Every approval in an organisation was added for a reason: a past mistake, a control failure, a compliance requirement, a risk event, or simply a manager’s desire to stay informed. Over time, approvals accumulate. New ones are added. Old ones are never removed. The result is an architecture that reflects the accumulated anxiety of the organisation rather than a rational assessment of which decisions require oversight.
The consequence is measurable. A purchase order that requires four signatures consumes four people’s time and adds four scheduling delays to the decision. If the total elapsed time is five days and the early-payment discount expires in three, the approval process has cost the company money while ostensibly protecting it from risk.
The Mapping Methodology
Map every significant approval in the organisation: decision type → approver → information the approver receives → threshold → time consumed → risk the approval is designed to reduce.
Then evaluate each approval against two criteria:
Risk reduction: Does this approval materially reduce the probability or impact of a bad outcome? An approval on a Rs 50 lakh capital expenditure by a person who understands the technology, the vendor and the business case may genuinely reduce risk. An approval on the same expenditure by a person who signs based on the previous approver’s signature adds a name but not a judgment.
Time cost: How much elapsed time does this approval add to the decision, and what is the economic value of that delay? A procurement approval that adds three days to a time-sensitive purchase has a quantifiable cost: lost discount, delayed production, missed customer commitment.
When the time cost exceeds the risk reduction, the approval is destroying more value than it protects.
The Redesign Principles
Proportionality: Approval depth should be proportional to the irreversibility and magnitude of the decision, not to the category. A Rs 5,000 petty cash disbursement and a Rs 5 crore capacity expansion should not require the same governance architecture.
Pre-approved parameters: For recurring decisions within defined parameters (procurement from approved vendors below a threshold, pricing within a defined range, credit within policy limits), pre-approval eliminates latency without eliminating control. The parameters are the control. The individual transaction does not need separate approval.
Parallel processing: Approvals that can be reviewed simultaneously should be, rather than processed sequentially. Three approvers reviewing in parallel consume one day. The same three approvers in sequence consume three days.
Exception-based escalation: Instead of requiring approval for every transaction, approve by default within defined parameters and escalate only exceptions. The control shifts from universal pre-approval (slow, high volume, low attention per item) to exception-based review (fast for standard items, concentrated attention for unusual ones).
In Northrop Management governance advisory work, approval matrix redesign is one of the most impactful interventions available: reducing decision latency by 30% to 50% while maintaining or improving control effectiveness by concentrating governance attention where it matters.
Ashish Chaudhary, frames the governance principle directly: “An approval is not a control. It is a potential control. It becomes an actual control only if the approver has the information, the incentive and the time to evaluate the decision meaningfully. An approval that is rubber-stamped because the approver is too busy to review it is not a control. It is a delay with a signature.”
Questions for the Boardroom
- How many approval layers does our most frequent high-value decision require, and is each layer adding genuine risk reduction.
- Which approvals could be replaced by pre-approved parameters without increasing risk?
- What is the total elapsed time for our top five approval processes, and what would that time be if we processed approvals in parallel rather than in sequence?
- Have we added any approval layers in the last three years, and have we removed any?
- If we eliminated every approval that adds time without adding judgment, how much decision latency would we recover?
Closing Implication
The approval matrix should be designed, not accumulated. Every approval should earn its place by demonstrably reducing risk proportionate to the time it consumes. Approvals that add latency without adding judgment should be eliminated, because their only effect is making the company slower without making it safer.
