Shopify COGS reports: what five tools give your accountant at month end

What actually fails at month end

The books close, and the question arrives in the same shape every month. What did the stock you sold actually cost?

Shopify can tell you how many units left the building, to the hour. What it holds against them is one cost per variant, typed in once, and that field has never seen your forwarder’s invoice. So somebody opens a spreadsheet and rebuilds the number, usually the person you pay by the hour to close your books.

The comparison people make at this point is a feature list: forecasting, purchase orders, multi-location, barcode scanning. None of that is what fails at month end. What fails is the handover. Counting stock and costing stock are two different jobs, and most tools sold to Shopify stores do the first well and the second as a screen you look at rather than a number anyone can post. So this list is scored on one question.

If you are here because you are replacing Stocky, this is the half of it nothing replaced by default.

How we compared

Landara publishes this guide and builds one of the tools on it. Product facts and prices are as of August 2026 and checkable on each vendor’s own site.

Four questions, in the order a bookkeeper asks them. The list is ordered by the answer to the first one, weakest first. Ask any vendor here how it treats the fourth: it is the fastest way to find out whether their costing was designed by somebody who has filed a return.

1. 1Does a journal come out, or only a report? A report is a screen you read. A journal is Dr COGS / Cr Inventory, in the right period, with a number that ties and working you can still explain a year later. Only one of those closes a month.
2. 2Where does the unit cost come from? A ledger of receipts, or a field somebody typed. Your stock arrived on four shipments at four prices, and a single field holds one of them.
3. 3What happens when the freight invoice is late? The forwarder bills weeks after the container lands and the broker bills after that, by which time part of the shipment is sold. A tool that can only cost a receipt at the moment of receipt cannot be corrected by a document that arrives afterwards.
4. 4Does it know which border charges are a cost and which are not? Recoverable import tax (GST, VAT, or whatever your border charges) is claimed back, so it never goes into stock value. Customs duty is not recoverable, so it does. Both arrive on the same document, which is why software conflates them, and conflating them overstates your inventory and leaves the tax off your return. US importers can skip this one: there is no recoverable import tax at the US border.

The arithmetic, so you can check the rest of this against it

A container of 1,000 units lands. The supplier invoice is $10,000, so Shopify’s Cost per item says $10.00. You sell 400 units in three weeks. Then the forwarder bills $1,800 of freight and the customs broker bills $700 of duty plus $250 of clearance.

• •True landed cost: $12,750 across 1,000 units, so $12.75 a unit.
• •COGS on the 400 already sold is understated by 400 times $2.75, or $1,100.
• •The 600 still on hand are carried at $6,000 when they cost $7,650, undervalued by $1,650.

1. Shopify’s own tools, plus a spreadsheet

What your accountant gets: nothing they can post.

What it is: Shopify Admin now has purchase orders, transfers and adjustments. Alongside it, almost everyone keeps a spreadsheet of costs.

Right for: a domestic store buying at stable prices from a local supplier, where the invoice you pay is genuinely the cost of the goods.

What it does not do: Cost per item is one static field per variant, and editing it revalues everything on hand retroactively, including units received at a different price, with no journal recording that it happened. Shopify’s own Stocky migration guide says plainly that landed-cost or weighted-average allocation requires a third-party app.

The spreadsheet fails for a different reason. The problem is not that it is manual; it is that it is disconnected. Your store sold forty units this morning and the file has no idea.

Cost, as of August 2026: included with Shopify. The spreadsheet is free; your bookkeeper’s time is not.

Verdict: the baseline everything else here is trying to beat, and what most stores run today without having decided to.

2. TrueProfit

What your accountant gets: nothing they can post.

What it is: a net profit and lifetime value dashboard. Shopify revenue, ad spend and fees pulled together into profit per order, per product and per channel, in real time.

Right for: deciding where the next advertising dollar goes. Attribution, ad return and cohorts are what it is for, and no inventory tool here does them.

What it does not do: COGS is a number you type per variant. No cost layers built from receipts, no allocation of freight, duty or brokerage into unit cost, and nothing posts to Xero or QuickBooks Online. On the example above, a dashboard fed $10.00 reports a margin $2.75 a unit too generous, and reports it confidently.

Pricing, as of August 2026: from $35 to $200 a month.

Verdict: keep it. It answers a question nothing else here can. It just does not produce anything your accountant can post, and it never claimed to.

3. Prediko

What your accountant gets: a number they must rebuild.

What it is: Shopify-native demand planning. AI forecasting, reorder timing, purchase order automation, supplier management, and a long list of warehouse integrations.

Right for: knowing when and how much to reorder. If that is the pain, pick Prediko, not Landara. Landara does no forecasting at all and is not planning to.

What it does not do: Prediko has a COGS (FIFO) report and an Inventory Value report, which is why it belongs on a list scored this way. It looks like it passes. The reports are reports. There is no COGS journal and no inventory asset posting, and the Xero and QuickBooks integration added in February 2026 sends purchase orders across as bills and nothing else. Landed cost is a manual extra-costs field on the purchase order, with no allocation method, no capture of the freight or duty invoice, and no true-up when the broker invoice arrives.

So on the 1st your bookkeeper opens the report, reads a number off it, and writes the journal by hand. That works. It is also the step everyone thought they had bought their way out of.

Pricing, as of August 2026: Starter is $49 under $100K GMV, Scale-up $119, Growth $199 up to $2M GMV, so it scales with your revenue rather than your order count. Rated 4.9 stars across 232 Shopify App Store reviews, as of August 2026.

Verdict: a FIFO report is not a FIFO ledger. Run it for the buying decisions and solve the costing somewhere else.

4. Cin7 Core

What your accountant gets: a journal they can post.

What it is: a full inventory and order platform, formerly DEAR. Real FIFO and FEFO costing, automatic COGS journals into your accounting system, warehouses, B2B ordering, point of sale and light manufacturing.

Right for: a business past roughly $5M selling across several channels, holding stock in more than one place, doing B2B as well as retail. If that is you, pick Cin7 Core, not Landara. Landara does none of those things.

What it does not do: landed cost is manual. A freight bill arriving separately, after the goods were received, needs a blind receipt workaround rather than a native late-cost path, which is question 3 answered by a procedure rather than by the software. Its AI document capture add-on reads customer purchase orders into sales orders, not your forwarder’s invoice.

Pricing, as of August 2026: Standard $349, Pro $599, Advanced $1,199, USD only. Standard covers 5 users, 6,000 orders a year and 2 integrations. Document capture is a separate $79 a month.

Verdict: the only tool here that both costs on FIFO and posts the journal, inside a platform that replaces how Shopify tracks your stock. That is the trade. You get the month-end answer and you get an implementation.

5. Landara

What your accountant gets: a journal to review and post.

What it is: the cost side of a Shopify store, and nothing else. Shopify keeps counting stock. Landara attaches a dated cost layer to every receipt, draws those layers down oldest-first as orders ship, and produces a ready-to-post Dr COGS / Cr Inventory journal for Xero or QuickBooks Online. You review it and you post it, always: it never posts to your ledger by itself. COGS is replayed from the ledger rather than stored, so a cost corrected after posting yields a correction journal, never a rewritten history.

Right for: Shopify stores holding imported stock, where the freight invoice lands after the goods and the books are done in Xero or QuickBooks Online.

On the four questions: purchase orders and the freight, duty and broker invoices go in as PDF, PNG or JPG and are read by AI, so the numbers arrive without retyping. Allocation is by value, quantity, weight or volume, chosen per charge, because freight follows volume or weight and duty follows value. A late invoice attaches to the shipment and the layers and the period’s COGS true up. Recoverable import tax cannot capitalise: the engine filters it out before a layer forms.

What it does not do: no forecasting, no reorder workflow, no warehouse management, no manufacturing, no B2B portal, no Amazon. If you need any of those, buy one of the tools above and run this beside it.

Pricing: free to start, then by monthly order volume, with no per-user charge.

Verdict: a cost ledger rather than an inventory system, which is either exactly what you needed or nowhere near enough.

What each one hands over

One line each, as of August 2026, from each vendor’s own documentation: where the unit cost comes from, what happens when the freight invoice is late, and what reaches your accountant.

• •Shopify plus a spreadsheet: unit cost is one typed field per variant, nothing catches a late freight invoice, and your accountant receives nothing they can post.
• •TrueProfit: unit cost is one typed number per variant, nothing catches a late freight invoice, and your accountant receives nothing they can post.
• •Prediko: unit cost comes from a FIFO report built from receipts, a late freight invoice goes into a manual extra-costs field with no true-up, and your accountant receives a number to read off and rebuild.
• •Cin7 Core: unit cost comes from FIFO cost layers, a late freight bill needs the blind receipt workaround, and your accountant receives an automatic COGS journal.
• •Landara: unit cost comes from FIFO cost layers at landed cost, a late freight invoice attaches to the shipment and trues up, and your accountant receives a ready-to-post COGS journal.

Which one should you actually pick?

The list above runs weakest first on one question. Here it is the other way round, by the situation you are actually in.

• •Prediko if your problem is when to reorder and how much.
• •TrueProfit if your problem is which ad spend is working. Keep it whatever else you run.
• •Cin7 Core if you need warehouses, B2B ordering, point of sale or manufacturing in one platform, and you are willing to implement one.
• •Landara if Shopify already tracks your stock correctly, the freight invoice keeps arriving after the sale, and what you are short of on the 1st is a journal your accountant can post.
• •Shopify plus a spreadsheet if you buy domestically at stable prices, in which case the invoice you pay really is the cost of the goods.