Fishbowl Standard Costing: Why Your Inventory Looks Right While the Variance Account Bleeds
If you run Fishbowl on standard costing, your inventory valuation will almost never look wrong. That’s not reassurance. It’s the whole problem.
Standard costing values every unit you own at a fixed number you set in advance. Ten units at a $5 standard is $50 of inventory, full stop, no matter what you actually paid or what it actually cost to build. The balance sheet stays clean and stable by design. So when reality disagrees with the standard, and it always does eventually, the difference cannot show up in inventory. It has to go somewhere else. In Fishbowl, that somewhere is a single account called Cost Variance, and it is the account nobody looks at until month-end.
We spend a lot of time inside standard-cost databases untangling exactly this, so here’s how it actually works and how to read the account before it turns into a month of detective work.
Scope: this describes Fishbowl Advanced, the desktop product, verified against current 2026 releases. Fishbowl Drive, the newer cloud product, behaves differently in several areas; if you’re on Drive, some of this won’t map.
The one rule that explains everything downstream
Every valuation Fishbowl reports under standard costing is quantity times the part’s standard cost. That’s it. It doesn’t look at what you paid the vendor or walk your cost layers. The standard is a fixed multiplier, so the inventory asset simply cannot drift.
That sounds like a feature, and in a stable world it is. The catch is arithmetic: if the value on the shelf is locked to the standard, then any difference between the standard and what actually happened has to be booked somewhere the balance sheet won’t see it. Fishbowl books it to the Cost Variance account. Pay a vendor more than standard, and the overage lands in variance. Build a finished good whose real inputs cost more than its standard, and the whole gap lands in variance. Everything that reality does differently from your standard accumulates in that one account, while inventory stays pristine.
Where the variance actually comes from
There are three doors into the variance account, and it helps to know all three because the account itself won’t tell you which door a given dollar came through.
Purchasing. When you receive a part at a cost different from its standard, Fishbowl books the inventory at standard and shoves the difference into variance. Buy a $10-standard part for $11 and you get a dollar of variance per unit, every receipt. If your standards are stale and you are consistently paying more than them, this door drips money into variance continuously.
Manufacturing. This is usually the big one. When a work order finishes, Fishbowl adds the finished good to inventory at its standard cost and routes the entire difference between the real rolled-up cost of the build and that standard into variance. For a manufacturer, this fires on essentially every work order, which is exactly why the account can balloon while finished-goods inventory looks perfectly reasonable.
Editing a standard on stock you already own. Change a part’s standard cost while you have units on hand and Fishbowl instantly revalues those units in every report. But it does not post a journal entry to move the general ledger. So your Fishbowl valuation jumps and your GL inventory balance, which was booked at the old standard when the goods came in, sits still. That gap is real, and it surfaces later, usually at a physical count, as a difference nobody can immediately explain. When you do a sweeping standard update, plan to book one manual journal entry in your accounting system to bring the GL back in line. Where you pull those dollars from, COGS or an expense account, is a case-by-case call for you and your accountant, but the entry has to happen. Fishbowl will not make it for you.
| Door | Fires when | The variance is |
|---|---|---|
| Purchasing | You receive a part | Actual cost paid − standard |
| Manufacturing | A work order finishes | Real rolled-up build cost − finished-good standard |
| Editing a standard | You change a standard with stock on hand | Old standard − new standard, times quantity on hand (revalues reports, but no GL entry) |
The $4 finished good with $500 of inputs
Here is the scenario that sends people looking for us. A finished good carries a $4 standard. Its bill of materials actually consumes $500 of raw materials per unit. Every time a work order completes, the finished good goes into inventory at its $4 standard, the $500 of real materials leaves raw inventory, and the $496 gap has nowhere to go but the variance account. As a journal entry, one build looks like this:
| Account | Debit | Credit |
|---|---|---|
| Finished Goods Inventory (at standard) | $4 | |
| Cost Variance | $496 | |
| Raw Materials Inventory (actual consumed) | $500 |
The balance sheet dutifully shows finished goods at $4 a unit and looks fine. Raw materials drop by the real $500. And the $496 that reconciles the two lands in Cost Variance, on every single build.
When we sit down with the team, the fix usually is not in the costing engine at all. It is that the standard is wrong, or the bill of materials is misunderstood. The mechanics are just multiplication: each BOM line is quantity times unit cost, you add the lines up, and that sum is the input. The inputs have to match the output. If the output is only valued at $4, the other $496 does not vanish and it does not come from a magical pool. It gets pulled from the per-part accounts and moved to variance. That is the entire trick. There is no other place for it to go.
So triage is transaction by transaction. Pull one completed work order, lay out what went in against what came out, and ask whether it makes sense. Put it in a spreadsheet, add it up, and find the difference. Once you can explain one transaction, you can explain the pattern.
Why a legitimately large variance means “maybe you shouldn’t be on standard”
Not every bloated variance account is a bug. Sometimes it’s telling you standard costing is the wrong tool for your business right now.
Standard costing earns its keep by smoothing out short-term blips. A snowstorm spikes your freight for three weeks; a standard rides through it so your margins don’t lurch. What it is bad at is a cost environment that keeps climbing. If your input costs grow one or two percent a month, quietly and continuously, that compounds, and a standard set last year is not valuing this year’s inventory correctly. A standard carried from 2019 into 2022 through a supply-chain shock is not evening anything out. It’s just wrong, and the variance account is where “wrong” accumulates. When we see genuinely large variances, the first question isn’t “how do we fix the account,” it’s “should you be on standard at all?”
The misconfiguration version looks different and gets weird fast. Picture three levels of bills of materials all pointing at the same variance account, with costs moving in both directions: some parts undervalued, some overvalued, work-order completions running consistently over standard, a BOM that quietly got more expensive to build. All of it lands in one bucket. You stare at a swinging pile of dollars in accounting and the records do not tell you whether it was the BOM, the standards, or an input mistake. A standard is a line you draw in the sand. Missing it is information, but a single account full of net movement does not tell you why you missed.
The account-mapping traps
Two mapping mistakes account for a lot of the confusion we get called into.
The first is people mapping variance where it doesn’t belong, usually blurring the line between a variance account and an expense account. In Fishbowl, the variance account always behaves like a variance account, no matter what you named it or where it sits on your balance sheet. The behavior is baked into the costing logic, not the account’s label. Renaming it or moving it doesn’t change what posts there. It just makes the postings harder to reason about.
To find these, we lean on Fishbowl’s own import/export tools and a simple query: list every variance account defined across all parts and summarize them. Almost always a handful of parts stick out, ten units pointing somewhere the other thousand don’t, and that mismatch is the thread you pull.
The second is scope. Fishbowl gives you exactly one variance account. That is a deliberately simple design. Bigger ERPs like NetSuite break variance into separate buckets for purchasing, receiving, manufacturing, and adjustments, and shops that want it will split further into labor, material, and overhead variance. Fishbowl does none of that out of the box, which is occasionally cumbersome. In practice, the most we see clients do is split by class of item, a finished-goods variance account and a raw-goods variance account. Going finer than that, down to a per-part variance account, is rarely worth it. If you need that level of detail, you are better off writing a query inside Fishbowl and getting the data directly than trying to model it in the chart of accounts.
The zero-standard landmine
A part with a blank or zero standard is the purest version of the whole problem. Under standard costing, Fishbowl values that part’s inventory at zero and routes one hundred percent of its actual cost into variance, on every receipt and every build. There’s no guard and no warning. It just transacts silently at zero.
Most of the time we catch this before go-live, because it is a trivial query: show me every part whose standard is zero, or under a dollar, and hand the list to the client with “are these actually right?” The dangerous ones are the parts that look fine until they don’t. A length-based part is the classic. You stock cable in inches or feet but buy it in 500- or 1,000-foot rolls. Someone puts the roll’s cost into the standard field, which is a perfectly reasonable-looking positive number, except the part is denominated in inches. If there is no stock on hand yet, say it is a part you won’t purchase until a few months after go-live, nothing flags it. It’s a positive number with zero quantity, so there is no valuation impact and nothing to compare against. Then the first receipt lands and someone asks why inventory just went to nine million dollars. Because the thing you count in inches is carrying the unit cost of a thousand-foot roll. (If that failure sounds familiar, it is the same unit-of-measure trap from our deep dive on units of measure, wearing a standard-cost hat.)
Switching costing methods without wrecking your valuation
Occasionally the right answer is to move a client off standard, or onto it. Fishbowl’s setup wizard tells you the costing method cannot be changed once you finish, and for a casual user that warning should be taken at face value. It can be done, but it is not something you do willy-nilly.
When we do it, it is a planned procedure with the accounting team’s understanding and sign-off, never a surprise. We pick a moment in time and stop transactions. We take an inventory valuation at that moment. Then we make the change and seed the destination method’s numbers so that, for every part, unit cost times quantity adds up to the same total before and after. A part valued at $377 under standard should read $377 under average when we’re done, give or take rounding. That equality is how we know the valuation landed correctly: if the total doesn’t meaningfully move, we’ve reseated it right. The work is not long once you know exactly what has to happen. It just requires knowing exactly what has to happen, which is most of what you are paying for.
Cleaning up a year of neglect
The harder engagement is a client who ran standard costing for a year with zero or blank standards on half their parts and piled everything into variance. Where you start depends on the part.
For raw materials, the cost source is purchases, so we validate against enough purchase history to establish a real number, or against a prior standard that existed but never got loaded. We give the data to the accounting team plainly: these parts are at zero, you have quantity on hand, they are supposed to be worth something, tell me what’s appropriate. Sometimes they take the last purchase price and let it float for the year. Sometimes they want a computed average. Sometimes it’s just “fill in the last price, get it correct, we’ll adjust.” We can’t make QuickBooks match COGS transaction by transaction after the fact, but we can revalue the transactions inside Fishbowl against a clean number, which is what actually matters for getting the valuation straight.
Finished and work-in-process goods are trickier, because their standard was wrong precisely because the input raws were at zero. Zero times any quantity is still zero, so the error compounds up through the build. There the approach is to work toward a stable number across raw goods and finished goods together, use that to value inventory, and make the adjustments from there. It can take real effort to sort out which way is up. It can also be done.
What we’d tell a controller running standard costing
- Watch the variance account more than once a month. The teams that use standard for a good reason do watch it, but usually only at month-end, and by then a problem has a month of history to investigate. A weekly glance turns a forensic project back into a quick correction.
- A clean inventory number is not proof costing is healthy. Under standard, inventory is clean by construction. The health signal is in the variance account, not the balance sheet.
- Editing a standard is a financial event even though Fishbowl won’t post one. After a sweeping standard change, book the corresponding journal entry yourself to keep the GL and Fishbowl aligned.
- The variance account behaves like variance no matter what you call it. Renaming or remapping it changes nothing about what posts there. Map it deliberately and audit which parts point where.
- Set standards to represent roughly 90% of the year’s real value, then review on a real cadence. Yearly at minimum, quarterly if you’re serious. If you find yourself changing standards constantly to chase actual cost, that’s your sign to just use average costing instead, which carries far less maintenance.
- Question standard costing itself if variances are genuinely large. Standard smooths short-term blips. In a continuously rising cost environment it just accumulates “wrong” in one account. Total variance movement of millions that nets to a little is a flare, not a rounding issue.
This piece comes out of years of client work on Fishbowl, verified on current versions: reconciliations, costing cleanups, and the reports we’ve had to build to explain what standard costing is actually doing. If your variance account is doing something you can’t explain, that’s the kind of thing we untangle.