Organizations are being told that AI will transform how work gets done.
So naturally, many start by looking for tools. There is no shortage of them. New platforms, features, and capabilities are promoted constantly.
What often gets overlooked is that successful AI adoption rarely starts with technology — it starts with operational problems.
While the discussion below focuses on AI, the underlying challenge is automation, and automation has always depended on clearly defined work.
Most work does not live inside a single clean workflow that technology alone can fix. It moves across systems, emails, approvals, and unwritten steps people rely on every day.
So, before deciding what AI to adopt, organizations should ask a simpler question:
Is our work structured enough for something else to perform it?
Automation can execute a process, but it cannot define one.
In many environments, especially the public sector, processes exist across a mix of systems, policies, and legacy human knowledge. The workflow exists, but parts of it live in different places.
Before automation is possible, organizations must understand what is actually happening today, not just what documentation or policy suggests should happen.
Automation works best when work is consistent. When the underlying process is inconsistent or misunderstood, automation will propagate that inconsistency.
Instead of starting with the tool, the sequence works better in reverse:
Consider a funding or permit application intake.
Two reviewers may request different supporting documents because parts of the process live in policy, legacy forms, and staff experience rather than one defined workflow.
If automation is introduced at this stage, the system does not resolve the difference — it learns one interpretation and applies it repeatedly.
Before automating application or permit review:
Again, before choosing an AI tool, organizations should ask:
Do we understand our work well enough for something else to perform it consistently?
Before AI can transform the work, the organization has to understand the work.
If something in this post resonated—or sparked ideas—we’d love to hear from you.
If something in this post resonated—or sparked ideas—we’d love to hear from you.
Join our mailing list to receive fresh insights, stories and updates.
