The costing method decision is made once at implementation and is painful to reverse. Here is what each actually does to your numbers.
Standard costing values inventory at a pre-set cost and posts the difference from reality as a variance. Actual costing values it at what was genuinely paid and consumed, so the cost moves with every purchase. Standard gives stable margins and visible variances; actual gives a truer unit cost and a noisier P&L. Most manufacturers run standard; most distributors run actual or moving average.
This is one of the few ERP decisions that is genuinely difficult to change after go-live, because the valuation method is embedded in every historical transaction. It also determines what your management accounts can tell you: under standard costing a margin problem shows up as a variance you can investigate; under actual costing it shows up as a margin that simply moved, with the cause buried.
A cost roll builds standard cost from the BOM and routing at current rates. This becomes the valuation for the period.
Receipts, issues and completions all value at standard, so inventory is stable and predictable.
Purchase price variance, usage variance, labour and overhead variances each post to their own account as reality diverges.
This is the point of the method. A persistent purchase price variance on one item is a procurement conversation, not an accounting entry.
Standards go stale. Most businesses re-roll annually, or more often where input prices move sharply.
A components business buys a part at a standard of 100 rupees. Over one month it receives three batches: 400 at 96, 600 at 104, and 300 at 118 after a supplier surcharge. It issues 1,000 to production.
Under standard costing every issue is valued at 100, so production cost is a clean 1,00,000 and the differences land in purchase price variance: 1,600 favourable, 2,400 unfavourable, 5,400 unfavourable — a net 6,200 unfavourable, sitting in one account where a buyer can see it and act. Under moving average the issue value drifts through the month and production cost comes out near 1,03,800, with no variance account and nothing that obviously points at the surcharge.
Neither total is wrong; they answer different questions. Standard costing tells you production consumed what it should have and purchasing had a bad month, which is actionable by two different people. Moving average tells you what the stock actually cost, which is what you want on the balance sheet. The real decision is not which is correct but which conversation you want the numbers to start — and whether anyone is actually reviewing the variance account if you choose standard.
| Term | What it means |
|---|---|
| Standard cost | A pre-set expected cost per unit, used to value transactions as they happen and to give a stable benchmark. |
| Actual cost | What the unit genuinely cost, known only after the fact. Precise, and volatile enough to be hard to manage against. |
| Moving average | A running weighted average that updates with each receipt. A middle path, and the most common default in mid-market ERP. |
| Purchase price variance | The gap between what you paid and the standard. The cleanest single signal of whether standards are current. |
| Revaluation | Restating stock when a standard changes. Necessary, and large enough that it should never be a surprise to finance. |
A standard cost set three years ago produces variances so large they stop meaning anything, and people stop reading them.
Actual costing removes the cost roll but also removes the variance analysis. The maintenance was doing useful work.
Different methods for different item classes is legitimate, but it has to be deliberate and documented, not accidental.
Switching method later requires revaluing inventory. The entry is large, visible and awkward to explain — decide correctly first.
| Metric | What it tells you | What distorts it |
|---|---|---|
| Purchase price variance | How far standards have drifted from the market. Persistent one-way variance means the standard is stale. | Netting favourable and unfavourable across unrelated categories. |
| Standard cost age | How long since standards were last reviewed. Annual is typical; volatile input costs need more. | Reviewing only the top items by value, which leaves a long tail quietly wrong. |
| Revaluation size at rollover | Whether standards were maintained during the year or allowed to drift and then corrected in one jump. | Nothing; a large revaluation is itself the finding. |
| Margin variance by product line | Whether costing is believable at the line level, where pricing decisions are actually made. | Overhead allocated by a single driver, which flatters high-volume lines. |
We build and support the systems this page describes. Start with a free working proof of concept — no cost, no commitment.