Maximizing Success in Automation Projects
Download as PPTX, PDF0 likes994 views
The document discusses best practices for successful automation projects based on common failures. It recommends having a strategic vision for automation that aligns with corporate goals. When designing automation projects, metrics should be created that align with goals, input should be gathered from various levels of an organization, and care should be taken not to automate valuable human tasks. The methodology should focus on getting a minimum viable product launched quickly and being prepared to pivot based on feedback. Examples of failed insurance claims automation projects are provided to illustrate best practices.
Maximizing Success in Automation Projects
- 1. Maximizing Success in Automation Projects Sandy Kemsley Independent Analyst Copyright Kemsley Design Ltd., 2022 1
- 2. The current state of business automation Copyright Kemsley Design Ltd., 2022 2
- 3. #fail • BCG: 70% of digital transformation projects don’t meet targets • E&Y: 30-50% of initial RPA projects fail • $260B spent on unsuccessful projects (just in the US) Copyright Kemsley Design Ltd., 2022 3
- 4. The automation imperative Technology makes more automation possible Competition makes more automation essential Copyright Kemsley Design Ltd., 2022 4
- 5. Automation technologies Copyright Kemsley Design Ltd., 2022 5
- 6. Copyright Kemsley Design Ltd., 2022 6 Process automation Business process management Case management Task/decision automation Robotic process automation Service APIs Decision management Intelligent analysis Artificial intelligence Machine learning Process mining Content/ capture Optical character recognition Natural language processing Content management Interaction Chatbots Intelligent agents
- 7. Best practices for automation success Copyright Kemsley Design Ltd., 2022 7
- 8. Strategic vision #goal 1 Make automation a strategic direction Copyright Kemsley Design Ltd., 2022 8
- 9. Strategic vision #goal 2 Pick the right process Copyright Kemsley Design Ltd., 2022 9
- 10. Design #goal 3 Create metrics aligned with corporate goals Copyright Kemsley Design Ltd., 2022 10
- 11. Design #goal 4 Listen to the experts “six levels down” Copyright Kemsley Design Ltd., 2022 11
- 12. Design #goal 5 Automate whatever possible Copyright Kemsley Design Ltd., 2022 12
- 13. Design #goal 6 Don’t automate out people where they add value Copyright Kemsley Design Ltd., 2022 13
- 14. Methodology #goal 7 Get something into production fast (MVP) Copyright Kemsley Design Ltd., 2022 14
- 15. Methodology #goal 8 Be prepared to pivot Copyright Kemsley Design Ltd., 2022 15
- 16. How to tell when you’re failing Copyright Kemsley Design Ltd., 2022 16
- 17. Strategic vision #fail 1 No one cares about the process being automated Copyright Kemsley Design Ltd., 2022 17
- 18. Methodology #fail 2 Waterfall methodology Copyright Kemsley Design Ltd., 2022 18
- 19. Methodology #fail 3 Taking too long and/or over-spending Copyright Kemsley Design Ltd., 2022 19
- 20. Design #fail 4 Fragile/inflexible processes and decisions Copyright Kemsley Design Ltd., 2022 20
- 21. Design #fail 5 Outdated procedures built into new processes Copyright Kemsley Design Ltd., 2022 21
- 22. Knowledge transfer #fail 6 Knowledge of process/tools is not internalized Copyright Kemsley Design Ltd., 2022 22
- 23. Transformation through automation Copyright Kemsley Design Ltd., 2022 23
- 24. Use case: Insurance claims Copyright Kemsley Design Ltd., 2022 24
- 25. Claims in theory: easy • Policyholder submits claim • Claims manager decides whether/how much to pay • Policyholder received payment Copyright Kemsley Design Ltd., 2022 25
- 26. Before automation… Capture Decide Pay Claim Supporting Documentation Auto-adjudicate/Auto-pay 26 Copyright Kemsley Design Ltd., 2022
- 27. Claims in practice: complex • Regulations • SLAs • Manual procedures • Exceptions • Human interpretation and decisions • Knowledge work Copyright Kemsley Design Ltd., 2022 27
- 28. Claims automation: high level • Goal: automate repetitive work to free up knowledge workers • Key technologies • Process drives overall flow and interacts with workers • Content captures and manages claims file • Decisions enforce rules and regulatory compliance • Machine learning determines “fuzzy” decisions, e.g. auto-adjudication Copyright Kemsley Design Ltd., 2022 28
- 29. Intelligent claims Claim Supporting Documentation Intelligent Capture Auto-adjudicate/Auto-pay Exception Handling 29 Copyright Kemsley Design Ltd., 2022
- 30. What is possible now? • Existing business • Automate repetitive tasks – more efficient • Decision guardrails – less risk - faster training • Fast, fair claims – more competitive • New business models • Adjacent low-value markets • Micro-insurance • Parametric insurance • Usage-based pricing with telematics Copyright Kemsley Design Ltd., 2022 30
- 31. Summing it up • Understand the automation imperative • A lot of companies are failing at business automation, but you don’t have to be one of them • Align automation metrics with corporate goals • Design the right thing the right way • Get something into production fast, then prepare to pivot • Watch for #fail indicators Copyright Kemsley Design Ltd., 2022 31
- 32. Questions? Sandy Kemsley sandy@kemsleydesign.com Read my blog at column2.com Find me on Twitter @skemsley Copyright Kemsley Design Ltd., 2022 32
Editor's Notes
- #2: The success of business automation projects is measured by the degree of business improvement or transformation that is achieved, relative to the business goals. As simple as this seems, why do so many automation projects fail? Join Sandy Kemsley as she delves into some of the reasons for automation project failure, and what steps you can take to make your project a success. As a technology and business architect who helps her clients create large-scale automation solutions, as well as an industry analyst researching the application of technology to improve business, she offers a unique perspective on the factors leading to success or failure of these projects. You’ll learn about: What business leaders need to know about key automation technologies and their application to business operations Methodologies that can align the outcome of automation projects with your operational business goals Indicators that a business automation project is off course – and how you can steer it back in the right direction What’s possible with business transformation through automation, including a use case for how automation can drive maximum business benefit
- #5: Pandemic plus technology advances have ramped up automation initiatives In the old days, it was (just) a waste of money Failure to successfully automate is a survival issue
- #12: Michael Lewis “Against The Rules” podcast about Athena Health
- #28: This is why it’s complex: Claims manager needs to figure out what is missing, and request it Each iteration, CM reviews file, figures out if they can make a decision, and makes it Too much manual decisioning and ad hoc processes Expertise is buried in the knowledge workers
- #29: Goal: knowledge workers should be focused on handling decisions and customer interactions that are better managed by people – true for most business process automation Process to drive overall flow and interact with workers Decisions to determine auto versus manual tasks (e.g., auto-adjudication) ML to learn parameters of auto-adjudication, know when to ask for more information – learn from knowledge workers – explainable decisions based on past decisions and regulatory guidelines
- #31: Usage-based pricing: Allianz BonusDrive for UBI using telematics, more competitive for different market segments Algorithmic: New Paradigm’s parametric insurance pays out automatically based on parameters, e.g., weather instead of handling individual claims