The problem
The forecast lands after the range architecture is already agreed. By then all it can do is validate what we decided or annoy people. The one thing it could actually improve — how the season is shaped — is settled before anyone opens it.
The season skeleton determines most of what happens afterwards. Once story count, option counts and category roles are fixed, everything downstream is optimisation within a structure that may itself be wrong. Intelligence that arrives after the skeleton is set is not intelligence, it is commentary — and it is why so many businesses experience forecasting as an expense rather than an input.
The method
Assortment planning translates a point of view into a structure. Trend intelligence should be shaping that structure, which means it has to arrive before it, not alongside it. Each of the six stages maps onto a specific structural decision.
How many stories does the season actually support?
Story count should follow from how many genuinely distinct, corroborated directions exist — not from a template inherited from last season. Seasons differ; some support six stories and some support three. Forcing a fixed count creates filler stories, and filler stories consume option counts that a real story needed.
What to look at
- Count corroborated directions with distinct drivers. Two directions with the same driver are one story.
- Resist the inherited template. A four-story season is not a failure.
- Filler stories are identifiable in hindsight by having no driver anyone can state.
Which categories lead this season, and which follow?
Directions do not enter every category at once. Usually one or two categories lead — often accessories, footwear or a specific garment type — and others follow a season later. Assigning category roles from evidence rather than from historical share is one of the highest-value structural decisions available.
What to look at
- Identify which category the direction is furthest advanced in and give it the leading role.
- Give following categories fewer options and a later phase rather than an equal share.
- Historical category share is a poor guide when a direction is entering somewhere new.
How should the season be phased?
Velocity should shape drop structure. Fast-accelerating directions justify early phases with held-back reorder capacity; stable directions can sit in a single main phase. Phasing built purely from a calendar template ignores the fact that directions arrive at different speeds.
What to look at
- Phase fast directions early with reorder capacity retained.
- Single-phase stable directions to reduce complexity.
- Where a direction peaks mid-season, plan the exit into the structure rather than discovering it.
Does the skeleton differ by market?
Where markets sit at different adoption stages or have different occasion structures, the skeleton itself should differ — not just the allocation within it. A shared skeleton with different quantities is the usual compromise and it systematically under-serves the markets that are not the primary one.
What to look at
- Allow story count and category roles to differ by market, not only depth.
- Where a market has a distinct occasion calendar, it likely needs a story the others do not.
- A common core plus market-specific stories is usually more efficient than either extreme.
What are the option counts, and what carries over?
The core structural output. Option counts per story, and an explicit split between carry-over and new development. The carry-over decision is where development capacity is really allocated, and doing it explicitly — rather than by default from what happens to still be in the range — is the single biggest efficiency available in most assortments.
What to look at
- Set option counts per story from the direction’s breadth, not from an even distribution.
- Make carry-over an active decision with a reason, not a residue.
- An effort ladder — which options get real development and which get refreshed — prevents capacity being spread evenly across things that do not deserve it equally.
Which colours and materials carry the depth?
Within the structure, colour and material decide where depth goes. This is the last structural decision and it is often left to supplier availability by default. Evidence-led colour depth is one of the cheapest margin improvements available, because it costs nothing extra to buy the right colour deep instead of the wrong one.
What to look at
- Put depth behind colours with market evidence, not just the ones in stock.
- Group colour by story rather than running one flat seasonal palette.
- Check material availability early enough that the structure does not have to change around it.
How F-Predict answers this
ScopeSS27 · Womenswear · multi-category · India · emotion: joy, pride
Range architecture is a first-class output of a F-Trend Predict run rather than something derived afterwards, which is what allows it to arrive early enough to shape the skeleton.
- Range architecture
- Stories, option counts per story, and an explicit carry-over versus new-development split laid against an effort ladder — the skeleton itself, as a proposal.
- Corroboration scoring
- How many genuinely distinct, supported directions the season contains, which is what story count should follow from.
- Catwalk, street & material intelligence
- Which categories a direction is furthest advanced in, which sets leading and following category roles.
- Narrative intelligence
- Velocity per direction, which drives phasing and how much reorder capacity to hold.
- Regional scope
- Market-specific evidence, including the festive and wedding calendar in India that shapes both story structure and phasing.
- Colour direction
- Evidence-led colour themes grouped by story with Pantone references — the input to colour depth decisions.
The same decision from another desk
The six stages are the same across roles; what changes is what each stage means when you are the one making the call.
How do design managers decide which trends their team should follow?
Design managers choose which trends a team follows by ranking candidates against development capacity rather than against each other on appe…
How can a design manager validate a trend before approving it?
A trend is validated before approval by testing one candidate against five conditions in sequence: it appears in independent domains, someon…
How do you turn trend intelligence into a seasonal design brief?
A seasonal design brief is built by extracting one decision from each stage of the trend analysis: the driver behind each story, where it si…
Frequently asked
When should trend intelligence enter the assortment planning process?
Before the season skeleton is agreed. Arriving after, it can only validate or trim decisions already made, which is the most common reason forecasting is experienced as an expense.
Should option counts be equal across stories?
No. Option counts should follow the breadth of each direction — how many credible expressions it supports. Even distribution is a template, not a plan.
How is the carry-over decision usually made badly?
By default. Options carry over because they are still there rather than because a reason was stated, which quietly allocates development capacity to whatever survived rather than to what deserves it.
Keep reading
How should merchandisers balance emerging and established trends?
Emerging and established trends should be balanced as a deliberate portfolio rather than a negotiated outcome. Established directions carry volume and marg…
How does F-Predict help product development?
F-Predict supports product development by converting a forecast into decisions a range plan can absorb: named colour themes with Pantone references, materi…
How do you turn trend intelligence into a seasonal design brief?
A seasonal design brief is built by extracting one decision from each stage of the trend analysis: the driver behind each story, where it sits on the adopt…
How should trend intelligence influence buying decisions?
Trend intelligence should change four specific buying decisions: whether to buy depth or breadth, when to enter, how to phase the commitment, and how much…