"Our operators are working round the clock, sometimes deep into overtime. Yet orders are still getting delayed. I requested additional manpower, twice but both times, it was rejected. The reason?Our sister plant runs the same production with far fewer people."  This was said by the head of one plant at a prominent Indian pharmaceutical company. It is not an isolated complaint. Across many large generic pharma exporters, the same dynamic is playing out: plants are under pressure to match the manpower efficiency of their best-performing counterparts, regardless of whether that comparison is structurally valid. The pressure has logic to it. Manpower costs compound persistently over time, driven by growing headcount and rising salaries. For companies operating multiple plants, benchmarking output per unit of manpower appears to be a rational lever. The problem is not the intention; the problem is that the comparison is often built on a flawed foundation, and the consequences of getting it wrong are severe. This article examines that paradox: how some of India's pharma companies are using benchmarking to optimise manpower while inadvertently undermining capacity planning. <h2>Why benchmarking leads to wrong conclusions about manpower capacity</h2>Two plants can produce identical monthly volumes and still have fundamentally different manpower requirements. Benchmarking misses this because it compares outputs and headcounts while treating plants as structurally equivalent. The key structural variable it most consistently ignores is the composition of what each plant actually manufactures.   Two cases illustrate this mechanism. <h4>Case 1:</h4>Two manufacturing plants, Plant A and Plant B, each produce 200 million tablets and capsules per month. At first glance, they appear equal in output. Their manpower tells a different story:<figure class="image"><img style="aspect-ratio:3355/1801;" src="https://web-assets.vectorconsulting.in/storage/editor_uploads//image_1778684687.png" width="3355" height="1801"></figure> A straightforward benchmarking exercise would immediately flag this: Plant B is producing the same volume with 30% more people. The critical variable missing from this comparison is batch size. Plant A operates with two anchor products featuring batch sizes of 5 million and 3.5 million tablets, respectively. Together, these two products alone account for 50%-60% of Plant A's total monthly output. Their scale justifies dedicated machinery and dedicated personnel. Plant B operates in an entirely different manufacturing reality. No single product has a batch size exceeding 0.8 million units. Batch size directly impacts manpower requirements, as we will see. To benchmark meaningfully, we need to strip out the outsized contribution of Plant A's two mega-batch products and examine what remains. When those products are excluded, the picture reverses entirely:<figure class="image"><img style="aspect-ratio:3229/722;" src="https://web-assets.vectorconsulting.in/storage/editor_uploads//image_1778684773.png" width="3229" height="722"></figure> With this revised view, Plant B appears more efficient than Plant A. <h4>Case 2:</h4>Plant A and Plant B produce the same monthly volume, run the same average batch sizes, and complete the same number of batches. Yet their manpower situations differ:<figure class="image"><img style="aspect-ratio:3287/1735;" src="https://web-assets.vectorconsulting.in/storage/editor_uploads//image_1778684818.png" width="3287" height="1735"></figure>At first glance, this appears to be a utilisation problem in Plant B. A detailed analysis reveals that Plant B manufactures relatively complex products, with processing times 35% higher than those of Plant A. Like batch sizes, processing time directly impacts manpower requirements. The table below illustrates the difference:  <figure class="image"><img style="aspect-ratio:3376/970;" src="https://web-assets.vectorconsulting.in/storage/editor_uploads//image_1778685040.png" width="3376" height="970"></figure>Benchmarking in this case would lead to the conclusion that Plant B's manpower is 50% higher than Plant A's, despite identical production volumes. In reality, Plant B genuinely requires 50% more manpower because of its higher processing time. In both cases, the gap in manpower is real. The error lies in interpreting it as inefficiency rather than as a structural consequence of the products being manufactured. <h2>The problems of getting manpower wrong</h2>In most pharma plants, machines are not the binding constraint, manpower is. The reason lies in how plants are built and staffed over time. Pharma plants, especially those focused on exports, enjoy gross contribution margins of 60%-80%. Capital costs per machine are modest relative to these margins, resulting in payback periods often measured in months. Because equipment lead times are long, organisations order machines well in advance based on long-term demand forecasts. Those forecasts frequently turn out to be inaccurate. A machine procured as a constraint may hold that status for only a few months before becoming a non-bottleneck, by which point it has typically paid for itself. Organisations recognise this pattern, so they do not hire full manpower for every available machine. The result is a plant with adequate equipment but structurally limited manpower. This becomes a problem in the short term for three compounding reasons:<ol><li>Products follow only approved production routes with fixed processing times and batch sizes, leaving no flexibility in how work flows through the plant.</li><li>Demand fluctuates at the product level across monthly horizons, continuously shifting the product mix.</li><li>Operators must undergo a qualification process before deployment, with each operator certified for only a limited set of operations. A plant cannot shift an operator from granulation to compression simply because the load has moved there.</li></ol> When product mix changes against this backdrop of fixed routes, fixed processing times, and rigid manpower deployment, the constraint machine shifts continuously across the plant: granulation, mixing, compression, coating, packing. The plant is not undermanned in aggregate; it is undermanned at the point that matters, at the moment that matters. The consequences then compound in two directions. <h2>1 Impact on regular orders</h2>Delayed orders attract penalties and trigger continuous follow-ups from sales teams and customers. Pressure builds on the production team. Two recovery options exist: overtime and air freight. Both come at a cost. Overtime appears to be an immediate solution, but it compounds the problem. Many companies pay overtime at premium rates. More critically, sustained overwork erodes the workforce: fatigue accumulates, absenteeism rises, and any output gains from overtime begin to disappear. When delays become severe enough to require Sunday operations, costs climb further. HVAC systems, a major driver of utility expenses, run continuously across production aisles and corridors that were never designed for seven-day operation, pushing energy bills significantly higher.Air freight resolves the delivery problem but at a steep price, typically 5-8 times the cost of sea freight. Neither option eliminates all delays. A plant facing delayed orders and poor manpower utilisation relative to benchmark figures is often labelled inefficient, and that perceived inefficiency is assumed to be the cause of delays. The plant is then pushed to improve efficiency. The logic typically follows: fewer changeovers lead to better efficiency, and longer campaigns reduce changeovers. As a result, the plant starts producing future orders in advance to build longer campaigns. The result? Future orders are produced at the expense of current ones. Output metrics improve, but so does the number of delayed orders. The efficiency drive ends up worsening the problem rather than solving it.<h2> 2 Impact on R&D projects</h2>The more damaging consequence plays out over a longer horizon and is largely invisible in any benchmarking exercise. Every pharma company simultaneously runs two operations within its manufacturing plants: the commercial business of today and the pipeline of tomorrow. New Product Development, the R&D work that eventually translates into pre-exhibit batches, exhibit batches, and ultimately product launches, depends on access to the same manufacturing equipment and operators already stretched across delayed commercial orders. When commercial and R&D batches compete for the same manpower capacity, commercial batches take precedence. R&D batches wait. But R&D batches that wait become delayed exhibit submissions. Delayed exhibit submissions push back regulatory approvals. In the generics business, where the window of opportunity after patent expiry is narrow and fiercely contested, being second to market is not a minor setback; it is a strategic failure. A company delaying R&D batch production may effectively surrender its first-mover advantage on a product representing years of investment. The return on that investment, already uncertain, diminishes further with every week lost on the production floor. These two consequences - today’s eroded margins and tomorrow’s weakened pipeline - are the true cost of a manpower model built on flawed benchmarking. The question, then, is how to calculate the right number to begin with.  <h2> Understanding manpower assessment</h2>Since benchmarking does not reliably establish manpower requirements, the correct approach is to calculate from first principles. Three parameters determine manpower requirements: <h4>1. Number of batches</h4>More batches mean more activity - more documentation, more line clearances, more equipment set-ups, and more quality checks. Even if each individual batch is straightforward, the cumulative coordination burden grows with every additional batch moving through the facility. <h4>2. Batch size</h4>As batch size increases, production shifts to larger equipment. The critical insight is that the number of operators and supervisors required per machine remains essentially constant, regardless of whether that machine is processing 100 kg or 1,000 kg. A granulator needs an operator. A coater needs an operator. The machine scales; the crew does not scale proportionally. This dynamic holds across granulation, blending, and coating operations. The implication is stark: a plant running smaller batches will inherently require more operators and supervisors than a plant running larger batches, even if total output measured, in kilograms or units, is identical. Compression, capsule filling, and packing operations behave somewhat differently; the manpower-to-batch-size relationship is more linear in these cases. But even on these machines, smaller batches bring one unavoidable consequence: more changeovers. Every time a batch or product changes, machines must be stripped, cleaned, reassembled, and re-verified. That work requires people. <h4>3. Processing time</h4>Processing time refers to the duration a batch must spend in a machine to be fully processed. The relationship is direct - longer processing time tie operators to that operation for longer periods, increasing manpower requirements. These three parameters do not operate in isolation, they interact, creating a compounding effect. A plant running a high-SKU portfolio with small batch sizes and long processing times faces a manpower burden that is structurally, not operationally, higher than that of a plant producing fewer, larger batches of simpler products. No amount of workforce efficiency or operator skill can fully close this gap. <h2>How to calculate actual manpower requirement</h2>Since benchmarking does not reliably establish manpower requirements, the correct approach is to calculate from first principles. Three parameters determine manpower requirements: The steps are as follows:<ol><li>Evaluate the load on machines based on demand. Machine load is the processing time required for all batches, plus batch changeover time and product changeover time. This represents the total machine hours required. This step ties manpower needs directly to actual work, rather than to a headcount figure inherited from last year's plan or borrowed from a sister plant. </li><li>From machine hours, calculate manhours. Machine hours indicate how long each machine needs to run; manhours tell you how many people you need standing next to those machines while they run. This conversion requires knowing how many operators are needed per machine type to run it safely and continuously. Not every machine requires the same number of operators. Some require one; others require two. For instance, a packing machine typically requires a minimum of two operators: one covering primary packing and one covering secondary. A plant running more packing-intensive products will carry a structurally higher manhour-to-machine-hour ratio than one with fewer packing requirements. The difference is legitimate, not a sign of inefficiency. </li><li>Sum up manhours across all machines to get the total manhour requirement for the plant. Dividing this total by per-operator manhours gives the total manpower requirement. Operator capacity should be adjusted for scheduled breaks. The netting against breaks matters more than it appears. In plants running extended shifts, scheduled breaks can represent 10%-15% of nominal operator hours, and ignoring this produces a manpower figure that appears adequate on paper but falls short in practice. </li><li>Adjust this manpower requirement for leaves, as companies provide statutory entitlements including casual leave, sick leave, and privilege leave. In most Indian pharma plants, this adjustment adds 10%-15% to the base manpower figure. </li><li>Add a buffer on top of the above to account for variability on both the demand and supply sides. The size of this buffer matters. It is not a fixed percentage to be applied uniformly. It depends directly on how much variability the plant is exposed to and, crucially, how much of that variability is avoidable.</li></ol> <h2>How to minimise buffers</h2>Some variability is unavoidable: genuine demand fluctuations, supplier disruptions, and quality failures. A buffer for these is legitimate and necessary. But a significant portion of the variability that drives buffer requirements in pharma plants is not market-driven; it is self-imposed through poor planning practices. Two patterns are particularly common. The first is skewed work-in-progress (WIP) flow: despite sufficient demand, some routes are overloaded while others are underloaded, a direct consequence of how material is released onto the shop floor, resulting in simultaneous congestion and idle capacity within the same plant. The second is raw material (RM) availability failure: RM shortages leave capacity idle, only to create overload when delayed material finally arrives. The plant lurches between starvation and flood, and the manpower model must absorb both extremes. Solving these planning-imposed sources of variability directly reduces the buffer required. Two interventions address them:<ol><li>Perpetual RM availability Quality Control-approved raw material for 15-20 days must always be available to ensure RM release on low-WIP routes and effective campaign formation. RM should be managed with an intelligent mix of stock-based and demand-based ordering. </li><li>Balanced WIP release RM should be released in a manner that maintains equal WIP across all routes, preventing both overload and underload. This stands in contrast to the common practice of dispensing against a fixed monthly plan.</li></ol> Meeting these conditions eliminates load spikes caused by planning dysfunction, makes manpower utilisation more efficient in execution, and reduces the buffer required in the manpower model. <h2>Closing comments</h2>The methodology described here calculates manpower from first principles, based on the inherent conditions of each plant. It is a ground-up calculation built not on what a sister plant does, but on what this plant, running these products and processing these orders, genuinely requires. The inputs are not estimates. Demand volumes, batch sizes, processing times, and changeover durations are all metrics that exist in every plant today. Once this calculation is in place, benchmarking becomes meaningful, but in a fundamentally different sense. The plant no longer measures itself against a sister facility or an industry average; it measures itself against its own past performance. With manhour consumption available at the product level, the plant can identify its highest manpower consumers and target them for continuous improvement: reducing processing time, increasing batch size, eliminating unnecessary changeovers. Each improvement directly reduces the plant's structural manpower requirement. What changes most fundamentally is the nature of the conversation inside the organisation. A plant head who can show, from first principles, exactly what their manpower requirement is and why, no longer needs to defend themselves against a comparison that was never structurally valid. And leadership, rather than applying pressure based on a cross-plant headcount ratio, can direct improvement efforts towards the variables that actually drive manpower consumption: processing time, batch size, changeover frequency. Acting on these variables reduces the plant's manpower requirement permanently. That is a more honest and more productive basis for improvement than cross-plant benchmarking has ever provided.

Key Takeaways

Stop comparing, start calculating. Discover why cross-plant benchmarking fails Indian pharma and how to use first-principle manpower modelling to slash delays, protect R&D, and optimise structural efficiency.

The benchmarking trap: How pharmaceutical plants are getting their capacity planning wrong

"Our operators are working round the clock, sometimes deep into overtime. Yet orders are still getting delayed. I requested additional manpower, twice but both times, it was rejected. The reason?Our sister plant runs the same production with far fewer people."  This was said by the head of one plant at a prominent Indian pharmaceutical company. It is not an isolated complaint. Across many large generic pharma exporters, the same dynamic is playing out: plants are under pressure to match the manpower efficiency of their best-performing counterparts, regardless of whether that comparison is structurally valid. The pressure has logic to it. Manpower costs compound persistently over time, driven by growing headcount and rising salaries. For companies operating multiple plants, benchmarking output per unit of manpower appears to be a rational lever. The problem is not the intention; the problem is that the comparison is often built on a flawed foundation, and the consequences of getting it wrong are severe. This article examines that paradox: how some of India's pharma companies are using benchmarking to optimise manpower while inadvertently undermining capacity planning. <h2>Why benchmarking leads to wrong conclusions about manpower capacity</h2>Two plants can produce identical monthly volumes and still have fundamentally different manpower requirements. Benchmarking misses this because it compares outputs and headcounts while treating plants as structurally equivalent. The key structural variable it most consistently ignores is the composition of what each plant actually manufactures.   Two cases illustrate this mechanism. <h4>Case 1:</h4>Two manufacturing plants, Plant A and Plant B, each produce 200 million tablets and capsules per month. At first glance, they appear equal in output. Their manpower tells a different story:<figure class="image"><img style="aspect-ratio:3355/1801;" src="https://web-assets.vectorconsulting.in/storage/editor_uploads//image_1778684687.png" width="3355" height="1801"></figure> A straightforward benchmarking exercise would immediately flag this: Plant B is producing the same volume with 30% more people. The critical variable missing from this comparison is batch size. Plant A operates with two anchor products featuring batch sizes of 5 million and 3.5 million tablets, respectively. Together, these two products alone account for 50%-60% of Plant A's total monthly output. Their scale justifies dedicated machinery and dedicated personnel. Plant B operates in an entirely different manufacturing reality. No single product has a batch size exceeding 0.8 million units. Batch size directly impacts manpower requirements, as we will see. To benchmark meaningfully, we need to strip out the outsized contribution of Plant A's two mega-batch products and examine what remains. When those products are excluded, the picture reverses entirely:<figure class="image"><img style="aspect-ratio:3229/722;" src="https://web-assets.vectorconsulting.in/storage/editor_uploads//image_1778684773.png" width="3229" height="722"></figure> With this revised view, Plant B appears more efficient than Plant A. <h4>Case 2:</h4>Plant A and Plant B produce the same monthly volume, run the same average batch sizes, and complete the same number of batches. Yet their manpower situations differ:<figure class="image"><img src="https://web-assets.vectorconsulting.in/storage/editor_uploads//image_1778684818.png"></figure>At first glance, this appears to be a utilisation problem in Plant B. A detailed analysis reveals that Plant B manufactures relatively complex products, with processing times 35% higher than those of Plant A. Like batch sizes, processing time directly impacts manpower requirements. The table below illustrates the difference:  <figure class="image"><img style="aspect-ratio:3376/970;" src="https://web-assets.vectorconsulting.in/storage/editor_uploads//image_1778685040.png" width="3376" height="970"></figure>Benchmarking in this case would lead to the conclusion that Plant B's manpower is 50% higher than Plant A's, despite identical production volumes. In reality, Plant B genuinely requires 50% more manpower because of its higher processing time. In both cases, the gap in manpower is real. The error lies in interpreting it as inefficiency rather than as a structural consequence of the products being manufactured. <h2>The problems of getting manpower wrong</h2>In most pharma plants, machines are not the binding constraint, manpower is. The reason lies in how plants are built and staffed over time. Pharma plants, especially those focused on exports, enjoy gross contribution margins of 60%-80%. Capital costs per machine are modest relative to these margins, resulting in payback periods often measured in months. Because equipment lead times are long, organisations order machines well in advance based on long-term demand forecasts. Those forecasts frequently turn out to be inaccurate. A machine procured as a constraint may hold that status for only a few months before becoming a non-bottleneck, by which point it has typically paid for itself. Organisations recognise this pattern, so they do not hire full manpower for every available machine. The result is a plant with adequate equipment but structurally limited manpower. This becomes a problem in the short term for three compounding reasons:<ol><li>Products follow only approved production routes with fixed processing times and batch sizes, leaving no flexibility in how work flows through the plant.</li><li>Demand fluctuates at the product level across monthly horizons, continuously shifting the product mix.</li><li>Operators must undergo a qualification process before deployment, with each operator certified for only a limited set of operations. A plant cannot shift an operator from granulation to compression simply because the load has moved there.</li></ol> When product mix changes against this backdrop of fixed routes, fixed processing times, and rigid manpower deployment, the constraint machine shifts continuously across the plant: granulation, mixing, compression, coating, packing. The plant is not undermanned in aggregate; it is undermanned at the point that matters, at the moment that matters. The consequences then compound in two directions. <h2>1 Impact on regular orders</h2>Delayed orders attract penalties and trigger continuous follow-ups from sales teams and customers. Pressure builds on the production team. Two recovery options exist: overtime and air freight. Both come at a cost. Overtime appears to be an immediate solution, but it compounds the problem. Many companies pay overtime at premium rates. More critically, sustained overwork erodes the workforce: fatigue accumulates, absenteeism rises, and any output gains from overtime begin to disappear. When delays become severe enough to require Sunday operations, costs climb further. HVAC systems, a major driver of utility expenses, run continuously across production aisles and corridors that were never designed for seven-day operation, pushing energy bills significantly higher.Air freight resolves the delivery problem but at a steep price, typically 5-8 times the cost of sea freight. Neither option eliminates all delays. A plant facing delayed orders and poor manpower utilisation relative to benchmark figures is often labelled inefficient, and that perceived inefficiency is assumed to be the cause of delays. The plant is then pushed to improve efficiency. The logic typically follows: fewer changeovers lead to better efficiency, and longer campaigns reduce changeovers. As a result, the plant starts producing future orders in advance to build longer campaigns. The result? Future orders are produced at the expense of current ones. Output metrics improve, but so does the number of delayed orders. The efficiency drive ends up worsening the problem rather than solving it.<h2> 2 Impact on R&D projects</h2>The more damaging consequence plays out over a longer horizon and is largely invisible in any benchmarking exercise. Every pharma company simultaneously runs two operations within its manufacturing plants: the commercial business of today and the pipeline of tomorrow. New Product Development, the R&D work that eventually translates into pre-exhibit batches, exhibit batches, and ultimately product launches, depends on access to the same manufacturing equipment and operators already stretched across delayed commercial orders. When commercial and R&D batches compete for the same manpower capacity, commercial batches take precedence. R&D batches wait. But R&D batches that wait become delayed exhibit submissions. Delayed exhibit submissions push back regulatory approvals. In the generics business, where the window of opportunity after patent expiry is narrow and fiercely contested, being second to market is not a minor setback; it is a strategic failure. A company delaying R&D batch production may effectively surrender its first-mover advantage on a product representing years of investment. The return on that investment, already uncertain, diminishes further with every week lost on the production floor. These two consequences - today’s eroded margins and tomorrow’s weakened pipeline - are the true cost of a manpower model built on flawed benchmarking. The question, then, is how to calculate the right number to begin with.  <h2> Understanding manpower assessment</h2>Since benchmarking does not reliably establish manpower requirements, the correct approach is to calculate from first principles. Three parameters determine manpower requirements: <h4>1. Number of batches</h4>More batches mean more activity - more documentation, more line clearances, more equipment set-ups, and more quality checks. Even if each individual batch is straightforward, the cumulative coordination burden grows with every additional batch moving through the facility. <h4>2. Batch size</h4>As batch size increases, production shifts to larger equipment. The critical insight is that the number of operators and supervisors required per machine remains essentially constant, regardless of whether that machine is processing 100 kg or 1,000 kg. A granulator needs an operator. A coater needs an operator. The machine scales; the crew does not scale proportionally. This dynamic holds across granulation, blending, and coating operations. The implication is stark: a plant running smaller batches will inherently require more operators and supervisors than a plant running larger batches, even if total output measured, in kilograms or units, is identical. Compression, capsule filling, and packing operations behave somewhat differently; the manpower-to-batch-size relationship is more linear in these cases. But even on these machines, smaller batches bring one unavoidable consequence: more changeovers. Every time a batch or product changes, machines must be stripped, cleaned, reassembled, and re-verified. That work requires people. <h4>3. Processing time</h4>Processing time refers to the duration a batch must spend in a machine to be fully processed. The relationship is direct - longer processing time tie operators to that operation for longer periods, increasing manpower requirements. These three parameters do not operate in isolation, they interact, creating a compounding effect. A plant running a high-SKU portfolio with small batch sizes and long processing times faces a manpower burden that is structurally, not operationally, higher than that of a plant producing fewer, larger batches of simpler products. No amount of workforce efficiency or operator skill can fully close this gap. <h2>How to calculate actual manpower requirement</h2>Since benchmarking does not reliably establish manpower requirements, the correct approach is to calculate from first principles. Three parameters determine manpower requirements: The steps are as follows:<ol><li>Evaluate the load on machines based on demand. Machine load is the processing time required for all batches, plus batch changeover time and product changeover time. This represents the total machine hours required. This step ties manpower needs directly to actual work, rather than to a headcount figure inherited from last year's plan or borrowed from a sister plant. </li><li>From machine hours, calculate manhours. Machine hours indicate how long each machine needs to run; manhours tell you how many people you need standing next to those machines while they run. This conversion requires knowing how many operators are needed per machine type to run it safely and continuously. Not every machine requires the same number of operators. Some require one; others require two. For instance, a packing machine typically requires a minimum of two operators: one covering primary packing and one covering secondary. A plant running more packing-intensive products will carry a structurally higher manhour-to-machine-hour ratio than one with fewer packing requirements. The difference is legitimate, not a sign of inefficiency. </li><li>Sum up manhours across all machines to get the total manhour requirement for the plant. Dividing this total by per-operator manhours gives the total manpower requirement. Operator capacity should be adjusted for scheduled breaks. The netting against breaks matters more than it appears. In plants running extended shifts, scheduled breaks can represent 10%-15% of nominal operator hours, and ignoring this produces a manpower figure that appears adequate on paper but falls short in practice. </li><li>Adjust this manpower requirement for leaves, as companies provide statutory entitlements including casual leave, sick leave, and privilege leave. In most Indian pharma plants, this adjustment adds 10%-15% to the base manpower figure. </li><li>Add a buffer on top of the above to account for variability on both the demand and supply sides. The size of this buffer matters. It is not a fixed percentage to be applied uniformly. It depends directly on how much variability the plant is exposed to and, crucially, how much of that variability is avoidable.</li></ol> <h2>How to minimise buffers</h2>Some variability is unavoidable: genuine demand fluctuations, supplier disruptions, and quality failures. A buffer for these is legitimate and necessary. But a significant portion of the variability that drives buffer requirements in pharma plants is not market-driven; it is self-imposed through poor planning practices. Two patterns are particularly common. The first is skewed work-in-progress (WIP) flow: despite sufficient demand, some routes are overloaded while others are underloaded, a direct consequence of how material is released onto the shop floor, resulting in simultaneous congestion and idle capacity within the same plant. The second is raw material (RM) availability failure: RM shortages leave capacity idle, only to create overload when delayed material finally arrives. The plant lurches between starvation and flood, and the manpower model must absorb both extremes. Solving these planning-imposed sources of variability directly reduces the buffer required. Two interventions address them:<ol><li>Perpetual RM availability Quality Control-approved raw material for 15-20 days must always be available to ensure RM release on low-WIP routes and effective campaign formation. RM should be managed with an intelligent mix of stock-based and demand-based ordering. </li><li>Balanced WIP release RM should be released in a manner that maintains equal WIP across all routes, preventing both overload and underload. This stands in contrast to the common practice of dispensing against a fixed monthly plan.</li></ol> Meeting these conditions eliminates load spikes caused by planning dysfunction, makes manpower utilisation more efficient in execution, and reduces the buffer required in the manpower model. <h2>Closing comments</h2>The methodology described here calculates manpower from first principles, based on the inherent conditions of each plant. It is a ground-up calculation built not on what a sister plant does, but on what this plant, running these products and processing these orders, genuinely requires. The inputs are not estimates. Demand volumes, batch sizes, processing times, and changeover durations are all metrics that exist in every plant today. Once this calculation is in place, benchmarking becomes meaningful, but in a fundamentally different sense. The plant no longer measures itself against a sister facility or an industry average; it measures itself against its own past performance. With manhour consumption available at the product level, the plant can identify its highest manpower consumers and target them for continuous improvement: reducing processing time, increasing batch size, eliminating unnecessary changeovers. Each improvement directly reduces the plant's structural manpower requirement. What changes most fundamentally is the nature of the conversation inside the organisation. A plant head who can show, from first principles, exactly what their manpower requirement is and why, no longer needs to defend themselves against a comparison that was never structurally valid. And leadership, rather than applying pressure based on a cross-plant headcount ratio, can direct improvement efforts towards the variables that actually drive manpower consumption: processing time, batch size, changeover frequency. Acting on these variables reduces the plant's manpower requirement permanently. That is a more honest and more productive basis for improvement than cross-plant benchmarking has ever provided.

Download PDF

Discover how manpower benchmarking in pharma plants can mislead capacity planning and how first-principles modelling protects margins and R&D pipelines.