The Manual Work Ledger
We treat a function’s manual work the way a balance sheet treats assets: inventoried, valued, and reconciled against what actually happened.
- 01
Inventory the work, not the process
Manual Load lives between the steps of the process map: re-keying, reconciling, chasing, checking, reformatting, routing, explaining. We build the inventory at task level from work sampling, system logs and structured interviews with the people doing the work.
- 02
Price it
Every task line gets fully loaded hours, cost, and a Manual Load percentage for the team and the function. It is usually the first time that number has existed in the organization.
- 03
Sort by realizability, not automatability
Every opportunity sits on two axes: can it be eliminated, and can the value be converted. Value converts through one of four paths: growth absorbed without hiring, a backfill avoided as people leave, redeployment to work with a measurable output, or a cycle-time gain that changes cash. No path and no owner scores zero, however automatable the work.
- 04
Eliminate in the operator's order: stop, simplify, automate
First ask whether the work should exist at all. Then simplify: a rule change or a field added upstream removes more manual work than a bot. Only then automate, with AI where the work is unstructured or judgment-shaped and plain integration where it is not. Automation is the expensive option and is spent on the work that genuinely needs it.
- 05
Realize through a Value Contract
Each elimination ships with a one-page Value Contract: the opportunity, its conversion path, a named owner, a date, the cost line it will move, and a finance sign-off. It closes the Realization Gap, and clients keep using it after we leave.
- 06
Compound
The Ledger becomes a standing management instrument: a quarterly Manual Load review owned by the function head, with your own people trained to run the inventory.
Automatable is not the same as realizable. The most automatable item on a list is often the least likely to change a cost line. The boring item with a willing owner and a clear conversion path is the one that lands.