Cash Flow Forecasting: The Model Most eCommerce Operators Skip
By Diosh — Founder, AHAeCommerce | eCommerce decision intelligence for $50K–$5M GMV operators
A profitable eCommerce business with 28% net margin can run out of cash in 90 days. The mechanism is straightforward: a $300,000 inventory PO commits in February for goods landing in May to drive a Q3 sales cycle. Payment terms with the supplier require 30% deposit on order ($90,000 in February), 70% on shipment ($210,000 in April). The goods sell over six months, generating revenue that lands in the bank account 2–7 days after each sale net of payment processing fees, with a 3% chargeback/return reserve held by the processor. The operator is profitable on every unit sold and broke from April to August because the cash timing — not the margin — is the binding constraint. Understanding contribution margin per unit is what confirms whether the underlying unit economics justify the inventory investment at all.
According to JP Morgan's 2024 Small Business Cash Flow report, the median small business holds 27 cash buffer days. The median seasonal eCommerce business needs 90–120 days. The gap is the reason the most common cause of eCommerce business failure is not unprofitability — it is cash timing, on a profitable operating model. This is preventable, but not by managing the bank balance.
Why the Bank Balance Lies
The bank balance is what your operating account holds today. It is a true number, but it answers the wrong question. The question that matters is: in 30, 60, 90 days, given the inventory commitments I have already signed, the marketing spend I have already committed, the customer payments that will arrive on a delay, and the operational costs that recur on fixed cycles — what will my balance be, and where are the troughs?
Operators who manage by bank balance are surprised by predictable events. The Q4 inventory PO that lands in October but does not generate revenue until December. The payroll cycle in the second week that always lands during the marketing-spend peak. The Black Friday revenue surge that creates a 5–7 day gap before the cash actually arrives, during which the BFCM ad spend has already cleared the account.
These are not surprises. They are visible 90 days in advance to any operator who has built a cash flow model. They become surprises only because the model has not been built.
The 13-Week Cash Flow Model
The standard cadence for eCommerce cash flow modeling is 13 weeks — one quarter at weekly granularity. The model captures every cash event in and out of the business, with timing accuracy at the week level. It is not a P&L, and it is not a budget. It is a timing model.
Cash Inflows (4 categories)
| Category | Timing | |---|---| | Customer payments via Shopify Payments / Stripe | 2-day rolling deposit (US), 3-day rolling (most international) | | Customer payments via Klarna, Afterpay, Affirm | Net 1-3 days, but with reserve held until return window closes | | Wholesale / B2B payments | Net 30–60 from invoice | | Other (refunds from suppliers, tax recoveries, factoring) | Variable |
The trap on inflows is double-counting. Revenue earned in a week is not revenue received. A $50,000 sales day on Shopify Payments generates approximately $48,400 in deposits over the following 2–3 days. The transaction fee ($1,150) and chargeback reserve ($450) are netted before deposit. New operators on Shopify Payments routinely have their first ~30 days of revenue partially withheld as reserve, which is not always communicated clearly during onboarding.
Cash Outflows (7 categories)
| Category | Timing | |---|---| | Inventory POs (deposits + balance on shipment) | Per supplier terms, typically 30/70 or 50/50 | | 3PL / fulfillment fees | Net 15 from invoice or auto-debit weekly | | Marketing spend (Meta, Google, TikTok) | Daily auto-debit, monthly invoice for some platforms | | Payroll | Bi-weekly or semi-monthly | | Software / SaaS / agency retainers | Monthly auto-debit | | Sales tax / VAT remittance | Monthly or quarterly per jurisdiction | | Card processing reserve releases / chargeback claw-backs | Variable — surface in real time |
The Critical Timing Variables
The 13-week model is most useful when it captures four variables explicitly:
1. Inventory float: The gap between cash committed for inventory and revenue earned from it. For a brand on 90-day production lead time + 30-day shipping + 60-day average sell-through, inventory float is ~180 days. Capital tied up in float at any moment is a function of inventory turn rate.
2. Payment processing reserve: Shopify Payments and Stripe both hold rolling reserves. Standard reserves are zero for established merchants but can be 5–25% for new accounts, high-chargeback categories, or rapid revenue growth.
3. Returns provision: For categories with 20%+ return rates, a non-trivial portion of revenue arriving in the bank account will be debited back within 30–60 days.
4. Marketing-to-revenue lag: Paid acquisition spend clears today. Revenue from that spend arrives across the customer's first purchase window — typically 0–14 days for direct-response Meta, 14–60 days for awareness/funnel campaigns. The gap is real cash exposure during the lag.
The Worked Example: A Q4 Cash Crunch That Was Visible in August
Operator profile: $2.2M GMV, beauty and skincare category, 32% contribution margin, 65% Q4-weighted revenue concentration. Modal supplier terms: 50% on order, 50% on shipment, 75-day production + transit lead time.
The August reality:
- $480,000 inventory PO committed for Q4 SKUs (50% deposit = $240,000 due August 15, 50% = $240,000 due October 30)
- August revenue running at $135,000/month (off-peak)
- Operating cash on hand August 1: $185,000
- Monthly fixed cost burn (rent, salaries, software, base agency): $42,000
- Marketing spend running at $28,000/month (off-peak)
The visible-in-August trajectory:
| Week | Beginning Cash | Inflows | Outflows | Ending Cash | |---|---|---|---|---| | Aug W2 (PO 1 deposit) | $185,000 | $32,000 | $310,000 | -$93,000 | | Aug W3 | -$93,000 | $33,000 | $25,000 | -$85,000 | | Aug W4 | -$85,000 | $33,000 | $24,000 | -$76,000 | | Sep W1 | -$76,000 | $35,000 | $26,000 | -$67,000 | | Sep W2 | -$67,000 | $36,000 | $24,000 | -$55,000 |
By August 15, the operator is in a $93K cash hole that does not close for 8 weeks. Net 30 supplier terms could close part of the gap. A line of credit set up in May would close it entirely. Without either, the operator misses payroll in late August — on a business that will produce $2.2M in revenue and ~$700K in contribution margin for the year.
This was visible in May. The operator did not have a 13-week cash model. The bank balance in May ($310,000) suggested everything was fine.
The corrective actions, sequenced by leverage:
- Renegotiate PO terms: 30/70 instead of 50/50 reduces August deposit from $240K to $144K. Cash impact: +$96K in August. Most suppliers will agree at 12+ month relationships and orders above $100K.
- Stage the PO: Two POs of $240K each, 6 weeks apart, instead of one $480K PO. Cash impact: spreads the deposit hit. Operational cost: 2x receiving labor, modest carrying cost, possible higher unit cost on smaller runs.
- Establish a working capital line: $200K–$400K line of credit at 8–12% interest. Capital cost on $200K drawn for 90 days = $4,000–$6,000. Insurance against the cash trough.
- Invoice factoring on B2B receivables (if applicable): For operators with wholesale, advance 80–90% of receivables for ~2% of invoice value. Specific to wholesale-mixed operators.
The point is not the specific tactic. The point is that all four tactics require lead time — typically 30–90 days — and none are available the week the cash hole materializes. The 13-week model is what surfaces the problem with enough runway to address it.
The Inventory Cash Lock-Up Most Operators Underestimate
For most physical-product operators, inventory is the single largest cash use. The model that matters is not "how much inventory is on hand" — it is "how many days of cash are tied up in inventory at any given time."
The formula:
Inventory cash days = (Average inventory value × 365) / Annual COGS
For a $1.5M GMV apparel operator with 35% COGS ($525K) and average inventory of $185K, inventory cash days = (185,000 × 365) / 525,000 = 129 days.
That operator has 129 days of cash perpetually locked in inventory. To free up working capital, the levers are: faster inventory turn (smaller, more frequent POs — at the cost of higher per-unit cost), shorter supplier lead times (often costs 5–15% premium for domestic or expedited production), or extended payment terms (Net 30/60 from suppliers, achievable at scale).
A 30-day reduction in inventory cash days frees up $43,000 in working capital for this operator. Most operators in this range have 110–160 inventory cash days when they audit. The first-time audit usually surfaces $30K–$80K in trapped cash.
The Reserve Gotcha That Only Hits Once
Shopify Payments, Stripe, and most processors hold a rolling reserve on new merchants and on accounts with rapid growth or high return rates. Per Shopify's published merchant agreement, reserves can be set at the processor's discretion based on chargeback risk profile and growth velocity.
The pattern: an operator scales from $50K/month to $200K/month over a quarter, hits a Shopify Payments reserve evaluation, and finds 5–10% of incoming revenue held in reserve for 30–90 days. On $200K/month at 10% reserve, that is $20,000/month in unavailable cash for the duration of the reserve.
This happens once to most operators. It is communicated in advance only sometimes. The defense is to model rapid growth periods with a 5–10% reserve provision built in, even if no reserve is currently in place. The provision rarely materializes; when it does, the cash plan absorbs it.
The Decision Framework
| Cash Position | Action | |---|---| | 13-week forecast shows positive balance throughout, lowest week >30 days operating burn | Healthy. Maintain quarterly model refresh. | | Lowest week 15–30 days operating burn | Watch list. Renegotiate PO timing or establish credit line. | | Lowest week <15 days operating burn or any negative week | Active planning. Sequence corrective actions immediately. | | Forecast shows negative position within 8 weeks | Crisis lead time. Credit line or factoring must close in 4 weeks. | | Forecast shows negative position within 4 weeks | Emergency. Immediate inventory deferral, marketing cuts, supplier renegotiation. |
The forecast is updated weekly with actuals. Each week, last week's projected cash position is compared to actual. If variance is >5%, the model assumptions are wrong and need rebuilding — not the model abandoned.
What to Build, What to Buy
Most operators between $200K and $3M GMV can build the 13-week model in a spreadsheet (Google Sheets, Excel) and update it weekly in 30–45 minutes. Above $3M GMV, the volume of transactions and the complexity of inventory commitments usually justify a tool — Pulse, Settle, Float, or accounting platforms with cash flow modules (Xero with Float plugin, QuickBooks with similar). Per Settle's 2024 customer benchmarks, brands using inventory financing tools maintain 35% lower inventory cash days than peers without them, primarily through better timing of supplier payments.
The choice between spreadsheet and tool is not financial — both are inexpensive. It is operational. A spreadsheet requires the operator to update it; the tool requires integration setup and a recurring fee. Below $3M GMV the spreadsheet usually wins on adoption. Above, the tool wins on accuracy and time.
The Verdict
Cash flow forecasting is not a finance department exercise. It is the operating discipline that separates eCommerce businesses that scale from eCommerce businesses that fail profitably. The 13-week model, updated weekly, surfaces every predictable cash event with enough lead time to act.
The model is not optional once seasonality, inventory commitments, or payroll exposure exists. It is what allows the operator to take on a $300K Q4 PO with confidence, to negotiate supplier terms with leverage rather than under duress, and to spot the difference between a profitable business and a profitable-but-cash-broken business — before the bank balance reveals which one it is.
This Week
Build the 13-week model. Pull last 30 days of bank transactions and categorize them into the inflow/outflow buckets above. Forward-project the next 13 weeks using known commitments (POs signed, supplier deposits due, payroll cycles, recurring software, average daily revenue). Identify the lowest projected balance in those 13 weeks. If that number is less than 30 days of operating burn, the corrective actions need to start this week — not the week the cash crunch arrives.
