← All posts

Forecasting

How to reorder when your sales are growing

The rate that sizes your order was true a month ago. The shelf it buys for is sold ten weeks from now, and a growing store is not the same store on both dates.

Marcus VolstedCo-founder, Restocio Published 27 October 202614 min read

Every source on this page was opened and read, and each one is dated in the Sources section at the foot.

A sales rate averaged over the last 60 days describes the store of a month ago. While you grow, an import order sized on it arrives short, because the shelf it lands on sells faster still: at 10 percent a month and a 72-day journey, the average sits about a quarter below the pace on landing day. Raise the pace only for growth you can prove: most of your catalogue selling faster, and holding for weeks, not one campaign.

This is a different undercount from the one empty-shelf days cause, where the rate is divided by days the product was not even for sale. Here the shelf stayed full and the arithmetic was honest. The number is simply old. It is also the mirror of a product whose sales are falling, where the same lag makes every order a little too big and strands cash instead of losing sales. Since 20 October 2026 Restocio counts a shop's measured growth into every order by default, and how it decides what counts is the second half of this post.

Why does a growing store run out when the reorder math is right?

Because the number the order is sized on was measured in the past and is spent in the future, and growth puts a slope between the two. An average gives every day in its window the same vote. Hyndman and Athanasopoulos say it plainly in Forecasting: Principles and Practice (3rd edition, OTexts, 2021, section 8.1): the average method “assumes that all observations are of equal importance”. On a 60-day window a day from early August weighs exactly as much as yesterday, so the figure's centre of gravity sits 30 days behind today. Harmless on a flat product. On a growing one it is a measurement of a smaller store, spent on a shelf that is sold ten weeks later, and the plan has quietly assumed the store stands still in between.

How far behind is a 60-day average when you grow?

Far enough to deserve a table. Take a store where every product grows at the same steady monthly rate, with no sold-out days and no sale days, and set its plain 60-day average against the pace today and the pace on the day a boat ordered today lands: day 72, two days at the supplier and 70 at sea. The last column shows what Restocio's growth switch would add for the same store, on a shop big enough for its 8 percent band, explained further down. This is a worked model, not a measurement from a live store.

The 10 percent row, so you can check it. Call today's pace 1. Sixty days ago, two months of growth earlier, the pace was 1 divided by 1.1 squared, about 0.826. The average of a pace growing smoothly across that window is (1 − 0.826) divided by (2 × ln 1.1), which is 0.91, and a day-by-day sum over the 60 days gives the same 0.91. Landing day is 72 days out, 2.4 months of growth, so the pace by then is 1.1 to the power 2.4, about 1.26. And 0.91 divided by 1.26 is 0.72: the average sits 28 percent under the pace on the day the stock arrives. The switch's column follows its own rule. Your last 30 days against the 30 before read 10 percent faster under steady growth, half of that would be 5, but just above the 8 percent band it counts less than half, so about 2.5 percent is added.

Is it growth, a campaign, or one big order?

Growth is the only one of the three that is broad, that holds, and that comes from many customers, and you can check all of that in Shopify before you touch an order. Compare each product's sales for the last month with the month before, look at how long the rise has lasted, and look at who bought. One account taking 400 units in an afternoon is one order inflating a rate, and the arithmetic for taking it out again is on that page.

A campaign belongs in your event days, not in your pace. Mark the dates on the Events page, in the sidebar just under Replenishment, and those days leave every product's rate at once; a sale you run every year goes in as a row that comes back on the same dates. If the campaign in question is November's, the Black Friday planning post works the quantity and the deadline for it.

How much of the growth should you order for?

Less than you measured, because the two ways of being wrong do not cost the same. A boat that lands short on a product that kept growing can be topped up by train or plane, at a freight price you will not enjoy, in weeks rather than months. A boat sized for growth that stalled lands anyway, and the extra sits on the shelf at the pace you actually have. That is how dead stock gets bought by careful people, and it ties up the cash a growing store is already short of, since cash goes out at tomorrow's sales rate and comes in at today's. If the next order is one you would need to borrow for, the asymmetry is sharper still.

So order earlier for any growth you can see, since a faster pace moves the deadline forward on its own, and order bigger only for growth that is broad and has held for weeks, and even then for part of it. A growing store sizing its pre-Chinese New Year double order faces this choice with twice the money; the CNY 2027 post has the dates.

Why does the order grow faster than the sales do?

Because an order is what the plan needs minus what you already hold, and the subtraction magnifies a small change in the pace. Take the example our own test suite pins: a product selling 10 a day, 800 on the shelf, two days at the supplier, a 70-day boat and 60 spare days. The plan buys 130 days of cover, the 70 at sea plus the 60 spare, less whatever is still on the shelf when the boat lands on day 72. Then the shop turns out to be selling 13 percent faster than the 30 days before, the switch counts half, and the product is planned on 10.65 a day instead of 10.

The 800 units already on the shelf are the lever: they drain faster at the higher pace, so less is left on landing day, while the target has grown, and the order absorbs both moves. These figures are what the engine returns for this case. They are also why every Restocio screen states the change in speed and never the change in the order: 6.5 percent printed beside an order that grew 10.8 percent would look like an error, and it is not one.

What does Restocio do when your shop grows?

It measures the shop's growth itself and counts a cautious share of it into every order, with one switch and a printed receipt, rather than asking you to guess a percentage. The switch is in Settings, under Inventory, in the step called “How fast you sell”: “Count my shop's growth into every order”, marked “Recommended: on”. It is on for every store, stores set up before it existed included, and one click turns it off, at which point every product goes back to its plain pace.

What it counts is narrow on purpose. The app compares a usual day in your last 30 days with a usual day in the 30 before, so it needs 60 days since your first recorded sale and counts down until it has them. It then reads two speeds, the shop as a whole and a typical product, the one in the middle when your products are lined up from the one that grew least to the one that grew most. The card's own words for the next step are “We count half of the smaller of the two (at most +20 %)”. A change under 8 percent, wider on a small shop whose ordinary ups and downs are bigger, adds nothing, and just above that line it counts less than half so there is no cliff. The typical product is what keeps the growth broad: a few fast sellers in a catalogue of hundreds cannot lift every product. Sold-out days, event days, days the whole shop sold more than one and a half times a normal day, and products that sold on fewer than ten days in either stretch are all left out, and one product can move the shop by 5 percent at most (in a shop of a few products, by a quarter of its own sales). If the newest week has fallen back to the speed of the month before, the growth is not holding and nothing is added for now.

A product with a rise of its own gets a bit more. When its own sales have run at one and a half times its usual pace, measured over its 60 days before, on most of the last two weeks and are still up, the app closes a third of the gap between the shop-grown pace and that faster pace. It keeps that while the product stays at least one and a quarter times its usual, for up to 60 days, with sold-out days left out, and the shop's growth is never added twice.

No sudden jump is guessed at. When sales step up for more than a week and leaving those days out would change the answer, the Dashboard asks: “Sales jumped from 20 Oct. Are you running a campaign?” Yes saves the dates as an event, with a from and to date or “Not sure when it ends”, in which case it ends by itself when sales are back to normal, after 30 days at most. “No, it's normal growth” counts it from then on. Until someone answers, the jump days are left out so a sale cannot make your orders too big, and if sales are still up after 30 days the jump counts as growth anyway, phased in over a week. While it waits, the same question also sits inside each affected product's “How this was calculated” on the Replenishment cards.

None of it is silent. The Settings card prints the whole sum, a usual day in each window, how much faster, the half, the multiplier, and two of your own products worked out. Every product's “How this was calculated” receipt prints its own version, and a pace the switch has moved carries a small “(estimate)” marker on the Replenishment card, explained in the page guide, because a forecast from your own sales can be wrong if a sale or a one-off order was never marked. The switch never touches how careful you asked the plan to be: your spare days, supplier cards, minimums and any dates you typed stay as they were. Before it shipped on 20 October 2026 it was tested by replaying a real store's year of sales with the switch on and off, counting the sales lost to empty shelves, the units bought and the stock left on the shelf. Where the pace sits in the rest of the decision, deadline and quantity included, is the Shopify replenishment guide.

When does the growth switch add nothing?

More often than a growing store might like, and that is a choice with a reason rather than a gap. Half of a 30-day change, nothing under 8 percent and never more than 20 from the shop's growth means a shop growing a steady 5 to 10 percent a month, strong growth by any ordinary measure, gets little or nothing added: at 10 percent a month the planned pace lands at 93 percent of today's instead of 91. That store still has to order earlier by itself, and the trend line on the Demand page is a view of the slope, not an input to any order. The reason for the timidity is the asymmetry from earlier. A boat bought for growth that stalls cannot be undone, and the app would rather leave you a little short on a rise it can see than long on one it imagined.

Four smaller limits, said plainly. It never lowers the pace for a shrinking shop; that remains your call and its own post. It needs 60 days since your first recorded sale before it adds anything. A seasonal peak counts as growth while it lasts and stops within about a week of its end, so the switch is no substitute for a seasonal product's own last-year pace. And a pace borrowed from an older stretch or from last year's season keeps its own number, with no growth added on top.

Common questions

How do I adjust my inventory forecast for growth?

Measure the shop's last 30 days against the 30 before, and raise the pace by a share of the difference only when most of the catalogue moved and the rise has held for weeks. Half the measured change is a defensible share, because a short order can be topped up by a faster route while a boat bought for growth that stalled cannot be undone. Order earlier for all the growth you can see; order bigger for part of it.

Should I use a shorter sales window when sales are growing?

It helps, with a cost. The window is one setting for the whole catalogue (in Restocio, Settings, Inventory, "Sales period the pace is measured over", 60 days recommended), so a shorter one follows a rise sooner but swings from week to week on every product, and more stock ends up on the shelf. A 60-day average sits about 9 percent under today's pace at 10 percent growth a month; a 30-day window roughly halves that lag and is noisier from week to week. Counting growth as a separate step leaves the window calm and still moves the pace.

How can I tell sales growth from a promotion?

Ask how many products moved, for how long, and for how many customers. Growth lifts most of the catalogue, holds for weeks and comes from many buyers. A promotion lifts the products on offer for days and then falls back, and a wholesale order lifts one product for one day. Restocio asks the question for you on the Dashboard when sales jump for more than a week: say it is a campaign and the dates become an event that leaves the pace; say it is normal growth and it counts.

Does Restocio account for growth automatically?

Yes. The switch "Count my shop's growth into every order" in Settings, Inventory is on for every store. It compares your last 30 days with the 30 before, takes the smaller of the shop's speed-up and a typical product's, counts half of it, and that shop-wide share never adds more than 20 percent. A product that has proven a rise of its own gets a third of the way to its own faster pace on top. A change under 8 percent adds nothing. One click turns it off and every product returns to its plain pace.

Why did my reorder quantity jump more than my sales did?

Because the order is what the plan needs minus what you already hold. In the example pinned in Restocio's own tests, a shop selling 13 percent faster gives a product a 6.5 percent higher pace, 10.65 a day instead of 10, and the boat order moves from 1,220 to 1,352 units, 10.8 percent more. The stock on the shelf drains faster at the higher pace, so less is left on landing day, while the target grows at the same time. That is why Restocio's screens state the change in speed, never the change in the order.

Does a growing store need more safety stock?

Not because of the growth itself. Growth is a pace problem and the right fix is a more honest pace, which moves the order date earlier and the quantity up together. Your spare days protect against a late boat and a strong sales week, and Restocio's growth switch never touches them: your spare days, supplier cards, minimums and typed dates stay as they were. Raise the cushion if deliveries have been late, not because sales are up.

Sources

Every source below was opened and read on the date shown against it. Rules, rates and schedules change, so check the current page before you act on a number that matters. If anything here is out of date or wrong, tell us and we will correct it.

• Hyndman and Athanasopoulos, Forecasting: Principles and Practice, 3rd edition (OTexts, Melbourne, 2021), section 8.1 Simple exponential smoothing, on the weight an average gives each observation (read 25 October 2026)

Marcus Volsted

Co-founder, Restocio

Marcus co-founded Restocio and works on it daily with a Swedish importer who plans their purchasing in it every working day. Restocio is built in Sweden by two founders, one Swedish and one Danish, and Marcus is the Danish one. Why we are building it.

Related reading

• Sales are declining, so should you still reorder the same amount
• How to tell whether your forecast is any good
• How to stop running out of stock on your best sellers
• What one big order does to your sales rate
• Inventory forecasting for Shopify, and what it cannot tell you