We Launched Too Much. The Markdown Paid for It.

No one could explain it.

The category launched.
The inventory was there.
The numbers didn’t reconcile.

The SKUs that should have carried the business ran out.
The ones that shouldn’t have been there stayed funded.

By the time anyone asked why, the capital was already committed.

The launch was already decided

We launched a new product category.

The development work had been running for nine to ten months. Sampling, costing, sourcing, line planning. By the time pre-season planning arrived, the decision to launch the full breadth was already embedded in the organization.

Not because demand validated it.
Because the work had been done.

When the capital exposure was raised, the answer was consistent.

We will figure that out later.
This product is hot. It will sell itself.

Nine months of development makes those sentences easy to believe. It also makes them easy to act on without scrutiny.

So everything launched.

No productivity threshold.
No reorder gate.
No exit strategy.

Every SKU that existed moved forward.

The development investment didn’t support the launch.

It became the launch strategy.

The sunk cost that kept everything in

Most teams understand sunk cost in theory.

In practice, nine months of development creates momentum that behaves like a capital commitment before a single purchase order is placed.

Every SKU that made it through development feels like it earned its place. Cutting it before launch feels like writing off the work.

So nothing gets cut.

The result is predictable.

The catalog fills with unproven SKUs.
Capital spreads across everything.
The SKUs with real demand get underfunded.
The ones without demand stay supported.

That is not a forecasting failure.

It is a launch architecture failure.

Nothing was built to stop it

The second failure was structural.

The category launched with an evergreen mindset.

These weren’t treated like products that needed to prove themselves. They were treated like products that would sell indefinitely.

That assumption removed the exit conversation entirely.

When no SKU has a defined threshold, the default state is simple.

Keep it funded.

No trigger forces a decision.
No threshold requires validation.
No mechanism stops capital from moving.

So it doesn’t stop.

Evergreen is not the problem.

Uncontrolled evergreen is.

Every SKU still needs a moment where it proves it deserves more capital.

Without that moment, the catalog becomes a permanent claim on cash.

When the OTB closed

This is where it shows up.

The SKUs that weren’t moving stayed funded.
The catalog stayed broad.
Capital kept spreading.

And when sell-through didn’t match the commitment, nothing stopped the next round of inventory from going out.

By the time the OTB closed, the capital was already locked.

Not into demand.

Into assumption.

At that point, there’s only one way to get cash back.

The markdown.

The cost the markdown didn’t show

The markdown clears inventory.

It doesn’t clear the consequence.

A premium product sells at full price.
Then it shows up at a discount.

The customer learns something.

Not about the product.
About the timing.

They don’t announce it.
They don’t complain.

They just wait.

Next season, they wait again.

Full-price sell-through slows.
Promotional dependency increases.
Margin erodes without a clear explanation.

Nothing about the product changed.

The behavior did.

And that behavior was trained.

What changes the outcome

The failure happened in two places.

Before the launch.
Before the reorder.

A launch-to-test structure changes both.

Before launch, every SKU clears a threshold.
If demand hasn’t been proven, it doesn’t go to full depth.
It launches to test.

Limited capital.
Defined window.
Clear criteria.

Before reorder, the same rule applies.

If demand hasn’t materialized, capital doesn’t recommit.

The SKU exists.
That doesn’t mean it deserves more inventory.

This is not a forecasting problem.

It’s a decision structure.

Capital should not move because a SKU exists.

It should move because demand proved it should.

Where this starts

If your catalog carries SKUs that launched without a threshold,
or reorders that moved without a gate, the markdown is not the problem.

It is the bill.

⚜️

Explore the Structural Performance Audit