Lessons From The Automation Trenches

EKATERINA CURRY, GLOBAL HEAD OF RATINGS OPERATIONS, S&P GLOBAL RATINGS

Watching many companies that tried to automate over the last few years, I often wondered why only a few succeeded, while some plateaued and many failed. In 2019, my team at S&P Global Ratings began such a journey - experimenting with automation to improve our workflow. Although we faced some obstacles along the way, we developed a framework to improve our approach to automation and better allow us to identify and respond to challenges.

I lead a global operations team with colleagues in 22 countries that strives to continuously improve what we do and how we do it. My team deeply understands the products, processes, and tools we use to serve our customers. We are fluent in operational excellence, efficiency, process design, and value stream mapping. However, requirements gathering, development work, user acceptance testing, user experience, and low-code robotic process automation (RPA) were foreign concepts to us until 2019. After all, operations teams are not technologists. Or so we thought.

We began by developing automation skills within the team. Operations specialists volunteered and became certified as “citizen developers,” non-technologists who can build applications with minimal coding in a safe, low-risk “sandbox”. We started exploring ways to automate, reduce waste, create capacity, and have fun. We developed and released more than 20bots to production over a few months - the “low-hanging fruit” or “quick wins.” We were delighted to win a prestigious industry award “Best digital transformation project under 90 days” from OPEX. As we progressed, we got swept up in the excitement of early adoption. Pleased with the progress, we thought our work was done. Then we started experiencing all kinds of challenges.

Every time there was a new release, the bots would break. We didn’t have the skills or the time needed to constantly fix and update the bots. We hadn’t sufficiently worked out the system with our technology colleagues who supported the bots in production, and massively underestimated the maintenance required to successfully operate the bots. We hadn’t incorporated best practices in bot development. We hadn’t prioritized the highest value opportunities and those that directly linked to our broader business strategy. We asked ourselves:

1. How do we balance the enthusiasm of learning RPA with return on investment (ROI)and business value?

2. How do we create capacity on our teams so operations specialists have time to learn these skills?

3. Who will support and maintain the bots?

4. How does each bot contribute to our business plan?

5. How do we measure success with each bot?

“Help your team support each other by creating a community of learners through company-operated RPA academies, individual learning and certifications”

We pondered what it would take for our team to succeed at automation and mature from an opportunistic “quick-wins” approach to a more strategic view. While there is no one recipe for success, we came up with a few building blocks to increase the likelihood of successful automation:

1. First simplify, then automate – Use Lean Six Sigma tools to streamline the process before automating.

2. Create a community of practice - Help your team support each other by creating a community of learners through company-operated RPA academies, individual learning and certifications.

3. Set aside time - Plan for and set aside team capacity to experiment with automation. Learn and fail forward.

4. Coordinate with technology – Get clear agreement from technology teams that they will support the bots deployed by non-technologists.

5. Get leadership support - Make non-technologist automation initiatives a leadership priority.

6. Decide how to prioritize - Clarify the framework for business value creation, ROI, and prioritization for the bots.

7. Link all automation work to the business plan - Ensure all automation projects have a direct link to the corporate strategy and team business plan.

We are now in our fourth year of our automation journey. Would we do it again? Absolutely. We have found so much value in learning by doing. But moving forward, we’ll approach automation with this new framework, giving us more discipline and clarity of outcome. We won’t rush to certify as many citizen developers and deploy as many bots as possible. We won’t get swept up in the excitement of experimenting and learning.

“Ensure all automation projects have a direct link to the corporate strategy and team business plan”

Instead, we’ll focus on the tried-and-true principles we’ve now established–simplify the process first, then automate; create time for the team to upskill and experiment; secure leadership and technology support; and be ruthless in prioritization, only picking projects with clear links to our strategy and business plan. We still have a long way to go in our automation journey, but with this new framework, we’re up for the challenge.