Merchandise Planning in Fashion: Balancing Trend Risk with Demand Precision

Most teams still frame merchandise planning as a forecasting problem. It’s not. Not in fashion.

You’re not chasing precision. You’re deciding where you can afford to be wrong.

Product lifecycles are short. A dress might matter for eight weeks. A denim trend might hit—or disappear halfway through the season. Historical data is thin, sometimes misleading. External signals move faster than your planning cycle. Weather shifts. Trends spike or stall.

So the real job is allocating capital under uncertainty.

Every buy is a risk decision. Not just “how many units,” but what kind of risk you’re taking on.

You’re balancing:

• Core products that keep cash flow stable
• Seasonal items with some predictability
• Trend bets that might outperform—or sit

And once you commit, you’re mostly locked. Lead times, MOQs, vendor constraints. You can tweak around the edges, but the exposure is already set.

Bad planning doesn’t show up in forecast accuracy reports. It shows up in inventory.

Stockouts on items that would’ve sold all day. Overstock in the wrong styles or sizes. Cash stuck in units that need discounting to move.

The goal isn’t eliminating error. That’s unrealistic. It’s choosing where error is acceptable.

Missing slightly on a trend piece? Fine. Running out of core sizes on a proven style? Not fine.

That’s the job. Everything else is tooling.

Structuring the Assortment as a Portfolio of Risk

Strong planners don’t start with SKUs. They start with risk.

Most assortments fall into three buckets, whether labeled or not:

• Core: repeat styles, predictable demand, reliable size curves
• Fashion: seasonal updates, variations on known winners
• Trend: new silhouettes, colors, cuts—high upside, high volatility

Each behaves differently. Treating them the same is where things break.

Core gets depth. You know your WOS targets. Replenishment is predictable. Size curves are mostly stable.

Trend is the opposite. You buy shallow. You’re buying optionality, not volume.

Fashion sits in the middle—and that’s where teams get uncomfortable. Enough data to feel confident, enough variability to be wrong.

Lean too hard into trend and you create markdown risk. Demand doesn’t show, inventory stacks fast.

Lean too heavily into core and you protect margin—but leave revenue on the table. The assortment gets safe. Maybe too safe.

The right mix isn’t static. It shifts by category, season, brand positioning.

A basics-heavy brand can run mostly core. A trend-driven brand can’t.

And this isn’t really about units. It’s exposure—margin dollars at risk.

Depth vs. Breadth: Where the Real Damage Happens

Most issues don’t start at the category level. They show up at SKU and size level.

You decide:

• How many styles (breadth)
• How deep to go (depth)
• How that depth splits across sizes

Then reality kicks in.

Demand is never evenly distributed across sizes. It just isn’t.

You’ll see:

• Core sizes gone in weeks
• Fringe sizes barely moving
• Size breaks killing sell-through on otherwise strong styles

A “balanced” size curve on paper rarely holds in practice.

Spread depth evenly and you trap inventory. Over-correct and you stock out in high-velocity sizes.

This is where margin leaks quietly. Not because the style failed—but because the size allocation did.

Most teams still manage this in spreadsheets. Static curves. Manual overrides. Slow feedback.

By the time the imbalance is obvious, it’s already expensive.

Why Forecasting Is Structurally Imperfect

Forecast accuracy gets too much attention.

In fashion, it’s inherently limited.

You’re dealing with short lifecycles, new products, and external demand drivers you don’t control.

Even strong models struggle with newness. You can’t fully predict how a new silhouette will land.

Core behaves differently. You can model it. Replenish it with confidence.

Everything else requires a mix:

• Historical data (solid for core, directional for fashion)
• Attribute-based forecasting (helpful, not definitive)
• Merchant judgment (trend awareness, market context)

Forecasting isn’t the problem. Treating it as truth is.

Teams that do get blindsided. Teams that treat it as a starting point tend to adapt better.

Testing and Early Signals

The only real way to reduce uncertainty mid-season is to test.

Small buys. Limited rollout. Controlled exposure.

Then you watch:

• Early sell-through
• Size-level performance
• Store-level variation

It’s not perfect signal, but it’s usable.

If something hits, you chase. If it stalls, you contain the downside.

Without testing, you’re guessing at scale. And that’s expensive either way—overbuying or underbuying.

Testing just gives you a steering wheel.

The Real Cost of Getting It Wrong

Planning errors don’t show up in dashboards first. They show up on the floor.

Stockouts in core sizes are obvious—you lose the sale, and potentially the customer.

More subtle: size breaks.

A style can be 60% in stock and still not sell. Missing key sizes kills conversion.

Then there’s overstock. Not just too much inventory—wrong inventory.

You’ll see mid-sizes gone, fringe sizes sitting. On paper, you’re in stock. In reality, you’re not.

That leads to discounting. Margin erodes, even when demand existed.

Same with trend bets. If demand underperforms, markdowns start earlier than planned. Then come the knock-on effects:

• Lower realized margin
• More reliance on promotions
• Inventory lingering late into season

Markdowns aren’t cleanup—they’re margin recovery. Timing matters.

Too late, you miss the window. Too early, you give margin away.

The cycle is familiar: overstock, late markdowns, margin pressure.

Better planning doesn’t remove it. It just reduces how often—and how badly—it happens.

From Plan to Execution

Even a good plan can fail in execution.

You can buy the right units and still underperform if inventory lands in the wrong places.

Allocation is where things get real:

• Which stores get what
• How depth is split
• How size curves vary by location

Uniform allocation rarely works.

Then replenishment. Core should flow based on actual sales—not static plans.

If something is moving, you need to feed it. If it’s slowing, you need to pull back.

The issue is speed.

Most teams are working with lag—weekly reports, manual updates, spreadsheet workflows.

By the time action happens, the moment has passed.

In-season control is about tightening that loop.

Watch sell-through, WOS, emerging stockouts—and act early.

Reallocate. Adjust replenishment. Cut exposure where needed.

Retailers that connect planning with execution still make mistakes. They just recover faster.

Others stay stuck with the original plan, even when it’s clearly off.

What Better Looks Like

The gap isn’t effort. It’s how the system is set up.

High-performing retailers treat planning as continuous.

Not seasonal. Not one-and-done.

They revisit assumptions. Adjust to real demand. Rebalance risk mid-season.

They don’t rely on a single input. They combine:

• Data models
• Merchant judgment
• Operational feedback

There’s tension there—but it’s useful.

They invest in visibility. Know where inventory is, and how it’s moving.

They shorten feedback loops. Faster data, faster decisions.

And for volatile categories, they plan lighter upfront and react faster in-season.

The core problems haven’t changed:

• Too much inventory
• Not enough of the right inventory
• Slow reaction cycles

Better forecasting alone won’t fix that.

You need tighter alignment between planning, allocation, and replenishment—and systems that can respond when reality diverges from the plan.

Because it always does.