AI Agents for Standard Cost
and Revaluation
A variance that is the same sign and similar size every month is not performance. It is a standard getting old.
- A variance with the same sign and similar size every month is a stale standard rather than performance, and the shape across periods is what tells you which.
- An item whose standard barely moved when its market price did usually had its new standard derived from the old one rather than from actuals.
- Items with no recent purchase have no actual to roll from. Those should be listed and reviewed rather than rolled silently.
- Inventory revalues once. Cost of sales carries the new standard all year, and that is the larger number and the one nobody forecasts.
- Where a material ledger exists, reconcile level by level from the lowest, because a raw material error appears three levels up as an untraceable variance.
AI agents for standard cost separate variance caused by a stale standard from variance caused by performance. ChatFin reads the shape of each variance across twelve periods rather than its size in one, because a difference that never changes sign or magnitude is telling you about the standard and not about the plant.
Before a roll it produces the review nobody has time for: items with no recent actual to roll from, standards that did not move when the underlying price did, stale bills of material, and work in progress carrying old standards across the roll date.
A standard cost is a decision that ages
A standard is set once a year against expectations that were reasonable at the time. Every month after that, the gap between standard and actual becomes a variance, and the variance is reported as though it were performance when a good deal of it is just the standard getting old.
| Variance reported as | What it often is | How to tell |
|---|---|---|
| Purchase price variance | The standard was set before a price rise | It is one directional |
| Material usage variance | A bill of materials that was never updated | It is constant |
| Labour efficiency | A routing time from two process changes ago | It never varies |
| Overhead absorption | A volume assumption that no longer holds | It tracks volume |
| Mix variance | Standards set at different times | It moves with product mix |
A variance that is the same sign and similar size every month is not performance, it is a stale standard, and ChatFin separates those by looking at the shape of the variance across periods rather than at its size within one.
ChatFin separates variance caused by a stale standard from variance caused by performance, by reading the shape across periods rather than the size in one.
The cost roll is where a year of drift lands at once
Rolling new standards is a large, infrequent, mostly manual exercise with a revaluation attached, and the revaluation hits inventory and cost of sales in one period. Most of the risk sits in what nobody checked before pressing the button.
ChatFin produces the pre roll review: items with no recent actual, standards that did not move when they should have, and work in progress carrying old standards.
Say what the revaluation will do before it does it
The revaluation entry is usually explained after it posts, which is the wrong order. It is entirely predictable beforehand, and the value of predicting it is that the number can be questioned while it can still change.
Split by cause
How much of the revaluation is material price, how much is labour and routing, how much is overhead rate. Three different conversations with three different owners.
Split by location and entity
The profit and loss effect lands where the stock is held. On a multi entity group that moves result between entities and, if they report in different currencies, adds a translation effect on top.
Show the gross, not just the net
Revaluations up and down net to a modest number and hide two large movements. The gross tells you how much of the standard set was wrong.
Model the cost of sales effect forward
Inventory revalues once. Cost of sales carries the new standard for the whole year, and that is the larger number and the one nobody forecasts.
ChatFin produces all four before the roll runs. The margin conversation that normally happens in the second month after a roll can then happen before it.
Where a material ledger exists, the standard is not the whole story
Groups running a material ledger carry both a standard and a periodic actual cost, and the reconciliation between them is a separate exercise that many teams treat as a black box.
| What it does | What it hides | What to check |
|---|---|---|
| Rolls actual cost per period | Which differences were absorbed | The settlement result |
| Revalues consumption | Multi level effects | Each production level in turn |
| Splits price and exchange | Only if configured to | Whether the split is switched on |
| Handles multi level BOMs | Errors compound upward | The lowest level first |
| Produces a periodic unit price | How far it drifted from standard | The spread |
ChatFin reconciles the standard, the periodic actual and what was posted, level by level from the lowest, because an error at raw material level appears as an unexplained variance three levels up where nobody can trace it.
ChatFin reconciles standard against periodic actual level by level from the bottom, because a low level error surfaces as an untraceable variance higher up.
Where this underperforms
Three.
Deeper in cost accounting
Where the rest of the inventory and cost work sits.
The pillar above this
AI for inventory and cost accounting
Sibling hubs
Inventory reconciliation · Manufacturing variance · WIP and COGS · Cost allocation
Neighbouring functions
Month end close · Plex, manufacturing finance
Questions cost accountants ask
How do you tell a stale standard from real performance?
By shape rather than size. Performance variance moves around, changes sign and responds to what happened that month. A stale standard produces a variance with the same sign and similar magnitude every period. ChatFin reads twelve periods per item and sorts by how constant the variance is.
What should be reviewed before a cost roll?
Items whose standard barely moved when their market price did, items with no recent purchase and therefore no actual to roll from, bills of material that have not been updated since the last process change, and work in progress that will carry old standards across the roll date.
Why predict the revaluation before it posts?
Because after it posts the number is explained rather than questioned. Split by cause it is three conversations with three owners, and split by location it shows which entity absorbs the effect. The cost of sales effect over the following year is larger than the inventory hit and is almost never forecast.
What does a material ledger change?
It gives you a periodic actual alongside the standard, and a reconciliation between them that is often treated as a black box. ChatFin reconciles standard, periodic actual and what was posted level by level from the lowest, because errors compound upward and become untraceable.
Read the shape, review before the roll.
ChatFin reads variance shape across twelve periods per item, which separates a stale standard from real performance.
Before a roll it lists the items with no actual to roll from and the standards that should have moved, and splits the coming revaluation by cause and by location.
Bring twelve periods of variance by item.
Book a demoAll images are served from the Pexels content network. No third party product screenshots, vendor logos or licensed media are used on this page.
Bring twelve periods of variance by item.
We will run them through ChatFin and show you how much of what is reported as performance is actually a standard that needs rolling.