How Athory decides what to reorder — and when.

Most inventory tools tell you when you're running low. Athory works out the right quantity, for the right product, from the right supplier, at the right moment — running the kind of supply-chain science large operations rely on, automatically, on your store's own data.

Re-evaluated every day, for every product, across dozens of signals.

The tip of the iceberg

You see one number. Underneath it, the work that earns it.

Recommendationlive
Ethiopia Yirgacheffe · 1kgYRG-1KG
240units
Reorder — within 4 days · Drammen Coffee Co.
Verifiedrecomputed 4 min ago
What you seeWhat earns it
Read the demandmillions of lines
Fit the forecasta model per product
Size the bufferboth directions
Time the orderordering vs holding cost
Re-checked as events moverunning, daily

Thousands of signals, weighed continuously. You read one line.

Here's each layer, in order.

Stage 1 · Read the demand

It starts by understanding how each product actually sells.

Every product sells its own way — a steady earner, a seasonal spiker, a slow long-tail item, or a brand-new SKU with little history. Before forecasting, Athory strips the distortions out of raw sales — a one-off bulk order, a split shipment double-counted, a quiet week that was really a stockout — and reads each product's true demand and rhythm on its own. Get this first read wrong and you over-order for months on a signal that was never real.

Why this is hard

A wholesale order or a stockout gap can look exactly like a jump in demand. Take it at face value and everything downstream inherits the mistake — which is exactly where most tools quietly fall down.

The Common Approach

Averages the last 30 or 90 days of raw sales.

How Athory does it

Separates genuine demand from bulk orders, blips, double-counts and stockout gaps before forecasting anything.

Stage 2 · Fit the forecast

It forecasts each product with the method that fits it best — not one formula for them all.

Most tools run one formula across the whole catalogue. Athory runs a contest instead: a library of methods competes for each product, and only the one that best predicts that product's own history survives — re-run as sales change. Every forecast carries an honest uncertainty band, on every SKU you carry — even the long tail no one would forecast by hand.

Why this is hard

One formula is too smooth for your spiky sellers and too jittery for your steady ones — wrong for most products at once. A plain average only knows what just happened, so it's always a step behind the season or trend that matters most.

The Common Approach

One forecasting formula, stretched across every product.

How Athory does it

A library of methods competes per product — the one that best fits its real history wins, and keeps proving itself as sales change.

Stage 3 · Size the buffer

It sizes a safety buffer around two kinds of uncertainty — not just your average sales.

Your safety buffer sits between two costs — stock out and lose the sale, or overstock and trap cash. Athory sizes it from two risks most tools ignore: how much demand swings, and how reliably each supplier actually delivers — measured from the POs you've received, not the lead time they quote. So a flaky supplier earns a bigger buffer than a dependable one for the same product, your heroes are guarded tighter than the long tail, and buffers loosen on their own as forecasts sharpen.

Why this is hard

A flat "weeks of cover" assumes demand and delivery never wobble. They always do — which is how you end up with stockouts and dead stock at the same time, in the same store.

The Common Approach

A fixed "weeks of cover" applied to everything.

How Athory does it

A buffer sized to each product's demand swings and its supplier's measured reliability — tighter on bestsellers, looser on the long tail.

Stage 4 · Time the order

It works out how much to order — and whether now is actually the right moment.

Knowing you'll need more isn't knowing how much, or when. Athory answers both as an economic call, per supplier — and the economics fit your store, not a textbook: ordering cost is learned from your own PO history, orders round to supplier minimums and case sizes, and extra items only join a PO when their margin justifies ordering early. It even nets out stock already inbound or spoken for, and will hold an order a few days when waiting is cheaper — then release it the moment the maths flips.

Why this is hard

"Alert me when stock crosses a line" over-orders, under-orders, and ignores the economics entirely — and ignores what's already on the way, so you double-buy. The right moment is a trade-off recalculated per supplier every day; nobody has time to do that by hand.

The Common Approach

Alerts you the moment stock crosses a reorder line.

How Athory does it

Weighs your real ordering cost, holding cost, supplier minimums and margin-aware bundling to pick the most economical moment to buy.

Stage 5 · The recommendation

The result: a finished decision per product, kept current for you.

It all collapses into one finished decision per product — what to order, how much, from which supplier, by when — as a supplier-ready PO you accept in a click, not a low-stock list you still have to interpret. You get value from day one, not day ninety: Athory works from your existing orders the moment you connect, then sharpens as each product earns its own tuned method. Every recommendation stays current on its own while you stay in control — accept, adjust, snooze, or order to a budget.

Why this is hard

Most systems make you wait months to "warm up," then hand you a list you still have to turn into decisions yourself — by which point you've made this season's calls by gut. The hard part is giving you a trustworthy, supplier-ready decision early, and keeping every one current as the store moves.

The Common Approach

A low-stock list you still have to turn into decisions.

How Athory does it

Finished purchase-order recommendations — quantity, supplier, timing — kept current for you, ready to accept, adjust, or order to a budget.

What it weighs

And behind all five layers, it balances factors like these — at once, for every product.

Not a checklist run once. A balance struck continuously, where changing any one of these can change what — and when — you should order.

Demand trendSeasonalitySales volatilitySales-history depthForecast confidenceLead timeLead-time reliabilitySupplier minimumsPack & case sizesBundling potentialOrdering costHolding costStockout costProduct marginRevenue importanceProduct costOn-hand stockUnfulfilled ordersIncoming POsWarehouse splitCash impact

See it on your own store.

Connect your Shopify store and watch Athory build its first recommendations within days.

Request Early Access