Banking functions engage in digital and Accenture has already helped clients kick off their digital journey, especially via Robotics. A series of entries is posted to share insights on digital for Banking functions. This one tackles accelerators of Robotics deployment.
Bank functions face the challenge of rotating their data production tasks towards data intelligence and anticipation, aiming better to position themselves as a service provider to business lines. Robotic Process Automation, RPA, helps reduce time to produce data within these functions and it is a first step in digital transformation.
What is RPA? It is a technology that automates human interactions across applications and office software if data is standard and structured. It has a value where tasks are repetitive and rule-based. The leading RPA providers are Blue Prism, UI Path and Automation Anywhere. The technology results in programs performed automatically by virtual assistants.
In a previous blog entry dedicated to use cases in Risk functions, I reached the conclusion that Robotics is cost effective for functions if Robotics is implemented at scale. This brings about the need to accelerate opportunities identification and implementation. What can help functions speed up the implementation of virtual assistants? Both company and consulting accelerators help.
How to identify opportunities early and exhaustively?
RPA aims at automating processes; therefore, an inventory of existing processes is a starting point for assessing process eligibility for RPA. This not only facilitates an early and comprehensive assessment of automation potential but also eases RPA development prioritization based on benefits expected and mutualization of components.
On the contrary, automation scope and savings magnitude might suffer from lack of process inventory; some good candidates for RPA might not be identified in the early stage of the robotics project and efforts might not be optimized if potential for reuse of RPA components is not visible.
How to improve robotic process reliability?
Ability to think of activities performed in terms of processes rather than tasks, knowledge of process performance improvement and process documentation accelerate RPA process design.
Although RPA mimics existing tasks, it is a common practice to modify and improve the underlying process before developing RPA components; this serves the objective to make the process more standard and still flexible enough to use parameters.
Process management maturity among the process owners eases the communication with RPA process analysts when the target robotic processes are designed; this investment in time and effort has a pay-off in the immediate next phases- development and tests – and it limits rework and frustration generated because specifications do not need to be modified again and again.
How to onboard stakeholders?
RPA projects generate a lot of questions relative to technology, security and operational risk and it is paramount to address those questions and fight resistance to change and move forward.
The population targeted by raising awareness in the early stage of the RPA project encompasses Virtual assistants’ users, their management and allies.
Communication towards virtual assistants’ users aims at:
– explaining RPA functionality and eligibility criteria
– illustrating how virtual assistants operate in production: assisted versus unassisted, scheduled or manually triggered, audit track and log understanding
– sharing the project methodology, bringing forward decision milestones and requesting what contribution is expected from them during the main phases.
Communication towards management aims at:
– explaining RPA functionality and eligibility criteria
– sharing methodology to help navigate through the prioritization funnel to the business case landing
– getting the management expectation in terms of saving materialization
Allies are representatives of functions such as IT, Organization, HR, Security, Risk and Communication. Their onboarding is more stakeholder management than RPA awareness raising. The main objective is to introduce the project to them and gather their concerns (new roles to include in job descriptions, virtual assistants’ credentials and access rights, email policy relative to mail sent by virtual assistants, operational and security risks incurred by using virtual assistants) to address them and incorporate remediation in the methodology.
How to raise confidence and accelerate deployment decisions?
Data driven communication favors clarity and sound decision making; this is the reason why eligibility assessment tools must be data driven, user friendly and based on experience.
Following the onboarding step, future virtual assistants’ users are willing to assess their processes eligibility. Experience derived from previous RPA projects helps kick start eligibility assessments; between 30 and 45 minutes should be enough to fill a drop-down list questionnaire that provides eligibility grade, quantitative benefits and total effort to implement.
Accenture developed an online platform Intelligent Automation Diagnostic (IAD ). It provides clients with a facility for assessment of process eligibility for RPA and the benchmarked benefits expected. For the client, this tool brings about a lightened consulting team, a shorter assessment (divide the assessment duration by 3) and a benchmark facility.
How to demonstrate RPA’s promise?
The pilot in the client’s own production environment, rather than a Proof of Concept running on a standalone workstation, is to be started in the first weeks of the RPA project. The objective is to rapidly test and demonstrate the RPA application and relevance to the client’s own context. This first experience considers specific difficulties and helps adapt implementation methodology and size more accurately implementation effort.
Furthermore, users of RPA pilots, or pioneers, take pride in communicating achievements in their own words, which is well perceived by their peers. This fosters a group-wide emulation prone to accelerate deployment, through either peer pressure or RPA components reuse.
Can definition of a target operating model wait?
It is obvious that operating model clarity accelerates deployment because it provides a clear vision of RPA future rules and principles. The period dedicated to implement pilots should coincide with the target operating model (TOM) definition.
The advantage of conducting both pilots and TOM definition in parallel is that the guiding principles for process selection can be updated at once; on the one hand pilots help finetune implementation costs, benefits materialization and selection/disqualification criteria and on the other hand TOM definition helps detail recurring run costs.
The risk of sharing TOM details too late is that run costs are not appropriately factored into the process selection equation.
Does pressure for benefit materialization hinder or accelerate RPA deployment?
A dilemma is experienced by many RPA project sponsors; a positive RPA business case requires savings materialization, yet process owners are reluctant to bring forward opportunities for RPA if this equates to losing their job. So, should sponsorship authorize soft savings to start the RPA momentum and request savings materialization later? Alternatively, should they press for savings materialization early, so that early hard savings subsidize later soft savings?
The Client’s own context can help chose between both options, especially team size, growth and operational cost reduction agenda. Automation of processes operated frequently by many resources and top management pressure on costs push for early materialization, while processes operated in central functions by a few resources, activity growth and baseline understaffing push for redeployment of capacity on other tasks.
Accenture has extensive experience in conducting RPA projects in their early stages – build of RPA eligible process portfolio, definition of TOM and run pilots – which help accelerate roadmap definition and bring value to their clients. Moreover, Accenture is experienced in running RPA projects at scale and in my next post I will share key success factors for sustainable RPA programs.