Intelligent Automation Week 2018 Adding Velocity to Your IA Program:

• Shopping for Knowledge with RPA Pilot
• Flip to a Broader IA Process Evaluation
• Paint an IA Governance Picture
• Expand Digital Vision across the Enterprise

Shopping for Knowledge with RPA Pilot You Will Learn More from a Pilot than a Conference! Answers to these questions differ for every company and RPA scenario:

• What types of licenses do I need to buy?
• How much RPA scripting can fit on a license?
• When should I run RPA applications?
• How much will this cost in licenses, training, communications?
• How long will it take me get this set up?
• Do I need a recording device?
• What types of access will I need?
• Can robots run on my laptops, or do I need servers?
• What level of skill do I need to program robots?
• What types of processes lend the most impact?

Don’t Overanalyze…Just Get Started! Move through a pilot in four phases:

1. Find a Business Partner
2. Select the Right Process
3. Develop and Test the Bots
4. Move into Production

1. Strong champions are not hard to find these days:

• Find a problem
• Find an influencer
• “Gut feel” what to fix
• Find the team
• Introduce the players

2. “Thunderous rivers” are more interesting than “trickling creeks”:

• Avoid flow charts
• Think “volume flow”
• Think “batching”
• Look for forms
• Ask your staff

3. You don’t need the perfect RPA vendor just yet:

• Path of least resistance
• Go get a trial license
• Set up environment
• Get a few laptops
• Find your “scripter”

4. This will force governance application to broader IA:

• Document
• Train/communicate
• Assign roles
• Point to production
• Showcase and expand

Your learning on smaller RPA pilots will be applicable to building out broader IA programs. Find a high-volume flow…with lots of easy follow-on. Automation Potential Also consider business criticality—less critical processes offer less risk but also less “wow” factor to gain enterprise momentum. Find the right mix for a pilot. More is better

• Are there any forms involved in the processes?
• What percentage of the process is rule-based (logical if… else… then) steps?
• How many applications and data sources does the process need to integrate with?

Less is better

• Has this process changed frequently in the past two years?
• Does IT make frequent changes to the systems involving this process?
• Does your process involve secure data?

Flip to a Broader IA Process Evaluation You’ve Successfully Completed Some RPA Pilots…What’s Next? Your options…

• Declare RPA pilot victory and quit
• Do another pilot, then another…and another
• Start randomly pushing to production as pilots become ready
• Develop/promote an enterprise-wide program

Broader Process Evaluation: Understanding Where to Go from Here Data structure and integrity is essential across all IA. A process breakdown will lend clues to IA opportunity. Process tools can assist with:

• Collecting process step data
• Identifying characteristics that lend themselves to RPA, Virtual Agents, and Artificial Intelligence technology

Build a process opportunity dashboard to help prioritize. A process tool helps:

• Organizations understand standardization in process approach, execution, efficiency, and technology use
• Identify process steps that can benefit from RPA and other IA solutions

Early identification of which process steps present the greatest opportunity for automation solutions will allow your organization to maximize the benefit of its automation roadmap. Paint an IA Governance Picture What Do You Want to Achieve with Intelligent Automation? Answers help define program design and funding requirements. There Is No “One-Size-Fits-All” Governance Model What others are saying (quotes from past SSON conferences)

• Yes, maybe the system should do this, but it doesn’t and it’s not going to because IT is busy, so we use RPA (Walgreens)
• Our IT preferred in-house servers (vs. cloud) for data security, others may not… (Walgreens)
• Standardize and eliminate waste before robotizing (Ascension)
• Processes may not require standardization or optimization before automating (ANZ Bank)
• ROI increases as people begin to think digitally – opportunity becomes increasingly apparent (AA)
• Fish where the fish are – if there are a lot of people working on particular tasks, it’s probably a good place for RPA (AA)
• We set up robots in the system as contractors (McKesson)
• We set up robots as existing workers who had access but did not work the process (Ascension)
• We reinvest our IA savings into IA (Aruna, MasterCard)

The answer is typically some degree of hybrid (but there’s a broad range): Centralized, Hybrid, or Decentralized. Centralized

• Description: RPA launched from within enterprise COE or shared services with high visibility and control of efforts
• Enterprise Role: Almost entirely enterprise-led approach (typically IT, though it could be led by another centralized business function)
• Challenges: Less responsive; development time reduces impact opportunity

Hybrid

• Description: Baseline guidelines, controls and QA; core team provides automation services for areas unable to support own needs
• Enterprise Role: Provides direction and leadership, orchestration and quality assurance, optional process selection/ design, and optional scripting
• Challenges: Potential conflicts between central and local efforts and authority

Decentralized

• Description: RPA efforts are launched locally within business functions; can have business or function-specific COEs
• Enterprise Role: Little/no involvement at enterprise level; informal ad-hoc coordination
• Challenges: Varying quality, potential bot impact on other departments/bots, and robot chaos

Governance – Program Decisions An IA program without central standards and quality assurance is a weaker “program,” but tradeoff may be a reduced speed to implementation. Governance – Execution Decisions IA prioritization and project execution warrants business unit knowledge of needs, but an IA COE can help keep this organized and avoid surprises. Governance – Maintenance and Planning Decisions Running IA programs without IT is a fallacy—IT will need to be involved in maintenance at a minimum. But, an IA COE can serve as a great “head on the horse” to promote IA visioning and planning. Expand Digitization across the Enterprise What Types of Work Will IA Impact? Transactional

• Low-cost transactional centers and BPOs
  • AP, AR, accounting transactions
  • Master data entry and maintenance
  • Customer issue assignment

Functional – Customer Facing

• Service hubs and call centers – “Amelia-like” virtual agent, chatbots
  • Vendor help desks
  • Employee service centers
  • IT help desks
  • Customer service order taking

AI Specialist

• Centers of expertise
  • Credit and risk analysis
  • Fraud prevention and controls
  • Forecasting, planning, and budgeting
  • Tax strategy
  • M&A targeting
  • Compensation design
  • Recruiting and retention
  • Workforce planning
  • Demand planning
  • Spend-reduction opportunity

General Distribution of Back-office Impact Potential Total FTE:

• Strategic (6%)
• Specialists (11%): Deep specialist focus typically associated with decades of experience (COE where talent is)
• Robotic (12%): RPA work fits a small COE for controlling work, developing programs, and maintaining “bots”
• Hands On (16%): Hands-on touching of work, government interface, legal requirements (local, on site)
• Transactional (32%): Repetitive task, often related to admin or routine data entry (low-cost labor spots)
• Customer Facing (23%): In-depth subject matter expertise and closer proximity to customer (regional support hub)

Expanding Digital Vision

• Transactional Shift – Away from Lost-cost Areas?
• Service Shift – Virtual or Smaller Service Hubs?
• Specialists COEs – Smaller, Intelligent, Data Driven?

Impact to Future Work Types Will Change Our Business Cases Future impact to our rationale for existing GBS models may require rethinking and result in completely different delivery frameworks than we have today. How Can You Prepare Now? Preparing the future conceptual model and business case now will avoid the potential for costly initiatives that may not add value in the near future (e.g., call center builds, BPO contracts, etc.). Avoid surprises and create your future model now:

• Find process steps that lend themselves to IA
• Observe process characteristics and align to IA technology (RPA, virtual agents, AI)
• Build the case for investment and a realistic pace for your company
• Prepare your data (governance, structure, cleaning)
• Start formulating a beginner governance model
• Assess upcoming long-term strategic decisions vs. a highly likely future of disruptive technology
  • BPO decisions
  • ERP decisions
  • Service support decisions
  • Location decisions