The Same “Point of No Return” Risk Lurking in Opposite Decisions
This week, two seemingly unrelated management decisions were reported. One was by U.S. payments giant Block (formerly Square), which, despite strong performance, announced a bold organizational transformation: cutting approximately 4,000 employees—about half its workforce—citing the use of AI. The other was a historic chicken restaurant in Kobe deciding on a price hike deemed “unavoidable” due to rising material costs, while continuing to invest in creating a workplace where people can work long-term.
One is a forward-leaning efficiency drive using cutting-edge technology; the other is a defensive decision to protect tradition. When viewed through the lens of “reversible management,” these two seemingly opposite choices reveal they share the same structural danger. It is this: executing a decision without clearly defining an “evaluation period” and “observation points” makes it extremely difficult to turn back.
Block’s “AI Transformation”: The Trap of “Singular Focus” in the Name of Efficiency
Block CEO Jack Dorsey explained this large-scale reduction as an “organizational transformation through AI.” Certainly, it’s a plausible scenario that AI introduction automates tasks and reduces required personnel. However, the question here is whether that decision is “reversible.”
Suppose, one year after AI implementation, situations arise such as “the expected efficiency wasn’t achieved,” “customer satisfaction declined,” or “there aren’t enough human resources to respond to new business opportunities.” Could the company easily call back 4,000 employees then? It would likely be nearly impossible. Dismantling an organization once, especially losing highly specialized talent, is a decision that is almost irreversible.
From the perspective of “reversible management,” this decision has a critical design flaw. It is that it bets everything on the single hypothesis of “efficiency through AI” and leaves almost no recovery means (backup plan) for when that hypothesis fails. If the decision is viewed as an “experiment,” they should have designed in advance an evaluation period, exit criteria, and buffers for failure, such as: “If customer response time is not reduced by XX% within 6 months of AI introduction, freeze XX% of the reduction plan,” or “Implement cross-training for remaining members to cover the roles of key personnel who left, concurrently with the reductions.”
The “Price Hike” at the Kobe Historic Shop: The Quality to Protect and the Customers Who Might Be Lost
On the other hand, what about the decision by the historic Kobe shop? Based on the clear value of “protecting quality,” they chose to raise prices. Simultaneously, they continue investing in creating a long-term workplace. This seems like a sound decision that clarifies the core to protect (quality and employees).
However, the risk of reaching a “point of no return” also lurks here. It is almost certain that some customers will leave due to the price increase. The problem is whether there is a mechanism to observe in advance whether that attrition is “temporary and within acceptable limits” or “unstoppable and threatens the shop’s survival.”
“For three months after the price hike, measure the number of customers and average spending per visit weekly. If customer count decreases by more than XX% and the increased revenue from higher prices cannot compensate, implement partial corrections such as reviewing set menus or resetting lunch price ranges.” By deciding in advance on such specific evaluation periods (3 months), observation points (customer count, average spend), and corrective actions (partial menu changes), the decision to raise prices changes from an “all-or-nothing gamble” to an “adjustable experiment.” While immediately reversing the price hike itself might be difficult, leaving room for adjustment through accompanying services or menu composition is a “reversible” design.
The Common “Decision Solidification” Process in Both Cases
This is not to say that Block’s staff reduction or the historic shop’s price hike are wrong decisions in themselves. The problem likely lies in solidifying those decisions as “final decrees” and not incorporating a loop for flexible observation and adjustment even when circumstances change.
One fundamental principle of “reversible management” is to “prioritize observation over solidification.” Whether reducing staff or changing prices, before making it a permanent “decision,” first position it as a “time-limited experiment” and decide on the metrics and frequency for observing its results. Then, secure an “exit” in advance to revert or adjust in another direction if the experiment’s results differ from expectations.
In Block’s case, did the means of “AI introduction” become an end in itself, weakening the perspective for multi-faceted observation of the results it brings (true productivity improvement and sustainable growth)? In the historic shop’s case, while the cause of “protecting quality” is clear, was the price hike chosen after sufficiently considering alternatives (e.g., reviewing menu portion sizes or provided ancillary services)? The question is whether “reversibility” in the form of alternatives or a phased approach was incorporated into that deliberation process itself.
Designing a “Reversible Experiment” You Can Start in Your Organization Today
So, how should you, as an SME leader, design “reversible decisions”? Consider the following three steps.
Step 1: Verbalize the Decision as a “Hypothesis”
“Introducing AI will reduce administrative tasks by XX%.” “Even with a 10% price increase, customers will understand the quality and continue to support us.” First, write down as concretely as possible the “this should happen” hypothesis underlying your decision. If this hypothesis is vague, subsequent observation becomes meaningless.
Step 2: Decide on the Evaluation Period and Observation Points
Set in advance the “period” and “metrics” for judging whether that hypothesis is correct. Avoid vague period settings like “until results are achieved” at all costs. Set clear dates like “after 3 months” or “after 6 months.” For observation points, choose metrics that allow you to view processes and impacts multi-dimensionally, not just outcome metrics like sales, such as “employee overtime hours,” “number and content of customer complaints,” or “new customer acquisition cost.”
Step 3: Define “Failure” and Decide on the Action to Take Then
This is the most crucial and most frequently omitted step. “At the end of the evaluation period, if observation point A is below XX and B is above YY, define this decision as ‘not achieving the expected results.'” Then, decide in advance—at a time when you can decide without being swayed by emotion—”whether to completely scrap the decision (e.g., withdraw the price hike), make partial corrections (e.g., continue the hike only for some menu items), or extend the observation period by another 3 months.”
Conclusion: The Quality of a Decision is Determined by the Clarity of its Exit Conditions
Whether it’s Block’s bold transformation or the historic shop’s decision to protect tradition, the essential risk differs greatly depending on how much thought is given to the “design of reversibility.” In management, the most irreversible decisions are those that entrust everything to a single hypothesis of “this should work” and do not consider what happens if it fails.
An excellent leader is not someone with strong decisiveness, but someone who can simultaneously consider and design “how to correct course with minimal damage if this decision is wrong.” When facing your next management decision, please consider “how to proceed back” just as much, or even more, than “how to proceed forward.” That small margin of thought becomes the most powerful safety net protecting your organization from irreversible failure.
Comments