Markdown pricing is for selling through inventory by a deadline. Everyday Price Optimization asks "what price makes the most money right now?" Markdown pricing asks a harder question: "what price gets this stock sold by the date it needs to be gone, and how much margin should I give up to make that happen?" Instead of a single ongoing price, it plans a path of prices over a time window, so the right amount sells before that window closes.
Under the hood it reuses the same demand models the system already learned for each product during Price Optimization. What it adds is a calendar, a stock count, and a cost attached to anything left unsold. If you're deciding between everyday optimization and a markdown, see Dynamic Pricing vs. Price Optimization for the broader picture, and the section below on when each fits.
When to use it
Reach for markdown pricing when the goal is sell-through against a deadline, not steady profit:
End-of-season or end-of-life stock that must clear by a date.
Perishables or dated goods.
Discontinued lines being cleared out.
Warehouse or shelf space that needs freeing up.
For ongoing, open-ended pricing where there's no deadline and no urgency to deplete stock, everyday Price Optimization is the right tool instead. For a fuller comparison of goals to approaches, see Choosing Your Optimization Objective.
The big picture
A markdown run takes three things: the demand models the system already learned, your strategy (the objective, the start and end dates, and your safeguards), and a live snapshot of each product's price, cost, and stock count.
From those, it forecasts how much will sell at each price across the whole window, accounts for how much stock is on hand, applies a cost to whatever would be left unsold, and lands on a price for each product. For products it can't model, and for clearance campaigns as the deadline nears, it follows simpler rules described below.
The deadline matters throughout. Every forecast is "over this window," and how far into the window you are changes the answer.
Two clocks: time and inventory
Markdown pricing watches two clocks at once and responds to whichever is running faster.
The calendar clock is simply how far into the window you are: halfway through the season means it's roughly time to be halfway to your floor.
The inventory clock is how fast stock is actually depleting versus how fast it should. If sales are slow and a lot of stock remains with little time left, the inventory clock runs ahead of the calendar and pushes prices down faster to catch up.
The system takes the more urgent of the two. A product selling briskly rides the gentle calendar pace; a product that isn't moving gets marked down harder as its deadline approaches. This is why two similar products in the same campaign can be discounted at different speeds: the pace follows each product's own sell-through.
You can't sell what you don't have
A markdown can only sell as many units as are in stock. The system knows this: it caps its demand forecast at the available quantity. Dropping the price below the point where you'd already sell out adds no extra units; it just gives away margin for nothing.
So if a product would sell out at a modest discount, the system won't keep cutting deeper. It's the same constraint any store manager knows: you can't sell ten coats off a rack that holds three.
Products without a demand curve
Some products never had enough sales history for the system to learn how their demand responds to price. They can't be optimized against a model, because there's no curve to work with. (For why a product may not have a model, see Why Some Products Aren't Optimized.)
For these, markdown pricing falls back to a schedule: it glides the price in a straight line from today's price down toward the floor, paced by the two clocks above. No learning required, just a steady, predictable walk-down timed to the deadline and the stock level. These products are listed in the run's notices so they're easy to spot.
Products carrying a pricing action
Some products carry a rule-based action from the learning pipeline (for example, "keep price" or "lower price"). In ordinary stock optimization, these keep their pre-set price and skip the optimizer.
In a markdown or clearance campaign, though, the priority is clearing stock, so these products are put on the same scheduled walk-down as the no-curve products above, rather than being held at their action price. The campaign goal wins.
Safeguards and end dates
Two guardrails are worth knowing about:
Safeguards still apply. Floor and ceiling safeguards clamp the final price just as they do in everyday optimization. (In clearance, the floor behaves differently, as described in Choosing a Markdown Objective; ceilings and other protections hold as normal.) For how to set these, see How to Add Safeguards to Your Pricing Strategy.
A valid end date is required. Markdown pricing is built around a deadline. If a product's end date is missing or invalid, the system can't plan a path for it and leaves its price unchanged, flagging it in the run's errors. A missing or misconfigured end date is the single most common reason a product "didn't get marked down."
For the questions we hear most often about markdown runs, see Markdown & Clearance Pricing: Common Customer Questions.