The problem is not distance, it is escalation design

Plants under foreign ownership rarely fail because headquarters is far away. They fail because no routine exists that moves a problem from the shop floor to the person who can act on it, before it becomes a customer issue.

Written for operations leaders running or evaluating manufacturing in Brazil.

A pattern we see repeatedly: monthly reports from the Brazilian plant look plausible for months. Output is roughly on target and variances get explained after the fact. Then a customer complaint or a stockout reveals that a line has been running well below its capability for weeks.

Nobody hid it. It simply never travelled. There was no mechanism that required it to travel, so it stopped at the first person who could absorb it.

Why information stops

Verbal culture, and it is not a flaw

Brazilian shop floors communicate well informally. Problems are discussed, and they are usually discussed with the person immediately above. What is frequently missing is a written path that requires escalation when the problem exceeds a defined threshold. Without that threshold, absorbing the problem locally is the natural and even admirable response.

Reporting built for the reader, not the operation

Many plants produce a monthly pack assembled specifically for headquarters, disconnected from anything used to run the plant day to day. It gets rebuilt each month, definitions drift, and it reports what the previous pack reported rather than what currently matters.

Time zone reality

Brazil sits at UTC−3. Overlap with European headquarters is a few hours in the European afternoon; with Asian headquarters it is close to nothing. If escalation depends on a live conversation, it is structurally delayed by at least a day, every time.

What to build

1. Tiered meetings with thresholds

Shift, area, and plant tiers, each with a fixed short agenda and an explicit rule for what escalates. The rule must be quantitative — a threshold, not a judgement. “Escalate if the line is below X for more than Y minutes” travels. “Escalate if it is serious” does not.

2. A weekly written pack, same format every week

Weekly beats monthly for a plant under remote ownership, and written beats a call. The same indicators, the same definitions, the same order, with commentary on variance. Comparability across weeks is worth more than presentation quality.

3. One scheduled call in your time zone

A standing call sized to the overlap window, with a fixed agenda driven by the written pack. Because the pack arrived first, the call is about decisions rather than about status.

4. Indicators with named owners and stated sources

Every number needs a formula, a source and a person. If a number cannot be traced to how it was collected, it will eventually be wrong and nobody will notice.

What to avoid

• A dashboard as the intervention. Real-time visibility into a plant with no daily routine gives headquarters faster access to numbers nobody is acting on.
• Reporting in English only. Management reporting in English is fine. Shop-floor standards and boards must be in Portuguese or they will not be used.
• Escalation that punishes the escalator. If raising a problem reliably produces an uncomfortable call, escalation stops within a month. This single dynamic destroys more reporting systems than any technical flaw.

How you know it is working

A simple test: when something goes wrong at the plant, does headquarters learn about it from the plant, or from a customer? If the answer is consistently the former, the system works regardless of how sophisticated it looks.