Management teams often name their largest processes first: annual accounts, due diligence or quarterly reporting. Those may be valuable opportunities, but small, frequent handovers deserve a place on the shortlist too. Choose by value and readiness, not visibility alone.
A first project should be important enough to matter and bounded enough to test. Look at frequency, time, review effort and the consequences of failure. The largest process is not always wrong; it is simply not the automatic winner.
Why does the damage sit in the interruption?
Consider an illustrative task taking ten minutes and arriving eight times a day. That is eighty minutes of direct work. Interruptions can add a recovery cost, but measure it rather than adding an arbitrary multiplier.
A person returning to a review may need to reread the previous section or recover their place. A short observation exercise can show whether this is a meaningful part of the workload.
Why are these tasks invisible?
Small tasks can be recorded, but often are not tracked separately. Ask the team to log a representative week, including exceptions and rework, before estimating the opportunity.
For example, suppose forty documents a month require a disclosure review. At fifteen minutes each, that is ten hours. If assisted preparation cuts review to eight minutes, the direct saving is about four hours and forty minutes before operating costs. These are planning assumptions, not a measured client result.
Why is uneven effort worse than high effort?
Variable effort makes staffing harder. A simple document may take ten minutes in a planning example while a large bundle takes much longer. Group cases by type and complexity instead of estimating every file from one average.
Record a range and identify why cases differ: unreadable scans, missing references, approval delays or unfamiliar formats. The variation tells you where a system needs an exception path.
What does a non-standard tool cost later?
Inconsistent tools can make it harder to tell which standard was applied later. Version records and a clear release checklist help make the result traceable.
That is a downstream cost worth including. A preparation process that saves minutes but creates uncertainty in the archive may not improve the overall workflow.
What does it mean when the work lands with management?
It may indicate missing ownership or cover. Ask why the work is escalated and whether the decision genuinely needs management. Sometimes a clearer rule or a backup owner solves the problem before any automation is needed.
Which question finds the right first process?
“What keeps interrupting your day?” is a good discovery question. Use the answers to build a shortlist, then compare business impact, readiness and risk. Irritation alone is not a business case.
A second question does the rest: what happens when the person who normally handles this is away. Wherever nobody has a clear answer, there is a process without cover.
Why not start with the largest process?
A large process can involve many dependencies and expensive mistakes. It may still be the right target, but isolate a useful sub-process that can be tested without committing to a complete rebuild.
Choose one owner, a baseline and a measurable acceptance test. Include setup, review and maintenance in the economics. The first project should give you evidence for the next decision, not just an impressive demonstration.
