Your first automation project should solve a problem your team can describe clearly. A modest workflow with a measurable result is easier to assess than a plan to connect every system at once. Look for repeated work, stable inputs and rules that people already understand.
Choose your first automation project from observed work
Ask staff to record recurring tasks for a short period. Note what starts each task, the information used, the time spent and the corrections required. Include work that happens only occasionally if it has a clear operational impact.
For an illustrative design studio, transferring approved project details into a task board might be a candidate. Deciding whether a complex client request is commercially suitable may need more human judgement and deserve a different approach.
Assess candidates against practical questions
- Does the task happen often enough to justify the work?
- Are the inputs and rules reasonably stable?
- Can the team agree what a correct result looks like?
- Is there a safe way to handle exceptions?
- Can someone own the workflow after launch?
Consider the cost of building, testing and maintaining the connection alongside the effort it may save. Use observed task timings as a baseline. Avoid presenting an estimated saving as a guaranteed result.
Simplify the process before connecting it
Ask whether a redundant step can be removed or a form made clearer. Automating a confusing process can make its errors travel faster. Decide where the authoritative record belongs and who approves important changes.
Keep the first scope small: one form, one service or one team. Write down what is excluded so that new ideas do not quietly expand the pilot while it is being built.
Keep judgement where it belongs
Some workflows benefit from preparing information for review instead of making the final decision. Define the point at which a person checks the result, and ensure the system waits for the appropriate approval.
If an AI component is proposed, establish why it is needed and how its output will be checked. A rule-based step may be more suitable where the decision is already precise. Test the chosen approach against realistic examples.
Review the pilot honestly
Compare the time spent on the old task with monitoring, corrections and support in the new process. Ask users whether the result is easier to trust and manage. Record failures and changed circumstances that affect the comparison.
Finish with a decision to expand, revise or stop. Keep the operating instructions and ownership clear whichever option you choose.
Explore website automation options and bring your process to Kilwhiss Automate. A well-chosen first project provides evidence for what to do next.