Routine actions are made to require explicit approval after an incident or audit finding, which concentrates discretion with the approvers and puts the queue in their hands.
Borderline: whether this move is justified depends on the situation, and only the person in it can judge.
An example
After a grocery chain complained publicly about a botched promotion, the division head at a frozen foods maker pulled a year of customer credits and found account managers granting them at very different rates for similar problems. She issued a memo: every credit, of any size, would now need approval from her commercial finance team, with a five-day turnaround target. Account managers had previously issued credits within limits, usually the same day. She added two analysts to the team so the queue would move. Two account managers who issued credits on their own that month were handled through the normal disciplinary process. By the end of the quarter credits had fallen by a third, commercial finance was deciding which customers got them quickly, and account managers were routing routine questions through it too. The memo set a review date for the following spring.
Why it works
Asking permission becomes the easy choice, because routing a decision upward moves the accountability to the approver: if it goes wrong, the form protects you. Acting on your own looks worse, since initiative that bypasses the chain is recast as cowboy behavior or a control weakness, whatever the result. And it gets costlier, because an unapproved action is now a documented breach that can surface in reviews, audits, and the next layoff list. The new gate needs neutral language about risk or audit, approvers with cover from above, and a few early cases made very visible.
The thinking behind it
Richard H. Thaler Thaler and Sunstein's Nudge: Improving Decisions About Health, Wealth, and Happiness (2008) shows that defaults are the strongest element of choice design, since what happens when people do nothing is what most people get. Moving credits from permitted to pending flips the default for every account manager at once. No argument is needed afterward: people route requests to the approvers because that is now the easiest path.
Richard Zeckhauser Samuelson and Zeckhauser's Status Quo Bias in Decision Making (1988) found that people choose the option framed as the current state disproportionately often, a finding a 2021 replication largely supported. Once an approval gate exists, it becomes the status quo, and removing it means arguing for change. That is why a gate installed after one incident tends to outlast the incident, and why a built-in review date matters.
Oliver E. Williamson Williamson's Markets and Hierarchies: Analysis and Antitrust Implications (1975) assumes that some people act opportunistically and that organizations build safeguards against it. Credits granted inconsistently for similar problems are evidence of that hazard, and they give an approval gate its legitimacy. The same logic limits the move: a safeguard that costs more than the opportunism it prevents will be challenged with numbers.
When it fits
Process owners with authority over how work is approved, after a real incident or audit finding has shown that scattered discretion is costing money or creating risk. It needs approvers with the capacity to handle the queue and support from above for the new gate.
How to use it well
Tie the gate to the incident and say plainly what it is for.
- Publish the new rule, who approves, and the target turnaround.
- Staff the approval team so the queue moves at the speed you promised.
- Enforce consistently from the first breach, through the normal process.
- Report what the gate has saved or caught, in money and in incidents avoided.
Set a review date, and publish the queue's cycle times each month. A gate you are prepared to defend with numbers is easier to keep than one you merely inherited.
What it costs and where it backfires
Queues cost customers and staff time, and if the gate slows the business more than it saves, the people it hurts will push to remove it, with numbers. Approvers who hold discretion become the target of every complaint. Purpose and proportion keep the gate defensible. Once the gate exists only to give you a hold over people, its cycle times are hidden, or approvals are used to reward allies and punish critics, it has become retaliation by queue, and the first executive to see the pattern will take the discretion away.
When it is used on you
Name the new chokepoints publicly as choices someone made, track what each queue costs the business, and ask for a threshold below which the gate does not apply. Follow the rule while you argue against it, and take the numbers to the executive who owns the result the queue is hurting.