The hidden cost of an order routing setup is easiest to see at the end-of-season markdown review. A category closes the season with weeks of stock still sitting in stores, all of it about to be discounted.
Look back over the season and the reason is usually the same: the routing engine shipped those styles from the distribution center week after week and it never drew down the store stock that is now stranded.
No one set out to do that. The team optimized order routing for the numbers they could see on every order, shipping cost and delivery speed, and the choice that set up the markdown was never made on purpose.
The size of that miss is easy to underrate. Between 30% and 40% of apparel sells at a discount, and much of it never sells at full price at all. A routing choice that saves a split shipment is worth a few dollars on the order. A routing choice that forces a markdown is worth 30% to 40% of the item. Retailers optimize hard for the first number and leave the second one unmanaged, which is the real question behind order routing optimization: what are you actually optimizing for?
Why sell-through loses at the routing table
Order routing optimization rewards the costs that arrive immediately. Shipping spend shows up on the carrier invoice that month, and a missed delivery date shows up as a service complaint the same week. Both are easy to see, easy to attribute to routing, and easy to tune against.
A markdown works differently. When routing sends full-price online demand to a warehouse and leaves slow-moving stock sitting in stores, that stock keeps aging until the season ends and it is discounted. The cost is real, but it lands months later, and it lands in the merchandising markdown budget rather than the routing report. By the time anyone sees it, it no longer looks like a routing decision.
That deferred cost is not small. Excess fashion inventory in 2023 was valued at $70 billion to $140 billion. Most guides treat routing as a logistics problem to fix with cheaper shipping. It is better understood as a question about inventory health that only looks like a logistics one. Sell-through loses at the routing table because of when its cost shows up, not because it matters less than shipping.
Why a single routing setting cannot serve a product's whole life
Most retailers do weigh inventory health, but they treat it as a tiebreaker. Routing runs as one static setting for the whole year: optimize for cost, or optimize for speed, with inventory health dropped to the bottom of the stack as a last resort.
The trouble is that what a SKU needs from routing changes over its life. A style that just launched needs its scarce stock protected, so it does not sell out in one region while another region misses the launch. The same style, weeks later and heading for markdown, needs the opposite: online demand should go to the stores sitting on it, so the stock draws down at full price before it is discounted. One global setting cannot do both.
Speed gets over-weighted the same way. When asked, 75% of shoppers choose free or cheaper shipping over faster shipping, the longest most shoppers will wait is about 2.6 days, down from a 3.5-day average. Speed still matters on the orders that need it, but paying for it on every order, at the cost of where inventory ships from, is the same objective set once and left alone.
Make routing a strategy you set
Which should your OMS optimize for, speed, cost, or sell-through? The honest answer is that you should not have to choose one and live with it, because good routing balances all three at once. What you need is not a fixed setting but a strategy you compose, and can recompose, without a developer.
HotWax Commerce captures every online order and decides which location fulfills it, and it turns that decision into a strategy the operations team builds from layered rules, not a default baked into the sales channel.
Balance comes from layering the rules, not from picking one. Say a category is heading for markdown. You route those SKUs first to the stores sitting on the most of that stock, so the aging units draw down at full price, and within those stores you ship from the nearest one, so you clear the markdown risk without giving up on freight.
Flip the goal and the same layering protects the other side: when speed leads and orders route to the nearest store first, you break ties by which store holds the most stock, so a fast rule never strips a location that was already low. One objective leads, and the others still hold.
You set that strategy once, as a no-code change, and the engine runs it on every order, rerouting to the next store on its own when one cannot fill, while store associates pick and pack through Ship From Store.
You do not keep one strategy all year, either. Ahead of the holiday peak you can build a speed-leaning strategy in advance and schedule it, and for end-of-season clearance, a sell-through-leaning one. Each is a change to the rules, not a rebuild.
Because a strategy moves real orders and real margin, you do not guess at it. Order routing simulation runs a strategy against real orders and shows exactly where each one would go, so you can see whether it hits the goal you set before it touches live traffic. You are not tuning one dial and hoping; you are choosing the strategy that gets the outcome you need.
The answer to speed, cost, or sell-through is that you should not answer it once. You compose the strategy the moment calls for, balance the trade-offs inside it, and change it when the moment changes.
Key takeaways
- Routing is an inventory-health decision disguised as a logistics decision.
- The cheapest lever to optimize is not the most expensive one to ignore. A saved split shipment is worth a few dollars; a forced markdown is worth 30% to 40% of the item's value.
- Sell-through loses at the routing table because its cost is deferred and invisible, not because it matters less.
- The same SKU should route to protect scarce stock early in its life and to draw down aging store stock late in its life.
- A routing objective that never changes by season, channel, or product lifecycle leaves margin on the table.
- If you cannot test a routing rule change before it hits live orders, you are optimizing blind.
Try it on one aging category
Pick one category heading for markdown, or one seasonal line you want to clear. Group the stores holding it, write a sell-through-first rule for those SKUs, and run order routing simulation to see how those orders reroute before the change touches live traffic. Book a walkthrough with HotWax to set it up on your own data.
