Skip to content This library is free, offered as dana. If it helps your work, donate here →
penguin tree ai
← back to libraryDonate here

This is one of 3,622 free prompts in the penguin tree ai library. Sign up once and the whole library opens, with new prompts by email.

open the library →

Assortment performance trend spotter

Analyzes weekly assortment performance data and surfaces emerging trends, breakout outperformers, and fading underperformers at category and SKU level — with sell-through velocity, week-over-week momentum flags, and…

download .md

Purpose

Turn a weekly performance export into a ranked, action-first read of the assortment — flagging which categories and SKUs are accelerating, decelerating, or stalling — so planners chase winners before they sell out and clear losers before margin erodes. Replaces manual sorting through spreadsheets with a repeatable trend diagnosis and a prioritized action list.

When to Use

Invoke this skill when the user asks to:

  • Analyze this week's sales and inventory performance
  • Find emerging trends or breakout SKUs in the assortment
  • Identify which products to reorder, chase, or expand
  • Spot underperformers to mark down or discontinue
  • Review category-level or SKU-level performance week-over-week
  • Prep a weekly merchandising or open-to-buy review

Inputs Required

Before starting, collect:

  • Weekly performance data: Units sold, revenue, and on-hand inventory per SKU or category — ideally with 4+ weeks of history for trend detection
  • Level of analysis: Category, subcategory, SKU, or all three
  • Reporting period: The current week and how many prior weeks are included
  • Cost/margin data: Unit cost or margin % per item (for markdown vs. reorder decisions)
  • Reorder lead time: Weeks from PO to receipt (determines chase feasibility)
  • Business context: Season, promotions, launches, or stockouts affecting the read
  • Targets/benchmarks: Sell-through goals or WOS (weeks of supply) targets, if any

If margin data or lead time is missing, flag affected recommendations as directional and note the assumption. If only one week of data is provided, report current-state rankings and note that trend detection requires history.

Instructions

Step 1: Normalize and Validate the Data

Before analysis, confirm the data is usable:

  • Identify the grain (SKU, category) and confirm which metrics are present: units, revenue, on-hand, receipts
  • Compute sell-through % = units sold ÷ (units sold + on-hand) for the period
  • Compute weeks of supply (WOS) = on-hand ÷ average weekly unit sales
  • Flag data gaps: missing weeks, zero-inventory items (potential stockouts, not true underperformers), new launches with no history
  • Note any item that stocked out mid-week — its sales are demand-capped and should not be read as a decline

Step 2: Calculate Momentum Signals

For each SKU/category, compute week-over-week and trailing signals:

  • WoW unit change %: this week vs. last week
  • 3-week trend direction: rising / flat / falling based on the slope of the last 3 weeks
  • Velocity vs. category average: is this item selling faster or slower than its peers?
  • Rate of change: is growth accelerating or decelerating (2nd derivative)?

Assign each item a momentum tag: Breakout (accelerating, above-average velocity), Steady (flat, on-plan), Fading (2+ weeks declining), or Stalled (low velocity, high WOS).

Step 3: Classify Into the Action Quadrant

Cross momentum with inventory position to produce the recommended action:

  • Chase — Breakout momentum + low WOS (under 4 weeks): high demand, running out. Reorder now if lead time allows.
  • Ride — Breakout/Steady momentum + healthy WOS: performing, well-stocked. Hold and monitor.
  • Watch — Steady/Fading momentum + moderate WOS: early warning; revisit next week before acting.
  • Clear — Fading/Stalled momentum + high WOS (over 8 weeks): overstocked and slowing. Markdown or promote.
  • Investigate — Anomalies: stockout-capped sales, data gaps, or launches too new to judge.

Step 4: Surface the Emerging Trends

Roll up SKU signals into narrative trends:

  • Rising categories: subcategories or attributes (color, price band, style) gaining share across multiple SKUs
  • Breakout SKUs: individual items outperforming their cohort — name the top 5–8
  • Fading segments: categories losing velocity over 3+ weeks
  • Cannibalization/substitution signals: a new item rising while an adjacent one falls
  • Tie each trend to a hypothesis (why it's happening) and a confidence level (High/Medium/Low based on data depth)

Step 5: Build the Ranked Action List

Produce 8–12 prioritized actions, ranked by business impact (revenue-at-risk or upside × feasibility):

  • Each action names the specific SKU/category, the signal that triggered it, the recommended move (reorder qty range, markdown depth range, expand/hold), and the deadline (driven by lead time or WOS runway)
  • Chase actions must confirm lead time allows the reorder to land before stockout
  • Clear actions must reference margin/WOS to justify markdown depth

Step 6: Set Next-Week Watch List

List 5–8 items in Watch or Investigate status to re-evaluate next week, with the specific signal to monitor (e.g., "if WoW stays negative 2nd week → move to Clear").

Output Format

  • Structure: Executive summary → Trend narrative → Action quadrant table → Ranked action list → Watch list
  • Length: 800–1,200 words plus tables
  • Format: Prose summary and trends; classification and actions in tables; watch list as a bulleted list
  • Sections: Weekly Summary → Emerging Trends → Outperformers → Underperformers → Action Quadrant (table) → Ranked Actions (table) → Next-Week Watch List
  • Tone: Merchandise planning specialist — quantitative, decisive, action-first; every observation ends in a move, not a musing

Quality Checklist

Before delivering output, verify:

  • [ ] Sell-through % and WOS are calculated for every analyzed item or category
  • [ ] Each SKU/category carries a momentum tag AND an action-quadrant classification
  • [ ] Stockout-capped items are flagged as Investigate, not misread as declining
  • [ ] The ranked action list has 8–12 items, each with a specific move and a deadline
  • [ ] Chase recommendations confirm lead time supports the reorder timing
  • [ ] Clear recommendations reference margin or WOS to justify the markdown call
  • [ ] Trend confidence levels reflect the depth of history provided

Notes

  • Trend detection needs 3–4 weeks of history to separate signal from noise; with one week, deliver a current-state ranking and say so explicitly
  • The single most common misread is treating a stockout as an underperformer — always check on-hand before flagging a decline
  • Breakouts with short lead times are the highest-ROI opportunities; a chase that can't land in time is a missed sale, not a reorder
  • When promotions or launches are in the period, isolate their effect before declaring an organic trend

How to use this skill

This file is a Claude Code skill in Anthropic's standard format — markdown with YAML frontmatter at the top.

Installing in Claude Code:

  • ~/.claude/skills/<folder-name>/SKILL.md — auto-loads in every session anywhere on your machine (recommended for skills you'll use across multiple projects)
  • <your-project>/.claude/skills/<folder-name>/SKILL.md — auto-loads only when working in that project (recommended for project-specific skills)

Restart Claude Code or open a new session to register.

Using in Claude.ai, ChatGPT, Gemini, or other AI tools:

  1. Drag this file into a new chat, or copy and paste its contents
  2. Send your task or request as your next message — the AI will apply the skill where it fits

For Claude Code skill details: https://docs.anthropic.com/en/docs/claude-code/skills