Written by: David Carneal – Digital Efficiency Consulting Group – DECG
Read Time: 3 min
Approval steps are not automatically bad. Some protect quality, margin, risk, compliance, or customer commitments. Others exist because someone, years ago, got nervous and handed the workflow a rubber stamp. The stamp survived. The reason did not.
A workflow diagnostic does not ask whether approvals feel familiar. It asks whether they change the outcome. That question makes approval bloat very uncomfortable, which is usually a sign it needed the conversation.
The approval value test
For each approval in a workflow, ask what decision is being made. If the approver rarely rejects, changes, redirects, or improves the work, the step may be ceremonial. Ceremonies have their place. Your order cycle probably is not one of them.
- Score each approval:
- 1 point if it prevents measurable risk.
- 1 point if it improves quality or customer outcome.
- 1 point if it controls cost, margin, or cash exposure.
- 1 point if it satisfies a documented compliance requirement.
- 0 points if it exists because "leadership wants visibility."
Watch for approval drift
Approval drift happens when controls accumulate but never get removed. A process designed for exceptions becomes normal operating procedure. A manager added for oversight becomes a permanent checkpoint. A one-time failure creates a permanent delay for every future transaction.
- Routine work requires executive review.
- Approvals duplicate checks already performed earlier.
- The approver lacks context and rubber-stamps anyway.
- Urgent work bypasses the process, proving the process is not trusted.
A safer way to reduce approvals
Do not simply delete approval steps like a caffeinated minimalist. Segment them. Keep approvals for high-risk, high-dollar, unusual, or exception-based work. Remove or automate low-risk routine approvals where clear rules already exist.
- Define risk thresholds.
- Clarify who has authority.
- Replace approval with notification where visibility is enough.
- Review exceptions monthly instead of stopping every transaction.
Approvals should protect value. When they only protect habit, they become delay with a title.
CTA: Review one approval chain and identify which approvals actually change decisions. Start there, not with the sacred button parade.
Replace blanket control with smart control
The better answer is usually not "no approvals." It is better approval design. Routine work should move through clear rules. Exceptions should get attention. High-risk work should be reviewed by someone with authority and context. That is governance. Everything else may be workflow furniture.
Approval redesign also protects managers from becoming human speed bumps. If every routine transaction needs a manager, the manager becomes the process bottleneck and the team learns to wait instead of own the work. That is not control. That is dependency with a calendar problem.
- Use thresholds for dollar amount, risk, customer impact, or exception type.
- Give routine approvals to the process, not the manager.
- Notify leaders when visibility is needed but permission is not.
- Audit rejected or changed approvals to prove the step earns its place.
A quick manager check
Before changing anything, ask the people closest to the work to walk through one recent example from start to finish. Do not ask for the policy version. Ask for the Tuesday-at-3:40 version, where the missing field, extra email, surprise approval, and system workaround all show up wearing tiny disguises. That is the version leaders need before they can make useful improvements.
- What slowed this example down?
- Who had to chase information?
- Which step would be hard to explain to a new employee?
- What small change would prevent the same delay next time?