Guides
Working backward from your seasonal peak to the day you have to order
A reorder point waits for stock to fall. A dated peak doesn't wait for anything — so the only way to find the order date is to start at the peak and count backward.
A peak has a date. A reorder point doesn't.
Rolling replenishment rests on a quiet assumption: if you miss this week's trigger, there's another one next week. Cross the line late, order late, land late, lose a few days of sales. The cost is real but bounded, because the loop comes around again.
A dated peak breaks that assumption. The date is fixed by something outside your operation, and it doesn't slide because the PO went out two weeks after you meant it to — you get one commit window, and it doesn't reopen. Worse, the trigger fires after the window shuts: a reorder point only moves once demand has pulled stock through the threshold, and for a pre-season buy, that pull is the peak starting.
Assisty's forecast marks which items run out before the period ends, and at-risk items carry straight into the Smart Buying Table, so seeing a risk and fixing it are the same motion. That tells you a horizon has a hole in it. It doesn't decide which week of that horizon is your peak, or when the order covering it had to leave. The calendar is yours.
Count backward in legs, not in one lead time
Manufacturing solved this shape of problem long ago and called it backward scheduling: in MRP, planned start and release dates are calculated backward from the planned end date using planned lead times — "lead-time offsetting" (mrpeasy.com). Run your own chain in reverse and every arrow is a number of days, every junction a date.
The trap is treating that whole chain as one number. A supplier lead-time field — anywhere, not just here — is usually the production leg, sometimes production plus transit. In Assisty you set each supplier's Lead Time, Reorder Cycle, Min Stock Cover Days, and order minimums and maximums once, and the buying table's suggested quantities use them automatically. But whatever a lead-time field spans, it stops at your door. Arrival to sellable is its own leg, and it's the one nobody counts: stock in the building but not yet put away is not stock a customer can buy.
commit date = peak start − (sell-in + receiving + transit + production lead time + safety time)
Add the legs up honestly and an import chain like that runs past three months.
The legs that quietly grow before a peak
Even then the arithmetic under-delivers, always in the same direction, because every leg is longer in the run-up to a peak than in the quiet middle of the year.
Production first. insight-quality.com puts the official mainland holiday for Chinese New Year 2026 at 15–23 February and notes that most factories fully close for two to four weeks around it, then ramp back up gradually. A production leg quoted in ordinary weeks doesn't describe a week that straddles a shutdown.
Transit next. Sea-Intelligence reported global container schedule reliability of 62.6% in June 2026, with an average delay of 5.31 days for the vessels that arrived late. The quoted transit time is the common case — and roughly a third of the time it isn't the case at all, by days.
Then your own dock, which is slowest at exactly the moment the most containers land.
The instinct here is to add units. Wrong currency. Safety time is days you add to the schedule; safety stock is units you add to the order. A dated peak is a date problem, so buy the days — more units don't help if they arrive in week two of a six-week peak. Commit earlier than the bare arithmetic says, and put the difference on the calendar.
Suppose your peak starts the first week of November. Working backward: 5 days to get received stock onto the floor and selling, 4 days for receiving and put-away, 30 days of transit and customs, 45 days of supplier production, and 14 days of safety time on top. That's 98 days — about fourteen weeks. So the November peak has to be committed in the last week of July, roughly three months before the first customer walks in looking for it. Change one leg and the commit date moves with it; add a factory shutdown to the production leg and it moves weeks, not days. (Illustrative numbers.)
Size the order you only get to place once
Backward scheduling gives you a date. It says nothing about how much, and a peak buy is the order you don't get to correct.
Start with the grain. Buying decisions happen at variant level — so do the forecasts, per product or per variant, over the period you pick, and the period you pick has to span the whole peak.
Then check what last season is actually telling you. If a variant sold out in week two last year, the sales record shows demand stopping in week two, and sizing this year's buy off that record repeats the stockout deliberately. Assisty can exclude the days you were out of stock, so the forecast reflects real demand rather than the days you couldn't sell. On a peak buy sized from last season, that adjustment is most of the number.
Everything after that is constraint-fitting. Suggestions are computed from your sales velocity, lead times, and incoming stock, then rounded to your case sizes and capped at your maximum stock levels; fit-to-budget defers the least-urgent items and shows exactly what got cut.
One last thing the calendar decides: your peak is a single date, but the last safe day to commit is per supplier, not per store. Six suppliers, six backward counts, six deadlines. Assisty splits your selection into one PO per supplier and asks which location receives it — from there you review the line items, email the PO to your supplier, and receive against it when stock arrives.
| Step in the backward count | Who decides it | What Assisty gives you |
|---|---|---|
| The peak's start date | You — the calendar | You choose the forecast horizon; make it span the peak |
| How much to cover it | Assisty suggests, you decide | Predicted sales per product or per variant over the period you pick, with out-of-stock days excludable so last season's sell-out isn't read as last season's demand |
| Production (supplier lead time) | Your supplier | Feeds the suggestion: quantities are computed from your sales velocity, lead times, and incoming stock, then rounded to case sizes and capped at maximum stock levels |
| Transit & customs | You and your carrier | Your count — the suggestion starts from the supplier lead time you set |
| Receiving & put-away | You | Your count — days between arrival and sellable belong to your schedule |
| The commit date itself | You | Once you've decided, the selection becomes one PO per supplier with a receiving location |
You own the calendar — the date, the legs, the safety time. Assisty owns the quantity, the risk flags, and turning the decision into purchase orders. So do the cheap half now: open a calendar, write the peak's first day, and subtract the legs one at a time. The date you land on has either passed or it hasn't.