How to Forecast Ecommerce Cash Flow: a 7-Step Method

Most ecommerce founders forecast revenue and assume cash follows. It doesn't. The money from a sale lands 2 to 5 business days later on Shopify and several days after a settlement window on Amazon, while the inventory that produced that sale was paid for 60 to 220 days earlier. That gap is the whole problem. This post is the method: seven steps to build a forecast that models when cash moves, not just whether it does, plus a worked example you can copy.

Why ecommerce cash flow forecasting is different

A cash flow forecast for a DTC brand is not a smaller version of a corporate budget. The thing that breaks brands at $5M to $50M is not unprofitability. It is timing. You can be profitable on the P&L and still hit a wall when a purchase-order deposit, payroll, and a quarterly tax payment all land in the same light-receipts week.

Three numbers explain why. First, the median small business holds only 27 days of cash buffer (JPMorgan Chase Institute, Cash is King). The 25th percentile sits under 13 days. So most brands have under a month of margin for error. Second, the typical DTC cash conversion cycle runs 60 to 120 days (Wayflyer): the cash you spend on inventory takes two to four months to come back as collected revenue. Third, scaled DTC brands carry 41 to 221 days of inventory, computed from their latest 10-K filings. When your inventory days exceed your forecast horizon, the trough is structural. It is baked into how you buy and sell, not a one-off you can dodge.

When I talk to founders running a brand this size, the pattern is almost always the same: they are not unprofitable, they are profitable and out of cash in the same week. They forecast the top line, the top line is fine, and then a supplier deposit clears three days before payroll and nobody saw it coming. A cash flow forecast that models timing is how you see that collision 8 to 10 weeks early, which is early enough to move a PO, pull receipts forward, or draw a line.

The chart and table below show how much cash different brands have locked in inventory. The "13-week forecast window" row is the 91-day horizon you are forecasting against. Any brand whose inventory days exceed 91 is carrying a cash trough that lives outside a single forecast cycle.

The data table below shows the per-brand detail behind the chart.

Brand	Inventory days (DIO), FY2025	Vs 91-day window
FIGS	221	2.4x the window
YETI	133	1.5x the window
Lululemon	117	1.3x the window
13-week forecast window	91	baseline
Warby Parker	41	0.5x the window

If you want to compute your own number before you build the forecast, divide your ending inventory by annual COGS and multiply by 365 to get your DIO. Add your DSO (roughly 0 to 5 days for pure DTC) and subtract your DPO. The result tells you how many days of cash you have tied up, which sets the horizon your forecast actually has to cover.

The method: 7 steps to forecast ecommerce cash flow

Here is the spine. Each step produces one input the next step needs. Build them in order and you get a forecast that survives contact with the bank feed.

1. Pick your horizon and method. Run a 13-week rolling direct forecast for liquidity and a 12-month monthly indirect model for planning. Update the 13-week view weekly. The 13-week direct forecast is the one that keeps you solvent, because it is built around cash timing. This post is one level up: how you generate the numbers that go into it. For the full mechanics of building the 13-week grid itself, see how to build a 13-week cash flow forecast.
2. Pull your opening cash and reconcile it to the bank. A forecast that starts from an unreconciled balance is wrong on row one. Tie your opening number to the actual cleared balance before you do anything else.
3. Forecast revenue from drivers, not a trend. Sessions times conversion rate times AOV, per channel. Then layer returns percentage, processor fees, discounts, and channel mix on top. Driver-based revenue is what lets you run real scenarios, because you can move one lever (conversion drops 15%, returns climb to 12%) and watch the cash effect.
4. Convert revenue to cash receipts using payout timing. This is the step most founders skip. Shift each channel's revenue forward by its payout delay (Shopify captured-payment plus 2 to 5 business days, Amazon settlement plus 3 to 5 days), net of fees and refunds. Revenue is not a receipt until the money lands.
5. Schedule disbursements by actual due date. Inventory PO deposits and balances on supplier terms (30/70 deposit-balance, net 30/60), freight and duty near shipment, payroll, ad spend, rent, SaaS, tax, debt service. Each on the date it actually leaves the account.
6. Calculate net cash and ending balance each week, then set a minimum-cash threshold. Net receipts minus disbursements gives the weekly movement; the running ending balance is what you watch. The threshold is the line that turns a dip into an alarm.
7. Layer scenarios and run the variance loop. Build base, downside (revenue down 20 to 30%, slower payouts, higher returns), and upside. Every week, compare actual to forecast, update assumptions, and roll the horizon forward one week.

The table below maps which forecast does which job, so you are not trying to make one model do all of it.

Layer	Method	Horizon	Update cadence	What it's for
Weekly liquidity	Direct	13 weeks	Weekly	Payroll, supplier payments, near-term cash risk, runway
Revenue outlook	Driver-based	13 weeks to 12 months	Weekly / monthly	Model demand and collections from operational levers
Planning	Indirect (3-statement)	12 months	Monthly	Budgeting, seasonality, financing decisions
Board / lender	Indirect	12 months	Monthly	Explain profit-to-cash conversion and covenant headroom

Step deep-dive: getting payout and inventory timing right

Steps 4 and 5 are where forecasts live or die, so they deserve their own pass. The error is almost never in the totals. It is in assuming the cash moves the day the transaction does.

On the receipts side, your processor sits between the sale and your bank. Shopify Payments in the US pays out 2 to 5 business days after the payment is captured. Amazon disburses on a settlement cycle and then takes 3 to 5 business days for the transfer to clear, net of reserves and returns. Stripe and PayPal run 2 to 7 business days on standard payout. A Friday-night sales spike does not become cash until the middle of the next week, and a new or higher-risk account waits longer. If your forecast books revenue as a same-day receipt, every cash position you show is a few days too optimistic, and a few days is the whole margin when you hold 27 days of buffer.

On the disbursement side, inventory is the big one. You rarely pay for a PO all at once. A typical term is a 30% deposit when you place the order and the 70% balance at shipment or delivery, with freight and duty landing near shipment. The brands that get caught plan the cash-out around the PO date instead of the deposit and balance dates. When I talk to founders this size, the inventory deposit is the single line item they most often forget to forecast on its real date, and it is usually the biggest single outflow in the quarter.

Cash flow	Timing rule	Notes
Shopify Payments (US)	Captured payment + 2 to 5 business days	New / high-risk accounts longer; reserves possible
Amazon disbursement	Settlement cycle, then + 3 to 5 business days ACH	Reserves and returns netted out
Stripe / PayPal	2 to 7 business days standard	Instant payout available for a fee
Inventory PO deposit	On order (often 30%)	Plan the cash-out date, not the PO date
Inventory PO balance	On shipment / delivery (often 70%)	Add freight and duty near shipment
Returns / refunds	Roughly 1 to 2 weeks after sale	Model as a percentage of GMV

Worked example: a $20M DTC brand hits a Week-6 trough

Walk a concrete one. This is an illustrative model, not a measured brand, but the inputs are anchored to the real benchmarks above so the trough is believable.

A $20M apparel brand opens the quarter with $420K in the bank and sets a $250K minimum-cash threshold, the floor below which payroll and supplier payments get risky. Revenue is built from drivers and runs steady. But in Week 4 the brand places its pre-peak inventory order: a 30% deposit on a large PO clears, freight follows, and a quarterly tax payment lands the same fortnight. Payroll runs as normal. Receipts, meanwhile, are lagged 2 to 5 days behind sales, so the cash coming in does not cushion the cash going out in the same week.

Here is what the ending-cash line does, week by week:

The data table below shows the per-week detail behind the chart.

Week	Ending cash ($000s)	Threshold ($000s)	Status
Week 1	420	250	OK
Week 2	438	250	OK
Week 3	402	250	OK
Week 4	356	250	OK
Week 5	243	250	Below threshold
Week 6	165	250	Trough
Week 7	228	250	Below threshold
Week 8	301	250	Recovered
Week 9	347	250	OK
Week 10	392	250	OK
Week 11	358	250	OK
Week 12	404	250	OK
Week 13	451	250	OK

The forecast surfaces the breach in Week 1, when it is built, even though the cash does not dip until Week 5 and bottoms at $165K in Week 6. That five-week head start is the entire value of the exercise. With it, the founder has cheap options: move the PO deposit out by two weeks, ask the supplier to split the deposit, pull a promotion forward to land receipts sooner, or pre-arrange a draw on the line. Without the forecast, the same brand discovers the hole the week payroll is due and the only option left is the expensive one.

That draw, by the way, is real money again. The US bank prime rate is 6.75% as of May 2026 (FRED, MPRIME), down from 8.50% at the end of 2023 but no longer free. A typical asset-based line prices at prime plus 1 to 3%, so roughly 8 to 10% on a draw. The point of seeing the trough early is that you arrange the facility before you are desperate, when the terms are better and the bank is not reacting to a cash emergency.

Scenarios, the variance loop, and how often to update

A single forecast is a guess. The value comes from running it as three and closing the loop weekly.

Build a base case from your driver assumptions. Build a downside (revenue down 20 to 30%, payouts a couple of days slower, returns a point or two higher) and an upside. Because you built revenue from drivers in Step 3, each scenario is a real model, not a haircut on a single number. The downside is the one that matters most: it tells you how deep the trough goes if Q4 disappoints, and whether your line is sized for it.

Then run the variance loop. Every week, drop the week that closed, pull the latest actuals from the bank feed, compare them to what you forecast, and update the assumptions that were wrong. Add a fresh Week 13 so the horizon always runs a full quarter forward. Update the 12-month indirect model monthly. This is where forecasting stops being a document you build once and becomes the instrument you steer with. When we've struggled with this, what worked was treating Monday's forecast refresh as non-negotiable as payroll: thirty minutes, every week, reconciling forecast to actual.

The brands that hit a wall are almost never unprofitable. They are profitable and out of cash in the same week, because they forecast revenue and assumed cash would follow. Cash never follows on the day you sold. Build the forecast around when the money actually lands and the trough stops being a surprise.

Tools and next steps

Three moves this week. First, compute your cash conversion cycle: DIO plus DSO minus DPO gives you the number of days your cash is locked up and sets the horizon your forecast has to reach. Second, lay out the 13-week grid using the seven steps in this post to generate driver-based receipts and timing-correct disbursements, then roll it weekly. Third, if you want a second set of eyes on the model before the next pre-peak inventory order, talk to a fractional CFO who has built these for brands your size. For a broader walkthrough of the forecasting workflow, see how to do an ecommerce cash flow forecast.

Sources and methodology

Inventory days (DIO). Inventory days were computed from FY2025 Form 10-K filings on SEC EDGAR, using DIO = ending inventory / annual COGS x 365. FIGS, Inc. (CIK 1846576) reported inventory of $127.966M and COGS of $211.259M (accession 0001628280-26-012333, filed 2026-02-26), giving 221 days. Warby Parker Inc. (CIK 1504776) reported inventory of $44.512M and COGS of $401.326M (accession 0001504776-26-000006, filed 2026-02-26), giving 41 days. The calculation uses period-end inventory rather than a two-point average, a reasonable simplification that can slightly overstate or understate the figure depending on seasonality. The YETI (133) and Lululemon (117) figures are carried from prior Eightx analysis of their respective filings and are not recomputed in this run.

Cash buffer and cash conversion cycle. The 27-day median small-business cash buffer, the under-13-day 25th percentile, and the over-62-day 75th percentile come from the JPMorgan Chase Institute report Cash is King: Flows, Balances, and Buffer Days (Finding Three, Figure 5). The 60 to 120 day typical DTC cash conversion cycle, and the 30 to 90 day range for Amazon marketplace sellers, come from Wayflyer's cash-conversion-cycle benchmarks. CCC equals DIO plus DSO minus DPO; for pure DTC, DSO is roughly 0 to 5 days because customers pay at checkout, so the cycle is almost entirely inventory days minus supplier terms. Category benchmark ranges quoted elsewhere (apparel, electronics, CPG) come from vendor estimates rather than primary filings and should be treated as directional.

Payout timing. Shopify Payments US payout timing (2 to 5 business days after capture) comes from the Shopify Help Center US payouts page. Amazon disbursement timing (settlement cycle plus 3 to 5 business days for the ACH transfer) comes from Amazon Pay help article 201212310. Stripe and PayPal standard payout timing reflects published processor SLAs. New or higher-risk accounts and rolling reserves extend these windows.

Financing cost. The 6.75% US bank prime rate is the latest observation (May 2026) of the FRED Bank Prime Loan Rate series (MPRIME), down from 8.50% at the end of 2023 and 7.50% in early 2025. The FRED daily series (DPRIME) corroborated 6.75% as of mid-June 2026. The prime-plus-1-to-3% pricing for asset-based lines is a typical market range, not a quoted rate.

Method best-practice. The direct-versus-indirect framing, the rolling 13-week recommendation, and the driver-based revenue approach were synthesized from Wall Street Prep's 13-week cash flow model guidance, Harney Partners / TMA turnaround practice, Ripple Treasury, and Atlar. The $20M-brand trough model is illustrative and clearly labelled as such; it is anchored to the real buffer, CCC, and payout-timing benchmarks above so the trough is realistic, but it is not a measured brand.

Frequently asked questions

how do you calculate projected cash flow for an ecommerce business?

Start from a reconciled opening bank balance. Forecast revenue from drivers (sessions times conversion rate times AOV per channel), convert that revenue to cash receipts by shifting it forward by each channel's payout delay, then schedule every disbursement on its real due date (inventory deposits, payroll, ads, rent, tax, debt service). Net cash in minus cash out, week by week, gives you the ending balance. The trick is timing, not totals.

when does cash actually hit the bank after a shopify or amazon sale?

On Shopify Payments in the US, payouts land 2 to 5 business days after the payment is captured. On Amazon, the platform disburses on a settlement cycle and then takes 3 to 5 business days for the ACH transfer to clear, net of reserves and returns. New or higher-risk accounts wait longer and can face rolling reserves. Forecast the date the money lands, not the date of the sale.

what's the difference between actual and forecast cash flow?

Forecast cash flow is your projection of receipts and disbursements by week or month before they happen. Actual cash flow is what really moved, pulled from the bank feed after the fact. The gap between the two is your variance, and closing that loop every week (comparing forecast to actual, updating assumptions, rolling forward) is what makes the next forecast more accurate.

how far out should an ecommerce brand forecast cash flow?

Run two layers. A rolling 13-week direct forecast for liquidity (updated weekly) catches near-term cash risk: payroll colliding with a PO deposit and a tax payment. A 12-month monthly indirect model handles planning, seasonality, and financing decisions. The 13-week view is the one that keeps you solvent; the 12-month view is the one that helps you plan.

how do you account for inventory timing in an ecommerce cash flow forecast?

Schedule inventory by its cash-out date, not its PO date. Most suppliers want a deposit (often 30%) when you place the order and the balance (often 70%) at shipment or delivery, with freight and duty landing near shipment. Because scaled DTC brands carry 41 to 221 days of inventory, the cash leaves months before the sale that recovers it shows up, so model the deposit and balance as two separate disbursements on their actual dates.

what's the difference between the direct and indirect method for forecasting cash flow?

The direct method lists actual cash in and out by line item and week. It is the right tool for the 0 to 13 week horizon because it is built around timing. The indirect method starts from net income inside a 3-statement model and adjusts for non-cash items and working-capital changes. It is the right tool for the 3 to 24 month horizon because it ties cash to profit for planning and lenders.

what numbers do i need before i can build a cash flow forecast?

A reconciled opening bank balance, your traffic and conversion data per channel (sessions, conversion rate, AOV), your return rate, your payment-processor payout timing, your open purchase orders with deposit and balance terms, your payroll and rent and SaaS calendar, and your tax and debt-service due dates. If you have those, you can build the forecast. If your opening balance is not reconciled to the bank, fix that first.

how often should i update my cash flow forecast?

Update the 13-week direct forecast weekly: drop the week that closed, add a new week 13, and reset every line to the latest actuals. Update the 12-month indirect model monthly. The weekly cadence is what turns a forecast from a document into a steering wheel.

how do i forecast cash flow for a seasonal ecommerce brand?

Drive the revenue forecast off last year's weekly seasonality curve, not a flat line, then scale it by your current growth rate. The cash risk in a seasonal brand is almost always the pre-peak inventory build: you pay for Q4 stock in Q2 and Q3 and collect the cash in Q4 and Q1. Model that deposit-and-balance outflow on its real dates and you will see the pre-peak trough months before it lands.