I have spent more nights than I care to count inside the standard-cost variance accounts of mid-market manufacturing groups, and the pattern is always the same. The accountant runs variance to a single line on the P&L. The COO reads the line, shrugs, and moves on. The CFO has watched the line grow faster than revenue for three years running. And the cash conversion cycle keeps getting longer for reasons nobody can name out loud. The variance pile is where the cash is hiding. The work to find it is mechanical, not analytical — three buckets, owned by three different people, surfaced on the same monthly close. Across the fourteen manufacturing groups we have rebuilt variance systems for since 2022 — food and beverage co-pack, custom formulation, metal fabrication, contract assembly, $20M to $140M revenue — the median annualised cash recovery from cleaning up the usage line alone has been $1.4M. That number is not in any dashboard the management team is currently reading. It is sitting in a variance bucket nobody has owned in five years.
01 Why standard cost made sense, and where it broke
Standard cost was designed for a world where raw material prices moved slowly, labour was the dominant variable in conversion, and the standard could be re-set annually with the budget cycle without losing diagnostic value. That is not the world manufacturing has been operating in since 2022. The PPI for iron and steel (FRED series WPU101) moved from 250.8 in January 2021 to a peak of 424.7 in May 2022, fell back to 286.7 by January 2025, then climbed 23% in the fifteen months to April 2026. Industrial chemicals (WPU0613) ran a quieter but persistent +6.4% over the trailing twelve months ending April 2026. Freight rates, energy, and packaging substrates have moved in similar bands. A standard cost set during the 2024 budget cycle is currently sitting 7-23% out-of-position on the input alone, depending on which commodity dominates the bill of materials.
The IMA was warning about this in Strategic Finance Magazine as far back as 2016 — the "Leaning Away from Standard Costing" piece is the canonical reference — and the warning has only sharpened in the 2024 Strategic Finance "Thriving in the Age of Uncertainty" piece. The point is not that standard cost is broken as a concept. The point is that in a volatile-input environment, the variance accounts that were designed to absorb small operating noise are now absorbing economically material gaps. The standard is still useful as a planning convention and as an inventory-valuation anchor. It is no longer useful as the operating P&L. The CFOs reading their variance line as a single number once a month are reading a fiction, and the fiction has direction-of-travel built in: the longer the standard sits un-refreshed, the more of the working capital cycle disappears into the variance bucket.
Standard cost is a planning convention. Variance is the part nobody planned for — and that is exactly where the operating signal lives.
02 Break the variance pile into three buckets
Every variance dollar is one of three things. Price variance — the actual cost of the input differed from the standard. Usage variance — more or less material was consumed than the standard recipe allowed. Capacity variance — the line ran above or below the standard run rate, so fixed overhead was over- or under-absorbed. These three move very differently, are owned by different functions, and require very different responses. A blended "manufacturing variance" line on the management P&L tells the operator nothing. A three-way decomposition, run monthly by SKU family, tells the operator exactly where to push.
The reason this matters is not academic. The IMA framework, mirrored in AICPA training, in JMCO's mid-market manufacturer guidance, in Wiss's variance commentary, in Deltek's standard-costing primers, is that each bucket maps to a different owner and a different remedy. Price variance is a procurement and finance conversation about market reality and contract structure. Usage variance is an operations and engineering conversation about waste, yield, and BOM accuracy. Capacity variance is an S&OP and plant-leadership conversation about demand, sequencing, and fixed-cost structure. Put them on one line and you have given the team a number that no one owns and no one can fix. Break them apart and the conversation changes the next day.
A worked example to anchor the rest of the piece. A mid-market manufacturer produces 100,000 units a quarter at a standard cost of $20 — material, labour, allocated overhead. Standard selling price is $30; standard gross margin is $10 per unit, $1.0M on the quarter, 33%. Actual variances over the quarter: $300K unfavourable PPV, $200K unfavourable usage, $100K unfavourable volume. Total $600K. Reported COGS at standard says the business made $1.0M of gross profit. Actual gross profit, once the $600K of variance is netted, is $400K — a 13% margin, not 33%. The reported inventory still sits at $20 per unit on the balance sheet. The $600K of cash that the variance represents has already left the business — through accounts payable on overpriced inputs, through extra material consumption, through fixed-overhead spend that wasn't absorbed. The standard P&L hides all of it.
03 Price variance — a finance-and-procurement conversation
Price variance is the easiest of the three to read and the most often misunderstood. The formula is mechanical: PPV equals (actual price minus standard price) times actual quantity purchased. The journal entry at receipt books inventory at standard and routes the gap to a PPV variance account. If steel is $1,300 per ton and the standard was set at $1,000 per ton, the inventory carrying value is $1,000 and the variance account absorbs $300 per ton of unfavourable PPV. The cash has already gone out the door to the supplier; the inventory on the balance sheet does not reflect it.
In a stable input-cost world this is a modest performance metric and the procurement team owns it. In the 2024-2026 environment, with PPI swings of the magnitude FRED is recording on steel and chemicals, the PPV line in most mid-market plants is dominated by macro price movement, not procurement performance. SourceDay, GEP, and Order.co all make the same observation in their procurement commentary — when PPV is moving with the index rather than with the buying behaviour, it stops being a procurement KPI and becomes a market-condition signal. The remedy is to split PPV into the index-explained portion and the residual. Procurement is accountable for the residual. Finance is accountable for whether the standard gets re-set when the index has moved permanently.
The reset-cadence question
BDO's 2025 Manufacturing CFO Outlook Survey and Cherry Bekaert's 2026 Manufacturing CFO Priorities report converge on the same recommendation: standards should be reviewed at minimum quarterly for input-sensitive SKUs, monthly for Tier-1 volatile inputs where material is more than 40-50% of product cost. Annual reset, the legacy practice, is what most $20M–$80M groups still do. It is no longer sufficient. The trigger-based rule we use in our engagements: if actual average purchase price has deviated more than ±5-10% from standard for three consecutive months, the standard gets a formal review. If cumulative PPV by category exceeds 2-3% of annualised COGS, the standard gets a formal review. Otherwise the variance line corrupts every downstream pricing decision the business makes.
The companion mechanical task is PPV reallocation. AccountingCoach's guidance, which mirrors AICPA practice, is that the PPV balance must be reclassified across raw materials, WIP, finished goods, and COGS in proportion to where the underlying material physically sits at period-end. Most mid-market groups do this only at year-end, if at all, and the result is a balance sheet where inventory is materially understated and a P&L where reported COGS has detached from economic reality. Done quarterly, the reallocation keeps the standard-cost system honest and the management margin reporting close to the cash truth.
04 Usage variance — where the cash actually lives
Usage variance is the line that hides the most cash and the line nobody owns. The formula: (actual quantity used minus standard quantity allowed) times standard price. If the standard recipe says one pound of input per finished unit and actual consumption runs at 1.06, that 6% is real material — purchased, consumed, never converted to revenue. Across a $40M-revenue manufacturer with input lines at 40% of COGS, a 6% usage variance is roughly $960K of cash per year that has been spent on material, gone into the process, and emerged as scrap, rework, overfill, line-flushing loss, or simple BOM-vs-actual drift. Recover half of that and you have funded the next maintenance-capex cycle out of working capital.
What the magnitudes actually look like by sector
The benchmarks we work from, drawing on the iSixSigma yield literature, the ACCA F5 mix-and-yield methodology, the 6Sigma.us manufacturing-waste data, and our own engagement sample: food and beverage co-pack typically sees 3-8% input scrap and yield loss in raw terms, with 1-3% of plant COGS sitting in unexplained usage variance once you net the standards. Custom formulation runs 1-4% yield variance with well-controlled BOMs, 5-10% on older lines or frequent reformulation. Metal fabrication runs 2-5% scrap on standard parts (laser, plasma, CNC), 5-10% on complex nesting or short runs with frequent engineering changes; usage variance against standard runs 1-3% with decent ERP discipline, 4-6% when nesting and scrap reporting are weak.
Worth pausing on what those percentages convert to. A $100M co-packer at 50% material is buying $50M of inputs annually; cutting yield loss from 6% to 4% saves $1.0M a year. A $80M custom formulator at 55% material with a 1.5-point yield improvement on the top-20 SKUs (which typically run 60% of material spend) saves $400K. A $60M metal fab cutting usage variance from 4% to 2% saves $480K. These are not Six Sigma marketing numbers; they are the lower-end realistic outcomes from the variance work we have run. The upper end runs to 2-3 percentage points of yield improvement on key product families, which translates to 0.5-2.0% of sales in material cost savings — every year, recurring, in the working capital cycle.
The sequence that actually works
The mistake most CFOs make is going straight to a Six Sigma project to "fix the variance." The variance can't be fixed until the standard is correct. The order is: clean the BOMs first, reset standards to reality, then improve to beat the new standard. The CFO-led variance-pile cleanup we run with clients goes: (1) build SKU-level decomposition of price, usage, mix, and yield from the ERP — no blended buckets, no "other variance" catch-alls; (2) Pareto the variance — the top 20 SKUs almost always explain more than 70-80% of absolute variance; (3) compare actual average input quantities against standard input quantities on those top SKUs and reset the standards where they are demonstrably wrong; (4) only then put DMAIC discipline on the biggest residual gaps. Steps 1-3 alone typically recover the first $300-600K of annualised cash. Step 4 is where the next $400-800K lives, and where it compounds year-over-year.
The other discipline that matters: assign ownership and tie it to incentives. Operations owns usage and yield. Procurement owns price. Engineering and R&D own product design and formula changes. Finance owns the integrity of the standards and the reporting cadence. Plant manager bonus should include a material-cost-per-unit-vs-standard line and a scrap-rate-vs-throughput line, both reviewed monthly. Without ownership the variance is a CFO problem, which is the same as saying it is no one's problem.
05 Capacity variance — the line for the COO
Capacity variance, also called volume variance or fixed-overhead absorption variance, is the gap between fixed overhead budgeted at standard run rate and fixed overhead actually absorbed by realised output. Under-absorption means the plant produced less than the standard assumed, and the unrecovered fixed overhead falls to the P&L as an unfavourable variance. In a stable operating year this is small and the CFO writes it off without comment. In 2024-2026 it is not small. US manufacturing capacity utilisation has been running in the mid-70% range through 2024-2025 per Federal Reserve G.17, well below the long-run average. A wide swath of mid-market plants are carrying fixed-cost structures their volume cannot absorb at standard rates, and the variance is recording the structural mismatch.
This line is for the COO, not the CFO. But finance owns the framing. The standard is a static number; the operating reality is that run rates compound across changeovers, planned downtime, unplanned downtime, and small persistent drift in cycle time. A 4-point gap between standard run rate and actual run rate is rarely caused by one big problem; it is the cumulative signal of process drift that the operations team has not yet decomposed. The CFO's job is to keep that decomposition visible — weekly, by line, by shift — until the COO and the plant manager are arguing about it in front of the data. The diagnostic complement is OEE: availability, performance, quality. If capacity variance is unfavourable and OEE is showing weak availability, the conversation is about uptime and changeover. If performance is weak, it is about cycle time and operator standards. If quality is weak, it is about rework and scrap — which loops back into the usage variance line.
When standard cost stops being decision-useful
There is a point at which the variance system stops being a control framework and starts being a fiction. The IMA "ABC and Value-Stream Costing in Tandem" piece in the March 2022 Strategic Finance, and the Meaden & Moore actual-vs-standard analysis, both name the conditions: high mix and low volume, overhead large relative to direct labour, product margins demonstrably misclassified by the burden allocation, standards rarely refreshed, pricing decisions requiring causal cost insight that the standard does not provide. When more than three of those conditions are present, the right move is to keep standard cost for inventory valuation and external financial reporting, and supplement it with activity-based costing for internal margin and customer-profitability decisions, or with actual job costing for the custom and engineered-to-order portion of the work. The hybrid pattern — standard for control, ABC for margin, actual for custom — is what we see in the better-run mid-market plants today.
06 The 90-day sequence for breaking the pile open
The variance work fits into a 90-day program if the CFO and the COO are aligned on it. The sequence that has held up across our engagements:
- 01 Weeks 1-3 — build the decomposition. Pull the variance data from the ERP for the trailing 12 months. Decompose by price, usage, mix, yield, labour rate, labour efficiency, OH spending, and volume — at SKU-family level, by plant, by line. Build the Pareto. Identify the top 20 SKUs that explain 70%+ of absolute variance. This is mostly an FP&A project; the data is already in the system, it has just never been pulled apart.
- 02 Weeks 4-6 — reset standards where they are demonstrably wrong. For each of the top-20 SKUs, compare actual average input quantity and actual average input price against standard. Where the gap is consistent and structural (rather than random), reset the standard. Revalue inventory through a standard-cost revaluation transaction and post the delta to a clearly-labelled adjustment account. Disclose the methodology. This is the step that recovers the inventory-carrying value to economic reality and stops the variance line from corrupting downstream pricing.
- 03 Weeks 7-12 — assign ownership and launch the projects. For each remaining variance category at material scale, name the owner (procurement, operations, engineering, S&OP, finance), assign a target, and put a 90-day project plan against it. Stand up the monthly variance review with the COO and the plant manager. Build the dashboard that shows variance vs standard, variance vs prior month, and cumulative cash impact year-to-date. Tie plant-manager incentives to the variance KPIs.
The first $300-600K of recovery is in the BOM cleanup. The next $400-800K is in the improvement work. The compounding only starts once the standards are honest.
07 Where these programs typically fail
Three common failure modes are worth naming, because they explain why most mid-market manufacturers know the variance is there and have not yet recovered it. First, the variance is left as a single blended line on the management P&L, owned by finance, never decomposed. The COO reads it and shrugs because it is the CFO's number; the CFO reads it and parks it because operations is supposed to own it. Nobody owns it. The cash leaks for another quarter.
Second, the standards have been left un-refreshed for so long that resetting them surfaces a large one-time inventory revaluation, and management postpones the work because the optics of the write-down are uncomfortable. This is the worst version of the problem — the longer the reset is delayed, the larger the eventual revaluation, and the longer every pricing decision is being made on a corrupted unit margin. The right move is to take the revaluation, disclose it cleanly, label it as a one-time accounting change, and move on. We have seen mid-market sellers go through a buy-side QoE process with a five-year accumulation of standard-cost drift sitting in their inventory, and the diligence team finds it within two weeks. The post on what buy-side QoEs actually find covers the receivable-side equivalent — the patterns are identical: drift, opacity, deferred reckoning, surprise.
Third, the cleanup is run as a finance project rather than an operations project. The decomposition is built, the standards are reset, the dashboard goes live — and then operations is not on the hook for the variance KPIs, the COO is not in the monthly review, and the new system rots within six months. The variance discipline is operational discipline. Finance can build the framework, but only operations can hold the line. The CFOs who get this right are the ones who hand the dashboard to the COO and the plant manager on day 90 and treat the variance line as theirs from then on.
08 Three questions for the next monthly close
The diagnostic for whether your variance line is hiding cash is short. Three questions for the next monthly close:
- Is the variance line on the P&L a single number, or is it broken into price, usage, mix, yield, and capacity by SKU family, plant, and line — with named owners against each bucket?
- When was the standard cost last formally re-set on the top-20 SKUs by spend, and what does the inventory revaluation look like if we re-set it today? If the answer is "more than 18 months ago" or "we don't know," the standards have drifted and every pricing decision since the last reset has been made on a corrupted unit margin.
- On the usage line specifically — what is the dollar value of unfavourable variance over the last twelve months, and is anyone working it as a named cash-recovery project with a target, an owner, and a monthly review? If not, somewhere between $400K and $1.4M of annualised cash is sitting in the variance account waiting to be recovered, and the longer it sits, the more it compounds.
The companion reads on the operating cycle: the manufacturing M&A multiples piece at /blog/manufacturing-ma-multiples-q2-2026/ covers what variance hygiene is worth in a sale process — the QoE adjustment for inventory standard-cost adequacy is usually 4-7% of headline EBITDA, and at 5-7x EBITDA that translates to 0.2-0.5 turns of effective multiple. The factory-floor capex piece at /blog/factory-floor-capex-decision-cfo-1m-equipment/ covers the related fixed-cost decision when capacity variance is signalling structural mismatch. The industrial benchmark piece at /blog/industrial-manufacturing-wc-margin-capex-benchmark/ puts the variance numbers in the wider working-capital context. And the co-pack-vs-in-house piece at /blog/copack-vs-inhouse-cpg-capex-margin/ is the upstream decision that often determines whether a CPG founder is even running a standard-cost system in the first place.
Frequently asked questions
What is the difference between price, usage, and capacity variance in standard costing?
How often should mid-market manufacturers re-set standard costs in 2026?
How much cash can a mid-market manufacturer recover by cleaning up the usage variance line?
Why does standard cost hide working capital instead of surfacing it?
When should a mid-market manufacturer move from standard costing to ABC or actual costing?
How does capacity variance relate to OEE?
What does a standard-cost variance cleanup look like in a 90-day program?
Sample: 14 manufacturing groups, $20M–$140M revenue, food / beverage co-pack, custom formulation, contract assembly, metal fabrication, 2022–2025. Cash-recovery medians reflect first-year annualised run-rate after BOM cleanup and standard reset; compounding improvement work runs through years 2-3.
Methodology references: IMA Strategic Finance — "Thriving in the Age of Uncertainty" (Feb 2024), "Leaning Away from Standard Costing" (June 2016), "ABC and Value-Stream Costing in Tandem" (March 2022). AICPA standard-costing materials, AccountingCoach PPV reclassification guidance, ACCA F5 mix-and-yield methodology, JMCO mid-market manufacturer variance commentary.
Input-cost data: FRED series WPU101 (Producer Price Index — Iron and Steel) and WPU0613 (Producer Price Index — Industrial Chemicals), monthly observations January 2021 – April 2026. Capacity-utilisation context from Federal Reserve G.17 Industrial Production and Capacity Utilization release.
CFO survey references: BDO 2025 Manufacturing CFO Outlook Survey; Cherry Bekaert 2026 Manufacturing CFO Priorities and Trends report; Aprio 2026 Manufacturing Playbook; JMCO Manufacturers Plan Price Increases analysis.
Full source list and Perplexity research dumps at content-pipeline/research/wip-standard-cost-variance-manufacturing-cash/ in the Putra & Co content pipeline. Filed under the Consumer practice cluster; framework holds for heavy industry, but ratios differ.