Reconcile invoices and credits notes
• Generating queries to suppliers and finance for resolution
For one major manufacturing & retail firm, we identified £600,000 benefits from invoice credits and potential cost reductions for additional competitive tenders over a 12 month period. There were additional unquantified benefits from resources freed from this manual and tedious work.
Depending on the number of payment methods supported, manually processing customer payments for goods consumes a significant amount of time and resources within financial teams.
• Most finance teams process payments once or twice a day. Increasing the payment processing frequency could require additional staff, which may not be warranted by the additional income.
• Failure to process payments regularly negatively impacts cash flow and liquidity.
• Delays in accounts receivables can affect sales revenue, as customers who have reached their credit limit are likely to buy from your competitors.
• Customers need to receive regular reminders, as well as being able to easily determine the net amount owed to businesses to ensure they can maintain positive cash flow.
Robotic Process Automation can automate every aspect of the accounts receivable process:
• Extract data from different locations and systems, collating them to make decisions based on a defined set of rules and conditions.
• Match check, cash, bacs & credit card payments against debtor invoices & credit notes, to identify payments that have already been processed.
• Apply payment type specific business rules, e.g. cash payments are only applied to exact match, BACS payments applied if payments within defined tolerance.
• Ensure reminders aren’t sent to customers who have already paid, improving customer relationships.
• Sophisticated “natural language” matching rules, with flexibility to add additional rules later, considered (a) age of debt, (b) precedence, (c) amounts of payment, (d) amounts of invoices, (e) amounts of credit notes, (f) tolerances.
• Update finance system to reflect invoices paid and credit limit released.
• Confirm that bank deposits are made by the end of each working day.
• Create a cash report and send it to credit control.
Automating these processes ensures greater accuracy in accounting and finance reporting and works almost three times as fast as the human workforce. By eliminating manual, repetitive and tedious work from daily schedules, not only is money saved, but staff members are empowered to focus on more creative, strategic and customer oriented tasks that drive real value into the business, such as financial analysis or collaborating with project teams on new product launches. Furthermore, increasing customer communication improves retention and drives new cross-sale and up-sale opportunities.
For one company, the estimated ROI for this amounted to £454,200 per year, with a full ROI after just 3.2 months, which doesn’t include the benefit of having existing resources redeployed to perform higher value tasks.
Excess stock increases operating costs and wastage. Your customers want goods delivered “Just in Time” (JIT) for their own needs. As sales lead times continue shrinking you need to continuously optimise design, materials procurement, production & shipping.
Planning and scheduling software can help monitor & reduce inventory but are only as accurate as the data entered. Solutions that consolidate multi-warehouse stock, incorporate vendor managed materials & auto-replenishment can be quite expensive to purchase and maintain. Your procurement team are working to reduce you materials costs but cheaper suppliers may require your team to invest more time managing your own supply chain.
The complexity of the supply chain network has increased with globalisation. Many businesses rely on a vast network of manufacturers, suppliers, distributors, intermediaries and retailers to source, distribute and sell their products. This leads to a number of stock management challenges:
What a robot can automate:
• Monitor emails to identify new/changing customer demand and update your manufacturing planning systems , using existing screens – no changes are required to IT systems & websites
• Send emails & texts alerts to your to confirm their needs immediately before their production run begins – updating your planning systems if they increase quantities
• Send emails & texts alerts to notify customer when goods will be shipped & expected time of arrival
• Monitor emails from your suppliers’ and supplier websites to confirm shipment date, quantity and specifications for materials – updating your planning systems to minimise startup & changeover delays
• Monitor production line equipment, updating planning systems and sending real-time notifications for stoppages, stabilisation issues & low yield – particularly important for unattended (lights-out) operation
The benefits include:
• Increase sales revenue with real-time refinement to customer order quantities JIT for production runs
• Optimise your planning system’s performance by ensuring data is entered in real-time, regardless of how it is received
• Optimise your production line efficiency, ensuring materials are available when required and quickly resolving production issues
Supply disruptions account for 31 percent decrease in profitability (Source: Georgian Institute of Technology 10-year study). Imagine what you could achieve for your business with a 31 percent increase in profits!