About: Cross Cultural Archives

When the numbers do not add up: investigating a Czech plant without triggering chaos

In brief Financial registers, scrap logs and inventory valuations in a Czech plant stop reconciling. German boards then face a real governance choice. An informal call to the plant manager or finance controller gives a compromised manager time to adjust the records. Formal scrutiny loses its element of surprise. Waiting also compounds fraud exposure and statutory director liability under Czech law. The approach that works is different: a discreet, dual-track investigation. An interim executive with genuine operational authority secures the facts on site within days. Production, customer deliveries and supplier payments continue without interruption. Audit triggers: why German executive boards hesitate to launch investigations A whistleblower alert or an anonymous tip can surface at any time. So can an inventory variance that will not reconcile, from a Czech plant in Plzeň, Liberec or Brno. Executive committees in Munich, Stuttgart or Frankfurt then face a genuine dilemma. A formal forensic investigation run visibly from headquarters carries real risks. It can not only destabilise customer deliveries but can also alienate a trusted local managing director, the jednatel. It can also become public in a way that damages the parent company’s reputation. Under that pressure, an informal call to the local plant manager feels like the cautious first step. It is understandable: nobody wants to escalate a discrepancy that might turn out to be a clerical error. The problem is different: even a well-intentioned call gives a compromised manager time to act. The manager can adjust production logs, correct stock counts, or delete electronic communications before investigators arrive. Unreconciled inventory, unexplained scrap, and unapproved scrap sales are rarely accidental. They usually mask a production yield problem, an unauthorised commercial arrangement, or margin diversion. McKinsey & Company’s research on data-quality investigations in manufacturing makes the same point. Boards need to isolate and resolve operational discrepancies through structured root-cause protocols, not a phone call. Each week an anomaly goes uninvestigated, financial exposure compounds and the evidentiary trail degrades. Cross-border manufacturing governance: German-Czech supply chain risks Manufacturing networks between Germany and Czechia operate on a highly integrated, often just-in-time basis. Organisations such as the German-Czech Chamber of Industry and Commerce (DTIHK) support that integration. A single plant disruption can affect German assembly lines within forty-eight hours. Investigating inside that network carries four distinct complications. 4 operational challenges in cross-border plant audits Local politics can reframe the investigation. An unannounced corporate audit team, arriving with visible legal scrutiny, changes how the plant reads the investigation. Local management can present it as headquarters acting against local workers, not as a specific financial question. That framing can trigger trade union resistance, work-to-rule behaviour, and the loss of hard-to-replace technical staff. The risk is real, but avoidable. The investigation needs an on-site posture that does not read as an attack from a distance. Manufacturing fraud is physical, not only digital. Falsified scrap logs can cover unauthorised overtime or off-the-books metal sales to local recyclers. Unrecorded work-in-progress can inflate a subsidiary’s balance sheet to hit bonus hurdles. Establishing what actually happened requires shopfloor knowledge, not a spreadsheet review alone. Statutory duties sit under Czech law, not German law. Under the Czech Act on Business Corporations (Act No. 90/2012 Coll.), a managing director, the jednatel, carries a statutory duty of care and loyalty. Czech law names this duty the péče řádného hospodáře. If the investigation confirms a statutory breach, the team must collect evidence carefully. It has to be admissible under Czech civil procedure from the outset, not retrofitted afterwards. Production cannot pause while investigators establish the facts. Customer orders still need fulfilling, raw materials still need receiving, and suppliers still need paying while the investigation proceeds. Accountants cannot simply review five years of invoices from Germany with the plant on hold. That is not realistic for a facility feeding OEM assembly lines. Balancing corporate governance with subsidiary operational realities None of this is a story about an unreliable local operation versus a vigilant headquarters. Local plant leadership usually works under its own pressures. Headquarters sets production targets centrally and margins stay thin. The plant often has no clear route to raise a concern before it becomes a visible discrepancy. Most supervisors and shopfloor staff have no part in a reporting scheme. Nobody should treat them as suspects by association. Both sides need the same thing: one verified set of facts, confirmed before anyone can alter them. Detecting financial anomalies: red flags in inventory and scrap reporting Three or more of these patterns, appearing together, signal deliberate distortion more strongly than any single anomaly alone. When several of these appear together, the situation has moved past a reporting query. It calls for on-site operational authority, not another round of emails. Interim management intervention: executing a dual-track forensic audit The sequence matters more than the individual steps. An interim executive needs genuine statutory authority from day one. That authority should not arrive gradually, once trust has grown. Securing plant evidence and establishing executive authority The first move is to place an interim Managing Director or interim CFO on site under a genuine operational mandate. The most credible mandate connects to a real business priority, such as a performance diagnostic or a planned capacity review. The executive actually leads plant performance and continuity from day one. Fact-finding then happens naturally from inside that authority. Announcing it as a separate exercise would only give a compromised manager time to alter the record. Within the first twenty-four to forty-eight hours, the priority is to secure the evidence. That means electronic records, ERP data, email servers and physical production logs, all without creating shopfloor alarm. It also means an unannounced physical inventory count of raw materials, work-in-progress and finished goods, checked against the general ledger. The count typically runs over a weekend, when it will not interrupt production. Reconciling physical inventory with ERP production data Reconciliation begins only once that evidence base is secure. The team checks machine runtime and energy consumption data against reported output. This shows whether equipment ran off-the-books batches, or whether someone

When headquarters must intervene: a board checklist for Czech manufacturing operations

In brief When a Czech manufacturing subsidiary keeps missing operational targets, the board’s natural response is to ask for more reporting. It usually means a revised turnaround plan, a weekly cash tracker, another review call. That instinct is understandable. It rarely closes the gap, because more detail from the same reporting line does not change what is happening on the shopfloor. Direct executive intervention becomes the right decision once specific, observable conditions are present, not once patience runs out. This sets out what those conditions are and which executive authority each pattern actually requires. It also covers how quickly that authority can be in place. The operational trigger: why increased reporting fails to resolve plant target misses The pattern arrives at a German board in a familiar form. A plant in Plzeň, Mladá Boleslav or Liberec has been reporting broadly acceptable output for several quarters. Yield keeps drifting and margin keeps contracting. Each quarter brings a revised turnaround plan from local management that has not closed the gap. Asking for more reporting is a reasonable first response. A board can pull several levers without stepping into the plant itself. A weekly cash tracker, a fresh recovery plan, another review call: all cost little to request. Under pressure, that instinct is sound. The difficulty is that reporting drawn from the same operation, at the same level of authority, rarely produces different facts. It produces the same picture in more detail. Timing matters more than it appears to at this stage. Operational turnaround research by McKinsey & Company on decisive executive intervention points to why timing matters. Turnaround situations respond to rapid, decisive executive action within the first thirty days. Every additional month spent reviewing plans while shopfloor scrap continues to rise uses up cash and customer confidence. It also narrows the range of options still open to the board. Cross-border management challenges: oversight limits between German HQ and Czech sites Four conditions make this specific corridor harder to manage than the geography suggests, and each has a reasonable origin. Compliant reporting can still hide operational drift. Czech plant teams typically maintain strong administrative discipline. The monthly pack reaching Stuttgart or Munich is usually complete and correctly formatted. That formal compliance is real, and it is not the same thing as operational visibility. Machines running below rated speed, unlogged micro-stoppages and rework cycles rarely appear in a top-level OEE or scrap figure. Nothing in the reporting template asks for them directly. Proximity does not substitute for shift-level knowledge. A facility in Ústí nad Labem or Plzeň sits only a few hours from Bavaria or Saxony. German executives reasonably read that distance as manageable oversight. A half-day site visit produces a clean tour and a useful conversation with the plant manager. It does not surface what changes between shifts, which is usually where the real variance sits. Legal and information barriers: Czech managing director liability and local knowledge risks Czech corporate law places personal liability on the local managing director. The jednatel carries statutory fiduciary responsibility for the entity, separate from the German parent’s own governance. Headquarters sometimes sets aggressive production targets without releasing the working capital or capital expenditure those targets assume. When that happens, the jednatel’s own legal exposure gives them a direct reason to protect their position. They have less reason to volunteer the full picture upward. That is a predictable response to how authority and liability are split across the border. It is not a sign of bad faith. Long-tenured local teams hold knowledge headquarters cannot easily audit. Supplier pricing history, maintenance records and shift scheduling logic often live in personal relationships built over years. Rarely does that knowledge sit in a shared system. When headquarters requests data, what comes back is a summary filtered through that same local knowledge. No other version of it currently exists to send. Board diagnostic checklist: key warning signs of operational breakdown in foreign subsidiaries These conditions are observable from headquarters, without commissioning a further review: Where two or more of these are present at once, passive governance has reached its limit. Multiple governance research studies on subsidiary oversight and operational systems point to the same conclusion. The decision in front of the board is which executive authority to put on site. It is not whether to request one more report. Executive intervention framework: matching operational patterns to interim leadership roles Intervention does not mean sending a corporate team from Germany for a further review. It means matching the pattern already visible in the conditions above to the specific executive authority it requires. That authority then needs to be on site quickly enough to still change the outcome. Pattern on the ground Executive role required Authority carried Typical duration Shopfloor execution has broken down: scrap above 5%, on-time delivery below 85%, downtime unmanaged Interim Plant Manager Direct authority over shift scheduling, shopfloor discipline, maintenance and quality gates 3 to 6 months Procurement, engineering and production are working against each other Interim COO Cross-departmental authority to realign supply chain, production flow and local engineering 6 to 9 months Cash burn has become a structural risk: negative EBITDA, creditor pressure Interim CRO Statutory managing director (jednatel) authority to restructure the balance sheet, renegotiate terms and resize the footprint 6 to 12 months Board trust in local leadership has been lost and compliance has broken down systemically Interim CEO or Managing Director Full enterprise leadership, direct interface with the group board, works councils, key customers and banks 6 to 12 months Implementation sequencing: deploying statutory authority for rapid plant turnaround Matching the pattern to the role is only the first decision. The second is sequencing. A Plant Manager mandate that later needs CRO-level authority to renegotiate supplier terms costs weeks in escalation and remobilisation. That is why the diagnostic above should be run honestly rather than optimistically. Run it honestly at the point the board decides what to commission. McKinsey & Company research on transformation in distributed operations makes a related point. Interventions that work

Stop Managing the Plant from France: How Shadow Management Destroys Local Accountability in Polish Plants

In brief When a French group’s functional leaders begin instructing a Polish plant’s supervisors directly, the site loses the authority it needs to run daily operations. Headquarters loses the accountability it was trying to strengthen. The answer is not less group involvement. It is a clearer division of it. Standards, capital and escalation thresholds stay at headquarters. Daily production decisions stay on site, under one accountable executive on the ground. Where local leadership has already weakened, an interim plant manager or managing director can carry that authority while the group rebuilds its permanent team. How Fragmented Decision-Making and Dual Ownership Erode Plant Efficiency Picture a single decision: a press goes down mid-shift near Katowice, and the supervisor needs to authorise overtime to protect tomorrow’s delivery to a French assembly plant. Eighteen months ago, that decision belonged to the Polish plant manager, taken in minutes. Today it also belongs, informally, to a group operations director in Paris, copied on every shift report since a delivery miss put the site under scrutiny. Neither person asked for this. When performance first slipped, increasing oversight was reasonable. A group quality director asking for daily scrap data instead of weekly is doing exactly what the situation calls for. Procurement retaining a supplier decision where the commercial exposure sits is sound governance too. Each step, alone, is defensible. The difficulty is what happens when several accumulate on the same plant at once. A daily call here, a request for raw data there. Eighteen months later, that overtime decision has two owners. A shift supervisor now takes direction from three people in France and one on site, and none of the four sees what the others have said. The plant manager, still accountable for the numbers, is no longer the person the shop floor actually asks. This is what is usually meant by shadow management: a second, informal instruction line from group functions into the plant’s operating layer, alongside the formal one. Research by McKinsey & Company on breaking up matrix complexity describes the mechanism. As decision rights spread across matrix lines, coordination work goes up while individual ownership goes down. A decision with two owners takes twice as long to make, or does not get made at all. Cross-Border Operational Challenges Between French HQ and Polish Subsidiaries Distance and time zones are the least of it. Three structural features of the corridor make the same drift more consequential than inside a single country. The Polish plant is a legal entity, not a department. Its managing director is a statutory officer with duties a group function in France cannot assume on their behalf. When instructions arrive from people holding no formal role in that entity, the person carrying legal responsibility is executing decisions they did not make. It is a common reason strong operational leaders resign from otherwise attractive roles. Customer and audit accountability attaches to the site, not the function that advised it. Under IATF and customer-specific requirements, the plant must demonstrate control of its own processes. A group function can set the standard, but only the site can prove it meets that standard. Group functions see the result but rarely the constraint behind it. A cycle-time target set in France is reasonable. Whether the plant can hit it this week depends on which press is down, which operator qualification has lapsed, and which container is late. That reaches headquarters, if at all, after the shift in which it mattered. Add the local consultation required before shift patterns change, and the pattern is clear. The instruction from France is usually sound. The route it travels to the shop floor is what causes the damage. The Operational and Business Costs of Unchecked Shadow Management The first thing lost is not a metric. It is the people who would have executed the recovery. Capable production heads, engineers and quality managers leave roles where accountability and authority have separated. They leave early, since good people are easy to place in Poland. A group that lets this run for a year often still has the original problem, and no leadership left to solve it. The second cost is harder to reverse. Once a customer’s programme manager learns that France, not the site, now decides on their parts, they escalate to France and stop calling the site. Local authority then must be rebuilt in front of the customer, a slower process than restoring it internally. Key Symptoms Indicating Centralized HQ Interference in Local Plant Operations The clearest sign is a change in how local management answers a performance question. When the answer points to a group instruction rather than a root cause, this can look like defensiveness. In fact, it precisely shows who made the decision, and where. Alongside it, group functional specialists find much of their week spent with the plant’s supervisory layer rather than its management. Few set out for this; it arrived one call at a time. Two symptoms tend to follow. Maintenance and tooling decisions that once took an hour now take two days. Nobody can name the approval step that added the delay, because no one ever wrote it into a process. Disputes once settled on the shop floor travel up two functional lines in France and return unresolved. The judgement is not how many are present, but whether the site’s operating layer has stopped absorbing normal variation on its own. Once it has, more reporting will not restore it. What is missing is a single point of authority both the group and the shop floor recognise. Initial Decision Mapping and Escalation Strategies for Operational Leaders An experienced operations executive does not start with scrap or OEE. Those are outputs, and by now both sides dispute what they mean. The first task is a decision map: for the twenty or so decisions that recur weekly, who actually takes them today, and how long does each take? Not who the organisation chart says. Who the supervisor calls. It takes two or three days, and is often

Why shopfloor discipline breaks down in German-owned Romanian plants

Shopfloor discipline in Romanian manufacturing plants rarely breaks down because of culture. Learn how leader standard work, visual management, clear escalation, and stronger supervisory authority restore consistent execution.

Why cross-border transformation starts with one verified fact base

In brief Transformation cannot be governed when headquarters and the local operation are working from different definitions, assumptions and figures. Before targets are set or initiatives launched, the first leadership task is to establish one verified view of the business: what cash is really available, what the order book actually commits to, what quality and capacity genuinely allow. Everything downstream, including the credibility of the plan itself, rests on that agreement. Where the fact base is contested, decisions stall or get taken twice. Why Subsidiaries and Headquarters Diverge: Operational vs. Financial Reporting No one sets out to run two versions of a business. It happens because a group needs comparability and a plant needs to run. Group finance defines revenue on a consolidated basis, recognises it under group policy, and reports monthly on a calendar the whole portfolio shares. The local operation measures what it can see and act on: what shipped, what the customer accepted, what is sitting in the yard waiting for a part. Both are accurate within their own frame. Neither is complete. Given eighteen months, the two frames drift far enough apart that a single word stops being reliable. “Backlog” means confirmed orders in one place and everything in the pipeline in another. “On-time delivery” is measured against the original promise at group and against the last revised promise locally. The gap is not concealment. It is the definition. Addressing Reporting Discrepancies: Why Boards Hesitate to Challenge Numbers This is where boards hesitate, and the hesitation is worth naming. Challenging the numbers feels like challenging the people. A group CEO who reopens the fact base is making an implicit statement about the managing director they appointed, the finance director who signs the pack, and their own judgement in accepting both for the last six quarters. There is also the quieter problem: if the definitions were wrong, then the decisions built on them were taken on the wrong basis, and some of those decisions were the board’s. So the fact base stays unexamined for longer than it should, while the operational explanation gets rehearsed instead. It was a timing difference. It was one bad month. The customer moved the schedule. Meanwhile the local team is usually working under constraints it has not been asked about directly. Group expectations set at portfolio level may not reflect what the site can produce with the tooling, headcount and supplier terms it has. Escalating that costs something politically. Absorbing it quietly costs less, until it cannot be absorbed. Two people keeping the household accounts in separate notebooks will both be honest and will never agree, and the argument when it comes will be about the money rather than the notebooks. The Complexity of Governing Cross-Border Industrial Operations Distance changes the mechanics, not just the mood. Information reaches headquarters aggregated and late, having passed through a local ledger, a statutory framework and a consolidation layer, each of which is legitimate and each of which removes detail. A group CFO in Munich reading a Hungarian subsidiary’s pack is reading an interpretation of an interpretation, and the operational facts that would explain the variance sit two translations away. The exposure is not marginal. Across the EU, foreign-controlled enterprises make up around 1% of market producer businesses but generate roughly a quarter of total value added, and in several Central European economies the concentration is far higher. Foreign-controlled enterprises accounted for 50% of value added in Slovakia, and 28% of jobs in both Slovakia and Czechia in 2023. A great deal of European industrial output is governed from a country other than the one it is produced in. Add statutory reporting that differs from group policy, ERP instances that were localised at implementation and never reconciled, and a management layer translating between two accounting logics every month, and the divergence becomes structural. It is not a language problem or a cultural one. It is a question of which numbers carry authority, who is permitted to change a definition, and how long it takes for an operational fact to reach the person accountable for it. Identifying the Signs of a Compromised Fact Base in Governance Not approaching this situation. Already in it. The last one is the reliable signal. Once decisions start waiting for agreement about the facts, the fact base has become the constraint on the business. Six Critical Metrics for Verifying a Single Business Truth Six areas carry almost all of the risk. Each needs a single agreed definition, an owner, and a documented source system. This exercise is unglamorous and it is where the value is decided. McKinsey’s research across 15 years of transformations found that completing a comprehensive, fact-based assessment of the business is one of three actions most predictive of a transformation capturing its full value, and that nearly a quarter of all value loss occurs during target setting, before implementation begins. Targets set on a contested fact base are compromised on the day they are agreed. The scale of ordinary error is easy to underestimate. In a Harvard Business Review study in which 75 executives assessed 100 of their own department’s records, 47% of newly created records contained at least one critical error, and only 3% of the resulting data quality scores were acceptable even on the loosest standard. The sample is small and self-assessed, and the study is now some years old, but the direction is consistent with what turns up whenever a group looks properly. Implementing a Fact-Based Decision-Making Framework Verification is not an audit. An audit establishes what happened. This establishes what is true now, so that a decision can be taken this week. The sequence that works is short. Agree the definitions in writing. Name one owner per figure. Fix the source system for each, so the same number cannot be produced two ways. Restate the last two quarters on the new basis, which is uncomfortable and necessary, because a new baseline without history gives the board nothing to judge movement against. Then set targets.