You walk into the stockroom on a Tuesday. Three boxes of your top seller arrived yesterday — but you already had six on the shelf, and you sold maybe two units last week. Meanwhile the item three customers asked for on Saturday is still backordered. That's the inventory planning problem in one paragraph: too much of what's slow, not enough of what moves.
Inventory planning isn't a software feature you turn on. It's a decision loop you run every week or month: look at what sold, predict what will sell, decide what to buy, and check whether last cycle's bets paid off. This guide walks through the loop with real numbers, including the safety-stock formula most SMBs avoid because it looks scary. It isn't.
Last updated: 2026-05-13
Inventory planning is the set of decisions that determine which SKUs you stock, how many units of each, and when you reorder. It sits between forecasting (what you think you'll sell) and execution (cutting POs, receiving stock). Skip it and you'll either tie up cash in stuff that won't move, or stock out on the items that pay your rent.
A working plan answers four questions for every SKU you carry:
Notice what's not on that list: gut feel, "we always get 100 of these," or "the supplier said we should bump it." Those are inputs to the plan, not the plan itself.
The mistake most small businesses make is conflating planning with purchasing. Purchasing is the act of cutting a PO. Planning is the policy that tells purchasing what to cut and when. Without the policy, every PO becomes a one-off judgment call — and judgment calls don't scale past about 50 SKUs before you start making expensive ones.
Treat planning as a repeating cycle, not a one-time setup. Four steps, in order:
1. Analyze. Pull last 90 days of sales by SKU. Compute units sold, revenue, gross margin, and current on-hand. Sort by revenue and look at your top 20% — that's where most of your planning attention should go. While you're there, glance at inventory turnover for each SKU. Anything turning less than twice a year is a candidate to discontinue or markdown.
2. Forecast. Predict demand for the next planning horizon (usually 4–12 weeks). For most SMBs a 3-month moving average is good enough. If you have seasonality, layer in a seasonal index. The inventory forecasting guide walks through the methods.
3. Set policy. For each active SKU, pick a reorder point, a reorder quantity, and a safety stock level. This is where the math in section 4 comes in. The output is a single row per SKU: "When on-hand drops to X, order Y units."
4. Execute & review. Cut POs against the policy. Then — and this is the step everyone skips — review last cycle. Did anything stock out? Did anything sit too long? Adjust the policy, don't just adjust the next PO.
That last step is what separates a plan from a wish. If you never look back, you're not planning — you're improvising on a clipboard.
People use these terms interchangeably and they shouldn't. The distinction matters because the two have different inputs, different owners, and different cadences.
Demand planning asks: what will customers buy? It looks at sales history, promotions, seasonality, and external signals (weather, school calendar, a competitor closing). The output is a unit forecast per SKU per week or month. The owner is whoever knows the customer best — usually sales or the owner.
Inventory planning asks: given that forecast, what should we stock? It takes the forecast as a given and overlays supply-side reality: supplier lead times, minimum order quantities (see MOQ), case-pack sizes, cash on hand, and shelf space. The output is a stocking policy per SKU. The owner is whoever buys.
Here's why mixing them gets expensive. If your demand planner says "we'll sell 50/week" and your inventory planner says "the supplier's MOQ is 500 and lead time is 8 weeks" — those are both true facts, and they imply a totally different stocking strategy than the forecast alone suggests. Run them separately, then combine.
Safety stock is the buffer you carry to absorb the difference between forecast and reality. The standard formula is:
Safety stock = Z × σ × √L
Where Z = service level factor, σ = standard deviation of demand during one lead time period, and L = lead time in those same periods.
Let's work through it for a real SKU. Say you sell branded coffee mugs, your supplier ships from overseas, and you've pulled the last 12 weeks of sales:
| Week | Units sold |
|---|---|
| 1 | 22 |
| 2 | 18 |
| 3 | 25 |
| 4 | 30 |
| 5 | 20 |
| 6 | 24 |
| 7 | 19 |
| 8 | 28 |
| 9 | 21 |
| 10 | 26 |
| 11 | 23 |
| 12 | 24 |
Average weekly demand = 23.3 units. Standard deviation (σ) across those 12 weeks ≈ 3.6 units. Lead time from your supplier is 4 weeks (so L = 4).
For service level, pick a Z from the standard normal table. Common SMB choices:
Go with 95% for a normal SKU. Plug it in:
Safety stock = 1.65 × 3.6 × √4 = 1.65 × 3.6 × 2 ≈ 12 units
Your reorder point becomes: (average demand during lead time) + safety stock = (23.3 × 4) + 12 ≈ 105 units. Drop a reorder calc into your software, or use the reorder point calculator to skip the arithmetic.
What changes if your supplier becomes unreliable and lead time itself varies? You'd use a more advanced formula that includes σ of lead time too. But for most SMBs, the version above gets you 80% of the benefit with 20% of the headache.
Inventory planning tooling sits on a gradient. Match the tool to where you are, not where you wish you were.
Spreadsheet (Excel, Google Sheets). Works fine up to ~200 SKUs and one location. You'll need a sales export from your POS or store, a tab per SKU or one big table with formulas, and discipline to refresh weekly. Cheapest, most flexible, most fragile.
Light inventory SaaS (like StockZip). Worth it once you hit one of: more than ~200 active SKUs, more than one warehouse or location, more than one person touching stock, or barcode-driven receiving. You get a single source of truth, reorder alerts, and the Economic Order Quantity (EOQ) math built in. Monthly cost is usually less than one stockout.
Full ERP (NetSuite, Acumatica, SAP B1). Only when you've got multi-entity accounting, complex manufacturing, or 1,000+ SKUs with deep supplier integrations. ERPs are powerful and expensive — both in license fees and in the consultant time to set them up. Most SMBs don't need one and shouldn't pretend they do.
Skip these unless you have a specific reason: standalone forecasting tools, demand-sensing AI for sub-1,000-SKU catalogs, anything sold to you with the word "AI-powered" in the headline but no clear answer to "what does it actually compute?" Vendors love selling complexity. You don't have to buy it.
A short list, ordered by how often they bite:
Most of these are obvious in hindsight and easy to commit to in advance. The trick is the review step in section 3 — that's where you catch them before they cost you another quarter.
A: Most small businesses re-plan monthly at the SKU level and quarterly at the category level. If your lead times or sales swing hard, move to weekly for the top 20% of items by revenue.
A: Forecasting predicts what you'll sell. Planning decides what you'll buy and hold based on that forecast plus lead time, supplier minimums, cash, and shelf space. Forecasting is input. Planning is decision.
A: Whoever signs the purchase orders. In a 1–10 person shop that's usually the owner or ops manager. If you split buying and selling between two people, the buyer owns the plan and the seller feeds them sell-through numbers weekly.
A: Yes, up to roughly 200 active SKUs and one warehouse. Past that, the spreadsheet becomes the bottleneck — formulas break, version conflicts pile up, and nobody trusts the numbers. That's when light inventory software earns its keep.
Try StockZip free — Stop juggling spreadsheets and start running a real inventory plan.
No credit card needed.
Straight answers about spreadsheets, scanners, offline work, existing systems, and the free period.
Most small businesses re-plan monthly at the SKU level and quarterly at the category level. If your lead times or sales swing hard, move to weekly for the top 20% of items by revenue.