Standard Costing: Setting and Maintaining Standards

A standard cost is what a unit of your product should cost when everything goes to plan. It is the benchmark you compare actual cost against — and that gap (the variance) is one of the most powerful management signals a manufacturer can have. Here is how to set standards that are genuinely useful.

What standard costing is for

Standard costing is not a different way of accounting — it is a layer on top of actual costing that lets you ask one critical question: how did we do this period vs how we should have done?

Without standards, "we spent ₹4.8 lakh on materials" means nothing in isolation. With a standard ("we should have spent ₹4.4 lakh for the volume produced"), the ₹40,000 overrun becomes a question worth answering.

The three standards you set

For each product, you set standards for the three cost components:

• Standard material cost — how many units of each material per finished product, at what price per unit
• Standard labour cost — how many labour hours per finished product, at what rate per hour
• Standard overhead — how much overhead is absorbed per unit, based on your absorption rate (see our post on manufacturing cost)

The standard cost per unit is the sum of these three.

How to actually set them

Three approaches, often combined:

• Engineering / BOM-based. From the bill of materials and standard times. Most accurate but assumes the BOM is current and times are realistic.
• Historical average. Last 6–12 months of actuals, with adjustments for known one-offs. Easy to compute, sometimes locks in past inefficiency.
• Market-derived. Look at what competitors achieve or what published norms suggest. Useful as a reality check.

For most SMEs the best starting point is BOM × current rates with a historical sanity check. If your engineering standard is wildly off your historical actuals, one of them is wrong — find out which.

What "standard" should mean — ideal or achievable?

Two schools of thought:

• Ideal standards — what is achievable in perfect conditions. Motivating but rarely met. Variances are always unfavourable, which becomes background noise.
• Currently attainable standards — what is achievable with reasonably efficient operations. Variances genuinely signal something. Most companies use this.

We strongly recommend currently attainable. The point of a standard is to flag exceptions worth investigating, not to manufacture guilt.

When to revise standards

A common mistake is setting standards and never revising. Standards drift out of date quickly because the world around them changes:

• Material prices change. Revise material price standards quarterly, or whenever a major input moves by more than 5%.
• Wage rates change. Annual cycle, plus on negotiated increments.
• Processes change. Whenever a new method, machine or BOM version is adopted.
• Mix changes. When the proportion of high-overhead vs low-overhead products in the mix shifts materially.

Set a calendar for review — "standards reviewed every quarter for prices, annually for everything else" — and stick to it.

The trap of stale standards

Standards that have not been updated for two years are worse than no standards at all. Every variance becomes a debate ("the standard is wrong") and the analytical value disappears.

If you have not reviewed your standards in a year and your variances are systematically large in one direction, your standards are stale. Re-set them before drawing any conclusions about operations.

Standards feed variance analysis

Standards are the enabling layer for variance analysis — splitting "we spent more than expected" into "we bought materials at a higher price than standard" vs "we used more material per unit than standard." That separation is what makes the number actionable. See our companion post on variance analysis for how to read each variance.

How Booksmor helps

Booksmor lets you maintain a standard cost on each item — built from its BOM, labour standards and an overhead absorption rate — and computes variances against actuals on every production run. Stale-standard warnings flag items whose actuals have drifted from standard. Start a 30-day free trial and put real standards in front of your operations.