The questions that come up in an automation strategy discussion set the tone for everything that follows. The best questions are those that have little to do with technology. They have everything to do with outcomes, risk, people, and focus.
1. Which Business Problems Are We Trying to Solve, and in What Order?
Start With Measurable Pain Points
Instead of asking where artificial intelligence can be applied, it is best to focus on three to five measurable business problems. These might be, for instance, long onboarding processes, high error rates, or slow collections and rising churn in an important segment.
Challenge your team to present clear metrics and baselines.** If it cannot be quantified, it is likely too vague to be automated.
2. How Will This Change the Day to Day Work of Our People?
Focus on Real Work, Not System Design
The value of automation is realized by changing how people work every day. How will it reduce tasks, increase tasks, and what new skills will be needed?
Challenge your team to provide examples.** A good approach is to ask how an employee’s day looks today and how it will look after the project is completed.
3. Which Journeys Are We Automating End to End?
Decide What Should Stay Human
Pockets of automation projects can be confusing, whereas end-to-end journeys have tremendous benefits. Instead, ask what customer or internal journey will be fully automated end to end.
The journey might be from contract to first invoice, from support issue to resolution, etc. At the same time, what should we deliberately choose to be non-automated, which needs judgment and relationship building?
4. Who Owns the Process After Go Live?
Assign a Clear Business Owner
Many automation projects have sponsors, but they do not have owners once the project is over and the system is live.
We should ask who owns the process on the business side. This is someone who owns it post-launch, not just someone who owns it pre-launch.
5. How Will We Measure Success Over Time?
Define Results at 90, 180 and 365 Days
Efficiency or optimization are not useful measures. Ask the group to come up with three to five clear measures that will determine if the initiative is successful.
Common measures include cycle time, error rates, customer satisfaction, cost per case, or revenue leakage. Establish goals for the first 90 days, 180 days, and one year. Tie further investments to the outcome of the initiative.
6. What Risks Are We Taking and How Are We Managing Them?
Consider Operational, People and Compliance Risks
Automation has the potential to generate substantial value, but it could also amplify risk if done incorrectly.
Ask the group to consider the areas where failure would have the greatest impact. This could include incorrect decisions being made incorrectly, exposure of data, customer satisfaction, or even the morale of the employees.
Make sure the group has thought about the “guardrails” in place, including approval mechanisms, audit trails, and explainability.
7. What Is the Smallest Version That Can Prove Value?
Start With a Focused Release
Many large-scale, all-at-once automation programs fail to deliver because they try to address too many issues simultaneously.
The team should discuss and decide on the smallest version of automation that will show value. This may require focusing on one particular journey, region, or product line.
The first release should be able to go live in weeks, not years, with a plan for learning and iterating afterward.
Turning Questions into Better Automation Decisions
CEOs who lead discussions on automation with these questions create a shift in thinking, moving the focus away from tools and toward results and accountability.
No more grand roadmaps with uncertain results, but rather smaller, more tangible programs with proven results and the ability to scale with confidence over time.