Retail markdown is not a strategy. It's a symptom.

The retailers using markdown as a default promotional lever aren't making a commercial choice. They're simply paying the price for decisions that weren't made earlier.

The first two articles in this series made two connected arguments. First: that pricing failure in retail compounds quietly through decisions deferred, signals missed, and granular differences absorbed into category averages. Second: that the structural disconnection between pricing and inventory is the condition that makes that compounding almost inevitable.

Reactive markdown—a permanent or temporary reduction in an item's original selling price—is where both of those arguments become visible. It’s the most tangible, most expensive expression of what happens when pricing intelligence and inventory reality aren't connected, and when the gap between decision speed and market speed has grown too wide to close without a blunt instrument.

Understanding markdown as a symptom, rather than a strategy, changes how you think about fixing it.

Reactive markdown is not a pricing decision. It’s the cost of not making the right pricing decision earlier. And it’s almost always more expensive than it needed to be.

How markdown became the default

The shift from planned markdown to reactive markdown is rarely a deliberate choice. It happens gradually, as the conditions described in the previous two articles accumulate. Pricing and inventory operating in silos means the full picture of a product's commercial position is never available in one place. Signals that should trigger early action arrive late, or arrive fragmented across systems that don't talk to each other.

Teams adapt. They get faster at reacting, and they get better at executing discounts quickly. But speed of reaction is not the same as quality of decision. A markdown triggered under pressure, without visibility into inventory cover, location-level demand, or SKU-level elasticity, will almost always be deeper than it needs to be.

Over time, the habit embeds. Something isn't moving? Discount it. Campaign underperformed? Slash the price even further. Stock building into the end of season? Time for a blanket promotion. Each decision has a logic, but collectively, they represent a drain on margin that compounds quietly and persistently across every trading period.

Revisiting the elasticity problem

My first article argued that elasticity is never uniform. The same price point on two products in the same category can produce completely different outcomes. Those differences disappear into the aggregate when pricing decisions are made at category level.

Nowhere is that more commercially damaging than in markdown. A broad category discount contains, within it, products that would have sold without that depth of intervention, either at full price or at a shallower discount trigger. Those units represent pure margin destruction: discount applied where it had no commercial purpose, invisible in the aggregate because the sell-through number moved.

At the same time, there are products within the same markdown where the category-level discount isn't enough; where demand is genuinely inelastic at that price point, in that location and at that moment. The blanket approach both over-discounts and under-responds simultaneously. The margin cost of each is real, and neither is visible without SKU-level elasticity intelligence.

A category markdown that looks decisive is often destroying margin on products that didn't need it, while failing to move the ones that did. The difference is invisible without granular elasticity data.

The 90-day principle

Consider what changes when agentic AI is continuously monitoring sell-through velocity, inventory cover, and location-level demand signals. And not at the point of crisis, but 90 days before it. The picture is clear long before it becomes urgent, and the system surfaces the insight at a point when there is still genuine commercial optionality.

High-performing inventory can be identified and transferred to locations where demand is stronger, before it accumulates in stores where it isn't and starts to gather dust. Markdown triggers can be calibrated to the actual demand trajectory of individual SKUs, in individual locations, rather than category averages. The depth of intervention required at exit reduces because the decision was made early, with full visibility, rather than late, under pressure.

The commercial difference between a store closing at 40% off rather than 70% off is not marginal. It is the difference between a managed commercial exit and a brand-damaging fire sale. The same principle holds at category level, campaign level, and SKU level. The depth of every markdown is, in part, a function of how early and how precisely the decision was made.

From symptom to precision instrument

The goal is not to mark down less. Markdown is a legitimate and powerful commercial tool in retail, and is crucial for managing inventory exposure, stimulating demand at the right moment, and protecting exit margin. The goal is to mark down right: at the right depth, at the right moment, for the right reasons, with full visibility of the inventory and working capital consequences before the decision is made.

Agentic AI makes that precision possible at a scale that manual processes cannot reach. We’re talking about continuous elasticity modeling at SKU and location level, scenario simulation that shows the margin versus revenue trade-off before anything goes live, and automated execution within guardrails that enforce floor prices and brand positioning consistently. The commercial team retains control, but what changes is the quality of the intelligence behind every call they make.

When those conditions exist, markdown stops being a reaction and becomes a decision. That shift, from symptom to instrument, is worth more to a retail business than almost any other operational change it can make.

In the final article, we'll bring the full argument together: why retailers building this kind of agentic AI capability now are creating a structural competitive advantage…and why the window to do so is narrowing.