Inventory

Inventory Planning for Shopify: Turn Demand into a Seasonal Buying Plan

What a seasonal buying plan must record

Inventory planning, for a seasonal Shopify buy, is one pre-season sheet: which SKUs are in the season, how many units are a base buy versus a stretch buy, what cash ceiling each line may consume, which dated purchase approvals and supplier deadlines authorize those buys, and which early-sales rule can later release or cancel the stretch.

A forecast answers how many units you expect to sell. This sheet answers a narrower question: of that estimate, what are you willing to commit on paper before any order is issued, and on which dates that commitment is allowed. If the sheet does not separate base from stretch, cap the cash, date the approvals, and name the rule that can change the stretch, it is a demand file.

This guide assumes the demand estimate already exists. How to build and check that estimate is in ecommerce demand forecasting. This sheet records the dated purchase approvals, the supplier deadlines those approvals must beat, and the calendar for acting on contingencies. Issuing the weekly purchase orders, reconciling incoming receipts, and editing an open order stay in inventory replenishment. Do not rebuild that weekly queue here.

Five fields belong on the sheet:

  • In, defer, or out for each seasonal SKU, with a reason tied to stock cover and forecast confidence.
  • Base units and stretch units, each with a cash amount, so the full estimate is not treated as authorized.
  • A five-date approval calendar: demand lock, base-buy approval, supplier deadline, contingency read, and latest stretch approval, plus the sell window those dates are judged against.
  • A contingency rule that compares the same number of days of sales with the same number of days of estimate, plus the cash cap that rule may release, and a supplier-delay rule for when arrival slips.
  • A cash reconciliation that labels committed cash, held-back stretch, returned cash, and any unassigned seasonal balance so the seasonal allocation adds up.

Skip formulas that other posts already own. Order quantity math lives in the economic order quantity formula. Buffer units live in the safety stock calculation. The trigger for a recurring reorder lives in the Shopify reorder point formula. Copy those outputs onto the sheet. Do not rebuild them in the planning meeting.

Build the sheet in this order. Each step below names the field or decision that step must record before you move on.

  1. Lock the season, the five approval dates, and the SKU list.
  2. Copy stock inputs onto the sheet.
  3. Split each estimate into a base buy and a stretch buy.
  4. Cap the sheet with a seasonal cash ceiling.
  5. Write contingency rules that use the same period, including a supplier delay.
  6. Score the plan after the sell window.

Step 1: Lock the season and the SKU list

Write the season before you look at quantities. A buying plan fails when before Black Friday is the only date, because nobody can tell which estimate the cash is tied to, or which approval still has time to reach the supplier.

The sheet records a five-date approval calendar. These dates decide which buys are authorized and when a contingency may still change them. They are not the weekly purchase-order queue. Placing the order still happens in inventory replenishment, against the dates and cash caps written here.

The five dates that govern each buy

  1. Demand lock date: the last day the forecast on this sheet can change. Buys approved after this date use this estimate only.
  2. Base-buy approval date: the day an owner signs the committed column, and the cash cap they may sign. A base unit is not authorized before this date. Nothing is added to the committed column after it unless the sheet records a written exception and a new cap.
  3. Supplier deadline: the supplier's last accept date that still lands the approved quantity before the sell window starts. If this deadline falls on or before the base-buy approval date, the row cannot be a base buy. Mark it deferred or out. A verbal note that the supplier will still take it is not an approval.
  4. Contingency read date: the day early sales are compared with the same number of days of estimate. Stretch cash is released or cancelled only on this date, up to the reserved cap, by the person named on the row.
  5. Latest stretch approval date: the last day a stretch buy can still be approved and arrive inside the sell window. After this date the stretch is cancelled even if sales are ahead, unless the supplier has confirmed in writing that the units will still land inside the window.

Also write the sell window start and end. Every approval date is judged against that window. A supplier deadline that cannot land units before the window starts does not get a purchase approval. A contingency read date that falls after the latest stretch approval date cannot release cash, because the buy would miss the window.

How approval timing and supplier deadlines govern the buy

Approval timing is the gap between the base-buy approval date and that row's supplier deadline. The gap is the working time to place the order after the signature. If the gap is zero or negative, move the approval earlier or drop the row from the committed column. Do not approve a buy you already know will miss the supplier deadline.

Supplier-deadline handling sits on the same row. If the supplier moves the deadline earlier, either pull the approval date forward so the gap stays positive, or move the quantity out of the committed column and return that cash to the unassigned seasonal balance. If the deadline moves later but the new arrival still misses the sell window, the same rule applies: no approval, and the cash returns to the unassigned balance. A long lead time on the row is only a note until it is written as this deadline.

Then list the SKUs that belong in the season. A SKU is in scope when its demand is expected to jump inside the sell window and a pre-season decision is required. Evergreen SKUs that keep their normal weekly reorder stay off the sheet, with their cash reserved so the season cannot spend it. Colorways stay on separate rows when they do not share the same estimate, the same cover, or the same supplier deadline.

Step 2: Copy stock inputs onto the sheet

Shopify does not show one inventory number. On the Inventory page and on the product inventory card, Shopify's inventory states split the quantity: Available is what you can sell, and it excludes Incoming. Committed is set aside and cannot be sold, including unfulfilled orders, draft-order reservations, and transfers marked ready to ship. On hand is every unit at the location, the sum of Committed, Unavailable, and Available. Incoming is stock on the way from transfers, purchase orders, or apps, and it is not sellable until it is received.

For each in-scope SKU, copy four figures onto the sheet before you touch the forecast:

  • Available units at the location that will sell the season.
  • Incoming units, with the expected arrival date already shown, not a lump sum with no date.
  • Any supplier quantity that is not yet reflected as Incoming, noted as an external figure so you do not treat it as zero.
  • Weeks of cover at recent sales velocity, labeled as current cover, not seasonal cover.

Subtracting only Available from the seasonal estimate ignores Incoming and can plan units that are already on the way. Planning against On hand overstates what you can sell, because On hand includes Committed and Unavailable units. Current cover is a starting fact, not the buy quantity. A SKU with 8 weeks of cover at July velocity can still be short in week two of a November peak if the seasonal estimate is several times summer demand. The opposite is common too: a SKU that looks thin in July is covered if leftover units plus Incoming already span the sell window.

If the counts on the sheet do not match a physical check, fix the record before the planning meeting. That diagnosis is in inventory discrepancy. Reconciling open supplier orders into the weekly buy queue stays in inventory replenishment. The weekly loop that keeps counts honest between seasons is in Shopify inventory management.

Step 3: Split each estimate into a base buy and a stretch buy

Import the demand estimate as a weekly or monthly unit number per SKU for the sell window, plus the confidence note that came with it. Do not reopen the forecast model in the planning meeting. The question is whether this estimate is firm enough to put cash against on the sheet.

Flag a SKU as uncertain when any of these are true, and treat the flag as a constraint on the sheet:

  • The SKU is new, or last year's comparable period had only a handful of weeks of real sales.
  • Last season's forecast error was large enough that buying the full estimate would have left a markdown problem or a stockout.
  • The estimate depends on an ad budget, a wholesale order, or a retail event that is not yet confirmed.
  • Lead time is longer than the period in which a second buy could still land inside the sell window, which means the latest stretch approval date falls before a useful contingency read.

How to measure that forecast error is in the forecast accuracy formula.

Split every seasonal estimate into a base buy and a stretch buy. The base buy is the quantity you would still want on the sheet if demand lands at the low end you are willing to own. The stretch buy is the extra units that only stay on the sheet as uncommitted until the contingency read date confirms the estimate, and only if that date is still on or before the latest stretch approval date. Uncertain SKUs should have a small base and a large stretch. Stable repeat SKUs can have most of the estimate in the base.

Grade the list before you assign cash. ABC analysis is the prior for which rows deserve a line-by-line decision. A-class seasonal items get their own base, stretch, approval date, and supplier deadline. C-class items can share one rule: base only, or out of the sheet if current cover plus Incoming already spans the sell window.

A smooth spreadsheet or AI forecast is not permission to skip the flag. If the prediction has not been checked against last season's misses, mark the SKU uncertain and leave the stretch uncommitted. When to question a prediction before cash is assigned is covered in AI demand forecasting.

Step 4: Cap the sheet with a seasonal cash ceiling

Cash is the constraint that turns a demand file into a plan. Start with the amount this season can spend without taking the cash that evergreen SKUs still need. That seasonal pool is the ceiling. The forecast is not.

Split the pool into three columns on the sheet, in money and in units:

  • Committed on the sheet: the base buy for SKUs whose current cover plus Incoming will not last to the sell-window start, whose supplier deadline is after the base-buy approval date, and whose lead time means a later decision would miss that window.
  • Held back: cash reserved for stretch buys that can still be approved on the contingency read date, before the latest stretch approval date.
  • Protected replenishment: cash the seasonal sheet will not touch, because core SKUs still need their normal reorders while the season runs.

Rank lines for the committed column with three tests, in order.

  1. Will Available plus confirmed Incoming cover demand until the sell window starts? If yes, the SKU does not need a pre-season base unless waiting would push the supplier deadline past any later approval date.
  2. Is the forecast flagged uncertain? If yes, the committed column gets only the base, and only if the base-buy approval date is before the supplier deadline.
  3. Does unit cost times the base fit in the committed column after higher-priority SKUs are funded? If not, cut the quantity or mark the row deferred, return the unfunded cash to the unassigned seasonal balance, and leave the deferral on the sheet.

Lead time only sets priority inside that committed column, and only after it has been written as a supplier deadline. A 70-day factory SKU that will miss November if it waits outranks a 14-day domestic SKU with the same revenue, because the domestic SKU can still be funded from held-back cash after the first sales weeks. Long lead time plus thin cover is the line you fund first, provided the approval date still beats the supplier deadline. Long lead time plus heavy cover is a candidate to leave off the sheet, even if the forecast looks large. Signing the approval authorizes the column. Issuing the order is the later replenishment step.

A worked example with hypothetical numbers

The figures below are an example, not a benchmark and not a result from a live store. They show how to fill the sheet, not a target to copy and not an order to send.

A store has a $40,000 seasonal pool for a November 1 to December 20 sell window. $12,000 is protected for evergreen replenishment. That leaves $28,000 for seasonal SKUs. Three SKUs are on the table:

SKUUnit costBase unitsStretch unitsCover now + incomingLead timeForecast flag
Gift set$184002503 weeks50 daysStable repeat
New candle$93005006 weeks21 daysNew, uncertain
Blanket$322008011 weeks55 daysStable, heavy cover

On this example sheet the demand lock is August 15 and the base-buy approval is August 20. The gift set supplier deadline is September 12, 50 days before November 1, so the August 20 approval still has a positive gap and the base can be signed. Its latest stretch approval is October 31, 50 days before December 20, which is before any in-window sales read. The candle supplier deadline for a pre-window arrival is October 11, also after August 20, so the base can be signed and the stretch waits. The blanket supplier deadline is September 7, but 11 weeks of cover already runs through most of the sell window, so no approval is signed.

Gift set base cash is 400 × $18 = $7,200. Cover will not reach November 1, the approval date beats the supplier deadline, and the lead time means the base has to sit in the committed column. A 50-day order approved after October 31 arrives after December 20, so October 31 is this SKU's latest stretch approval date. The November 15 checkpoint, 14 days after November 1, is after October 31, so it cannot release this stretch. Cancel the 250-unit stretch ($4,500) on the sheet before October 31 and move that cash to the unassigned seasonal balance. Do not hold it for the November 15 read.

New candle base cash is 300 × $9 = $2,700. It is uncertain, so the 500-unit stretch ($4,500) is not committed. Lead time is 21 days, so a contingency read on November 15 (14 days after November 1) is still before the latest stretch approval date of November 29 (21 days before December 20). The stretch can remain held back until that read.

Blanket base cash would be 200 × $32 = $6,400, but 11 weeks of cover already runs through most of the sell window. The sheet marks the blanket out, signs no approval, and labels that $6,400 as returned cash, not as held-back stretch. A later top-up is allowed only from held-back cash, only on or before the latest stretch approval date, and only if the supplier deadline still lands inside the window. Otherwise the written rule is to stop.

The $28,000 seasonal allocation reconciles in four labeled amounts. Committed cash is $7,200 + $2,700 = $9,900. Held-back stretch is the candle only, $4,500. The gift set stretch is cancelled, not held. Returned blanket cash, not assigned to a buy, is $6,400. Unassigned seasonal balance is $7,200: the $2,700 that was never tied to a SKU, plus the cancelled gift set stretch of $4,500. That $7,200 is not the gift set base and not the candle base. $9,900 + $4,500 + $6,400 + $7,200 = $28,000. Protected replenishment of $12,000 sits outside that $28,000. $12,000 + $28,000 = $40,000, the full seasonal pool. The sheet has not spent the forecast. It has assigned the portion that current cover, the approval date, and the supplier deadline actually require, and it has named the cash that is still unassigned.

Compare recent sales velocity and stock cover with the demand estimate on every line before you lock the row. A SKU whose recent weekly velocity, scaled only by a clearly stated seasonal factor, cannot support the base buy should be cut before the meeting ends, and the cut cash moves to the unassigned seasonal balance. Skymetrics stock intelligence is one way to put velocity and cover next to the SKU so that comparison is not a manual export. The base, the stretch, the approval dates, and the cash cap still have to be written on the sheet.

Step 5: Write contingency rules that use the same period

A contingency on this sheet is a rule two people would apply the same way. It is not a meeting to renegotiate the season, and it is not the purchase order itself. Each in-scope row needs the five approval dates, the metric, the contingency read date, the cash the rule may release or cancel, and the person allowed to make that call up to a stated cap.

Compare 14 days of sales with 14 days of estimate. The two-week estimate is the weekly demand estimate multiplied by 2. These rules are an example for the sheet above, not an industry standard:

  • If units sold in the first 14 days are below 60% of the two-week estimate, cancel the stretch and do not replace it with a different SKU on the same row. Move the cancelled cash to the unassigned seasonal balance.
  • If units sold in the first 14 days are above 120% of that same two-week estimate, held-back cash remains, and the contingency read date is on or before the latest stretch approval date, the sheet allows the stretch up to the reserved amount.
  • If you prefer a one-week checkpoint, compare the first 7 days of sales with the weekly estimate, using the same 60% and 120% cutoffs. Do not compare 14 days of sales with one week of demand. The one-week read still has to fall on or before the latest stretch approval date, or it cannot release cash.
  • If Incoming already on the sheet arrives above the quantity you recorded, reduce the stretch by that surplus before anyone treats the row as still open. If the surplus is larger than the stretch, reduce the unreceived base by the remainder and move that cash to the unassigned seasonal balance.
  • If the supplier confirms a delay that pushes arrival past the sell-window start, cancel the stretch first, including any stretch not yet released. Keep the unreceived base only if the new arrival date still covers the remaining sell window and a revised supplier deadline still sits after the approval date. If it does not, mark the unreceived base deferred and move that cash to the unassigned seasonal balance. Do not swap in a different SKU on the same row.

On this example, the November 15 read applies only to the candle. The gift set stretch is already cancelled, because its latest approval date is October 31 and a 50-day buy released on November 15 would miss December 20.

Set your own percentages from last season's forecast misses and from how late a second buy can still land inside the sell window. A store that can still receive a domestic reorder in 10 days can use a tighter checkpoint than a store whose second buy takes 8 weeks. The tighter checkpoint is still the same comparison: the same number of days of sales against the same number of days of estimate.

The supplier-delay rule governs the cash on this sheet. Chasing the missed receipt, editing the open order, and placing any replacement that replenishment later approves stay in that weekly queue. Hand that team the locked rows, the approval calendar, the supplier deadlines, the held-back caps, the unassigned balance, and the rules above.

Check the sheet before you hand it off

  • Every in-scope SKU is marked in, deferred, or out.
  • The five approval dates are filled. The base-buy approval date is before that row's supplier deadline, or the row is deferred or out.
  • Base units and stretch units are separate numbers. Uncertain forecasts are not fully committed.
  • Committed cash, held-back stretch, returned cash, and the unassigned seasonal balance add up to the seasonal allocation. Protected replenishment is separate and is not spent by this sheet.
  • Available and Incoming were copied the same week the sheet was locked.
  • Each row's contingency compares the same number of days of sales with the same number of days of estimate, names an owner and a cash cap, and states the one-week option, the incoming-surplus cut, and the supplier-delay cut. A stretch whose latest approval date is before that read is cancelled before the deadline, not held for it.

If a row fails, fix the sheet. Do not send a partial order to hold a slot from this meeting. A deposit with no cancel right is a commitment, and it has to be written in the committed column, dated on or after the base-buy approval date and before the supplier deadline, before it leaves the plan.

Step 6: Score the plan after the sell window

Once the sheet is handed off, do not run it as a second reorder system. Open orders, incoming receipts, and mid-season cuts follow the replenishment calendar, using the approval dates and contingency rules already written here.

After the sell window closes, keep three figures with the sheet so next season starts from a record:

  • Base units on the sheet versus units sold in the window, per SKU, and whether that base was approved before its supplier deadline.
  • Stretch units the rule allowed versus stretch units actually received before the window ended, including stretch cancelled by a supplier delay or an incoming surplus.
  • Cash left in held-back and in the unassigned seasonal balance at the close, and whether it was returned or spent on late adds that missed the latest stretch approval date.

Those three figures tell you whether the sheet was too timid, too late, or too loose with uncertain SKUs. They do not replace a forecast accuracy review. They tell you whether the buying rules and approval dates you wrote were the ones that got used.

A miss on one row does not mean you should raise every other SKU, and a beat does not mean the forecast method failed. Hand a pattern of misses back to the forecasting workflow after the season. When the next demand lock date arrives, start again from stock, Incoming, and a fresh estimate, with last season's sheet on the table.

Common mistakes that blur the plan into a buy queue

Most weak seasonal plans fail as documents, before anyone argues about a formula.

  • Putting the full forecast in the committed column because the spreadsheet looks precise. Precision is not a cash ceiling.
  • Leaving Incoming off the sheet and planning the gross demand number.
  • Using one date for every SKU. A 14-day vendor and a 60-day factory do not share the same supplier deadline or the same latest stretch approval date.
  • Signing a base-buy approval on or after the supplier deadline, or leaving the five dates blank.
  • Writing stretch units into the same authorized quantity as the base.
  • Ranking by revenue alone. A high-revenue SKU with 12 weeks of cover does not need early cash on this sheet.
  • Rebuilding the forecast in the planning meeting. That delays the approval date and produces a number nobody locked.
  • Turning this sheet into a weekly purchase-order queue. That job already has a home. The sheet's job is the dated approval and the rule that can still change it.

How to know the plan worked

You can check the plan before a seasonal unit sells. On the demand lock date, every in-scope SKU is in, deferred, or out. The five approval dates are filled, and no base buy is signed if its supplier deadline is already past the approval date. Base and stretch are separate. Held-back cash and the unassigned seasonal balance are still unspent on the sheet. The contingency compares 14 days with a two-week estimate, or 7 days with a weekly estimate, names who may release the cap, and states what an incoming surplus or a supplier delay does to that cap. A stretch whose read would fall after its latest approval date is cancelled before that deadline.

After the window, a working plan shows two patterns. Seasonal SKUs that were short on cover were either approved on the sheet in time for the supplier deadline, or the miss is explained by a date the replenishment calendar recorded after that approval. Uncertain SKUs did not consume the whole pool up front. If every new SKU was fully committed in August, the sheet was a forecast printout.

Build the next season from this sheet. Send every order through the replenishment calendar, against the caps, approval dates, and rules already written.