A record sales month can squeeze your cash even while the P&L shows a profit. When the reorder is bigger than the payouts, the best month of your year can be the one that leaves the least in the bank.
Izzy Rosenzweig ran an ecommerce brand called Browze for more than a decade before he founded Portless, and on episode 479 of the podcast he described a moment almost every product brand hits once it starts to work: “Now you got to double the PO to your factory. Your million dollar month of sales is now a $800,000 PO to the factory. That’s all out the door.”
That’s the whole problem in three sentences. Sales went up, so the reorder went up, and the reorder leaves the bank before next month’s payouts arrive. Nothing in your Shopify dashboard is wrong. It’s just answering a different question than the one your bank balance answers.
This guide is about the gap between those two answers, and it’s organized by stage, because the finance work that matters at $8K a month is not the work that matters at $800K a month. If you’re just starting, read the first two sections and the Starting stage, then stop. If you’re scaling, the record month math and the Scaling stage are where the money is.
A record month leaves less cash in the bank when the extra cash you pay for new inventory, beyond the product cost the P&L already counted, plus sales still waiting to settle, adds up to more than the month’s profit. On accrual books, the P&L counts product cost when a unit sells, while the bank counts it when you pay the supplier under your terms. For a brand that pays its factory before the stock sells, those dates can sit months apart, and in a growth month the gap runs against you.
Here is the situation Izzy described, worked through. This is an illustrative example: the ratio of sales to reorder is his, and every other number is my assumption, chosen to be plausible rather than typical. It also assumes no unsettled sales carried in from the month before, every other bill paid in the month, and sales counted before payment fees, with the fees counted once in cash out. A store does $1,000,000 in its best month. Product cost runs at 30% of sales, so the P&L records $300,000 of cost of goods. Ads cost $250,000, fulfillment and payment fees $120,000, and overhead $100,000. The P&L shows $230,000 of profit, and on paper that’s a great month.
Now follow the cash. The $300,000 of product the store sold was paid for months ago, when that inventory was ordered, so it doesn’t leave the bank this month. What does leave is the $800,000 reorder to restock for the next season. About $40,000 of the last few days’ sales is also still settling: Shopify Payments lists a minimum settlement time of three business days in the US and Canada, plus one to three days for the bank to post it, and weekend sales are grouped into one payout (as of November 2026). Cash in is $960,000. Cash out is $1,270,000. The bank balance falls by $310,000 in the month the store made $230,000.
Here’s my read on it. The fix for this store is not a better accounting tool, because its books are already correct. The P&L is right and so is the bank. What it’s missing is anyone asking, before the peak month, “what does the reorder do to the bank in week six?” If I were sitting with this founder in September, my first call would be to split the $800,000 order into two or three smaller drops tied to sell through, and my second would be to hold ad spend flat until the first drop is paid for. That second call only fits a timing problem, where margin is healthy and the bank can’t fund the reorder yet. A margin problem can need the opposite, as you’ll see in the next section. Growth that the bank can’t fund is growth that ends with a credit card float or a rushed loan.
I’ll be honest about why this one lands for me. At VisionPros, the online contact lens retailer I co-founded, we were growing fast and looked healthy from every angle, and the lesson on cash conversion hit me later than it should have. Every extra dollar of monthly revenue needed more cash up front for inventory and marketing, and I didn’t know that ratio for our business until much later.
Bookkeeping and decision metrics are two different jobs: bookkeeping makes the past complete and accurate for taxes, lenders and buyers, while decision metrics turn a handful of those numbers into a call about spend, inventory or price this week. In my experience, stores that get into cash trouble often have one of these jobs covered and assume it covers the other.
Stefan Dobrev, Sales and Onboarding Lead at Link My Books, put the bookkeeping side plainly on episode 496: “You know, the operational side is so heavy and so intense that accounting becomes a bit of an afterthought. So it’s easy to look at how many sales have I done this month and then be confused about what you’re seeing in your bank account.” His point is that the merchant and the bookkeeper each see half the picture. The bookkeeper doesn’t know why inventory moved, and the merchant doesn’t know why the deposit is smaller than the sales report.
Adam Callinan, founder and CEO of Pentane, described the decision side failing on episode 449: “But you can be running an eight return on ad spend. Those dashboards are like green, you’re crushing. You couldn’t be doing anything better and still be losing $100,000 a month.” He said that happened with Pentane’s very first client. Clean books would show the loss, but they wouldn’t tell the operator which mix of fixed costs, ad spend and margin to change. In that case the answer was counterintuitive: Adam said the brand was underspending on ads and carrying heavy fixed costs, so it tripled ad spend, let ROAS fall from eight to about two and a half, and was profitable within three months. A decision metric found that. The books alone didn’t.
So you need both, and they’re owned differently. Bookkeeping is a process you can buy: software, a connector, a bookkeeper. If you’re choosing the software, the ecommerce accounting software roundup compares the main platforms and add-ons by stage, so I won’t repeat it here. Decision metrics are a habit the operator stays accountable for, even when someone else calculates them: picking the one number that matters at your stage, reviewing it on a rhythm and letting it change what you do. Here’s how that splits across the four stages. Treat the revenue bands as rules of thumb, not accounting thresholds: a $40K a month brand paying factory deposits months ahead may need a cash forecast before a $150K brand with easy supplier terms does.
The table is deliberately thin. One bookkeeping job and one decision number per stage is the point. Adding a second number before the first one changes a decision is how a finance stack gets complicated without getting useful.
At the Starting stage, from pre-launch to about $10K a month, the finance job is separation and recorded costs: one business account that only the store uses, a cost per item on every product in Shopify, and a monthly look at gross margin by product. That’s it, and I’d resist anything more.
The cost per item matters more than it looks. Shopify’s profit reports only count products that had a cost recorded at the time they sold, so a product you add a cost to in March won’t show profit for its January sales. Enter the cost the day you list the product. Once it’s there, the Gross profit by product report shows which products actually earn money after discounts and refunds, which is the first decision number a store needs: what to restock, what to reprice and what to stop selling.
Separation matters for a different reason. A Shopify payout is a net deposit with fees and refunds already taken out, and if it lands in a personal account mixed with groceries, you can’t rebuild your margin later without hours of untangling. If you want the account structure laid out, this breakdown of separating payouts, sales tax and idle cash covers operating, tax reserve and inventory accounts in detail.
A spreadsheet is fine here. We ran VisionPros on a simple Google Sheet for its first six months, and it stopped working once we got to about $30K a month, which was our cue to move to real bookkeeping software. What you should skip at this stage: paid forecasting tools, a fractional CFO and any dashboard app that charges monthly to show you numbers Shopify already reports. In most cases, none of them will change a decision you’re making at $5K a month.
At the Growing stage, from $10K to $100K a month, reconcile every payout to its orders, fees and refunds, close the books monthly, and track contribution margin per order, because a payout is a net number that hides the costs that decide whether your growth pays for itself. This is often the stage where a deposit-based spreadsheet starts producing wrong answers, not just slow ones.
Reconciliation is the bookkeeping job. A connector such as A2X or Link My Books splits each payout into sales, fees, refunds and adjustments and posts them to your accounting system, so revenue is recorded gross and fees are recorded as fees. Stefan made a point on episode 496 that catches a lot of stores at month end: revenue is recognized when you’ve fulfilled your obligation to the customer, which for most stores is effectively the order or shipment date under your terms, while the payout arrives later. So a payout that lands on the 2nd can carry sales earned on the 30th. A good connector splits that settlement across both months, which helps the close but doesn’t replace your accountant’s call on when revenue is earned. A deposit-based spreadsheet books it all in the wrong month, and your month end P&L is off by however big your last weekend was.
Contribution margin per order is the decision number. I worked one out loud on episode 454 with Matt Raminick: a $100 product that costs $30 to make, $8 to ship, $3 in Shopify fees and $20 in ad spend leaves $39, a 39% contribution margin. Matt’s caution was that brands differ on which costs they include. I’d add two lines stores often leave out: a returns allowance and the discount you actually gave, not the list price. To turn this into an ad ceiling, work out contribution before ads first ($59 in that example), then subtract what each order needs to leave for overhead and profit; what remains is the most you can spend to win that order. Be clear whether that limit is per order or per new customer, since repeat orders cost less to win. That logic is why profit on ad spend ranks channels better than ROAS does at this stage.
If you’re near the top of this band, set a close date and keep it, say the 10th of each month. A P&L you see in week six of the next quarter is history, not a decision tool.
At the Scaling stage, from $100K to $1M a month, run the business on a rolling 13 week cash forecast and know your cash conversion cycle, because inventory timing can become the biggest swing in your bank balance. The forecast sits on top of reconciled books and contribution margin; it doesn’t replace them. This is the stage where the record month problem from the top of this guide usually starts to bite.
The forecast itself is a mechanical job, and this guide to building a 13 week cash flow cadence covers starting from reconciled bank balances, dating every receipt and payment, and holding a 30 minute weekly review. The decision number it produces is simple: the lowest projected cash balance in the next 13 weeks, and the week it happens. Every purchase order, ad budget increase and hire should be tested against that one number before it’s approved.
Your cash conversion cycle tells you why that low point is where it is. It’s days of inventory, plus days to collect from customers, minus days you take to pay suppliers. Illustrative example, with my assumptions: a store at $300,000 a month with product cost at 35% spends about $3,500 a day on cost of goods, using a 30 day month. If it holds 90 days of inventory, waits about 5 days for payouts and pays its factory when the stock is received, with no earlier production deposit and no payment terms, the cycle is roughly 95 days. Cut inventory from 90 days to 60 with smaller, more frequent orders, with sales and unit costs unchanged, and the cash tied up in inventory falls by about $105,000 as current stock sells through and replacement orders shrink. That’s working capital you didn’t have to raise. A deposit paid months before the goods arrive makes the cycle longer than this simple version shows.
Izzy’s model is one way to get there for brands it fits. On episode 479 he described keeping stock at facilities near the factories in China and Vietnam and flying individual orders out, so a brand buys four to six weeks of inventory instead of four to six months. He was clear on the fit: brands that manufacture in that region, with orders averaging under 3.5 pounds, at roughly $1 million a year in sales or more. If that’s not you, the same logic still applies through supplier terms, order size and reorder timing.
One more call at this stage. Financing fixes timing gaps, not margin gaps. Revenue based financing or a line of credit can carry a peak season PO, but if contribution margin per order is thin and repeat purchases don’t make up for it, borrowing just lets you lose money faster. Commitments that land beyond 13 weeks, like a factory deposit for next season, need a longer planning view than the rolling forecast gives you. My bias is to stay self funded as long as the business can grow on its own cash, and to borrow only for a dated gap the forecast can show you’ll repay.
At the Leading stage, $1M a month and up, pay for judgment: a controller or fractional CFO who owns the close and the forecast, and one agreed definition of each metric that finance, marketing and operations all use. At this size the tools are usually not the gap. The gap is that three teams report three versions of “profit”.
Adam’s story about Pentane’s first client shows the problem this stage has to solve, and it only gets more expensive as the business grows. An eight ROAS on the ad dashboard and a $100,000 monthly loss in the books were both true at the same time. His point on episode 449 was that context lives in finance, in how the business actually spends money, and an ad platform can’t see most of it. At this size, the cash conversion cycle should be tracked by product line and by channel, because a wholesale account on net 60 terms and a DTC line that settles in three days are different businesses sharing one bank account.
Two boundaries are worth naming here, because they get lumped into finance and they’re separate jobs. Automating the back office with AI, from reconciliation to reporting, is an operations project with its own risks and belongs in an AI operations plan. Warehousing, 3PL contracts and shipping cost are fulfillment decisions. Finance should see the numbers from both, but it shouldn’t be where those decisions get made.
Before adding another finance tool, remove the dashboards nobody acts on, any second source of truth for revenue, and every metric that doesn’t have an owner and a decision it changes. In my years as a Merchant Success Manager at Shopify, the stores I worried about at the $500K to $2M stage were rarely missing a tool. They usually had too many, and nobody could say which one was right.
Here’s the subtraction test I’d run on any finance stack. For each tool and report, answer three questions. What decision did this change in the last 90 days? Who owns that number? Could I export my data and leave? That last one comes from the story of Seller Ledger’s founder, whose first bookkeeping product was shut down by its acquirer, and it’s the cheapest insurance in finance software. If a discretionary dashboard or report fails all three, cancel it this month, after you’ve exported its history. Your accounting system and anything you keep for tax or lender records stays; the test is for the extras layered on top.
How the call changes by stage. If you’re Starting, the only thing to add is cost per item, and the thing to remove is any paid app. If you’re Growing, add a payout connector and remove the deposit-based spreadsheet. If you’re Scaling, add the weekly cash meeting and remove any revenue dashboard that disagrees with your reconciled books. If you’re Leading, add a person who owns the definitions and remove the reports that exist only because each team built its own.
A note on relationships, so you can weigh this guide. eCommerce Fastlane earns affiliate commissions from Shopify, and some tool names on this site, including Shopify, Pentane, Portless and Seller Ledger in this guide, link automatically through referral links that can pay us a commission. The guests quoted here (Izzy Rosenzweig of Portless, Stefan Dobrev of Link My Books and Adam Callinan of Pentane) appeared on paid episodes of the podcast, and Seller Ledger currently sponsors the podcast. None of them reviewed or approved this guide, and no tool named here paid to be included.
Ecommerce accounting is recording and interpreting an online store’s money: sales by channel, platform fees, refunds, product costs, inventory, sales tax and the timing of payouts. It differs from general small business accounting because revenue arrives as net payouts days after the sale, with fees and refunds already deducted, and because inventory ties up cash long before it shows up as cost of goods. In practice it covers two jobs. Bookkeeping keeps the records complete and correct for taxes, lenders and buyers. Decision metrics, such as contribution margin per order or the lowest cash balance in a 13 week forecast, turn those records into weekly calls about spend, pricing and inventory.
Bookkeeping records every transaction accurately, while accounting interprets those records for taxes, reporting and decisions. For a Shopify store, bookkeeping means splitting each payout into sales, fees, refunds and adjustments, recording product costs and reconciling to the bank every month. Accounting builds on that: preparing statements, handling sales tax and income tax, valuing inventory and advising on structure. Neither one automatically produces decision metrics. A store can have clean books and an accountant and still not know its contribution margin per order or its lowest cash week. Someone inside the business has to own those numbers and review them on a set rhythm.
Your Shopify payout doesn’t match your sales because a payout is a net amount for a settlement period, not a copy of the sales report. Payment processing fees, refunds, chargebacks and any reserve are deducted before the deposit, and sales near the end of a day, a weekend or a month may settle into a later payout. Shopify Payments lists a minimum settlement time of three business days in the US and Canada, plus one to three days for the bank to post it (as of November 2026). To reconcile, use the Shopify Payments activity report or a connector that splits each payout into its parts.
A store can be profitable with no cash because profit counts product cost when units sell, while the bank pays for inventory under your supplier terms, often well before it sells. In a growth month you reorder more than you sold, often months ahead, so cash leaves faster than profit arrives. Payout settlement delays, wholesale customers on payment terms, loan principal and annual prepaid bills widen the gap. The fix is not usually better books. It’s a weekly cash forecast that shows the lowest balance in the next 13 weeks, plus smaller and more frequent purchase orders so less cash sits in inventory at any one time.
Hire a bookkeeper when reconciling payouts and closing the month takes more than a few hours, or when you sell on more than one channel; for many stores that happens somewhere between $10K and $100K a month. Work with an accountant from the start for tax filings, entity questions and sales tax, even if only once a year. Once inventory, channels and financing make the numbers hard to hold in one head, often around $1M a month, a controller or fractional CFO who owns the monthly close and the cash forecast tends to earn their cost. Whoever you hire, the decision metrics stay with the operator, because a bookkeeper records the past and can’t decide your next purchase order.
