Shopify Demand Forecasting for Seasonal & Promo Planning (2026)

Reordering is close to a solved problem. Purchase orders, receiving, supplier emails: those are the easy part. Seasonal and promotional planning is where forecasting quietly falls apart, because most tools are velocity engines. They take your recent sales, project the trend forward, and call it a forecast. Two failure modes follow directly from that design.

• Reason 01The promo-baseline problemA BFCM or flash-sale week sells five times normal. A naive model reads that spike as your new baseline and keeps over-ordering for weeks after the campaign ends, tying up cash in stock that will now sit.
• Reason 02The censored-demand problemWhen a product sells out, its recorded sales for that period reflect whatever was on the shelf, not what customers actually wanted. A velocity model reads a sold-out week as a low-demand week, then recommends a smaller reorder. That is how tools systematically under-buy the exact SKUs that are selling fastest.

Neither failure shows up in a demo. Both show up on a buying decision, months later, as either a warehouse full of dead stock or a stockout during your best week of the year. So the useful question is not "does it forecast," it is "does it forecast correctly when the data is messy."

Before you trust any tool with a buying decision, it should pass three tests. They map directly to the two failure modes above, plus the transparency you need to spend real money on a number.

• Insight 01Does it correct for stockouts?When a product was out of stock, is that period excluded from the forecast? If not, the tool will under-order your fastest movers, because a sold-out week looks identical to a slow week in the raw sales data. This is the single most important question.
• Insight 02Does it show you the math?A tool that outputs "order 400" with no breakdown will not survive contact with a real buying decision. You need the reasoning: starting stock, coverage period, forecast demand per day, safety stock, incoming inventory, then the final quantity. If you cannot audit the number, you cannot trust it.
• Insight 03Can you tell it about planned promotions?No forecast can infer a future sale on its own, because the campaign is not in the historical data. The tool has to let you enter a planned event over a date range, with an expected lift, so the forecast and the stock you buy reflect the campaign before it happens.

Monocle is built around all three. Here is how each one works.

1. It excludes stockout periods from the forecast

This is the censored-demand test. When a product was out of stock, Monocle leaves that period out of the forecast rather than treating it as low demand, so a sold-out week is never mistaken for a genuinely slow one. The forecast reflects true demand, not shelf availability, which is what stops the tool from quietly under-ordering your best sellers. Stockouts are flagged automatically, and you can confirm or adjust them yourself.

2. It adjusts to your store's specific demand

Monocle learns how your store actually sells rather than projecting last month's numbers forward. It picks up your real seasonal peaks and slow stretches, adapts as your sales shift, and still gives a sensible forecast for newer products that have little history of their own. The forecast is shaped around your demand, not a one-size-fits-all template.

3. A promotional event calendar that moves the forecast

You define events such as "Black Friday" or "Summer Sale" over a date range and give each an expected lift, either a percentage you set (say, a 25% bump) or one Monocle estimates for you, applied to all products or just certain tags, collections, or SKUs. The forecast adjusts for that window, so the stock you plan to buy reflects the campaign in advance. This is the answer to the reality that no tool can infer a future promo on its own: Monocle gives you the place to tell it.

4. A planning grid that shows where you will stock out

Monocle's planning workbench is a spreadsheet-style grid where you can override the forecast per product, per location, and per period, then watch a day-by-day end-stock projection update. It simulates incoming purchase orders against forecast demand and shows where you go negative, honoring whether a variant backorders or treats unmet demand as lost sales. Model a supplier delay by moving a PO's arrival date, or a flash sale by editing demand, and see the downstream stock effect before you commit to an order.

5. The rest of the replenishment loop

Beyond forecasting, Monocle closes the loop: ABC-XYZ classification and market basket analysis to prioritize the products that matter, purchase orders with multi-supplier lead times, a login-free supplier portal where vendors review and suggest line-item changes without an account, and a receiving workflow that reconciles what arrives against what you ordered. Forecasting, reorder math, and the PO all live in one place.

Do not judge a forecasting tool on the demo. Run it against a season you already know the answer to.

• Step 01Load last year's promo calendarenter the events, generate the forecast, and compare it to what actually sold. A tool that models promos well should land close; one that cannot will miss the spike or smear it across the wrong weeks.
• Step 02Check a SKU that sold outpick a product that went out of stock last season and confirm the forecast does not read that week as low demand. This is the censored-demand test in practice.
• Step 03Open one recommendation and read the mathif a tool cannot show you why it is recommending a quantity, it will not survive your first real buying decision. Monocle shows the full breakdown on every line.

How does Monocle forecast demand for seasonal products?

Monocle adjusts to your store's specific demand rather than assuming a generic pattern. It learns how your products actually sell through the year, so real seasonal peaks and quiet periods are built into the forecast instead of being smoothed over, and newer products with little history still get a sensible forecast. The result is shaped around your store, not a template.

How does Monocle handle stockouts and lost sales?

When a product sells out, Monocle treats that dip as a shortage, not a drop in demand. That keeps it from under-ordering your best sellers, the ones that keep selling out because they move fastest, so they stay in stock instead of quietly running dry.

Can Monocle account for planned promotions like Black Friday?

Yes. Monocle has a promotional event calendar where you enter events over a date range with an expected lift, either a percentage you set or one Monocle estimates for you, applied to all products or just certain tags, collections, or SKUs. The forecast adjusts for that window in advance, so the stock you buy reflects the campaign rather than scrambling after it.