Purpose: ABC classifies products by impact so you can make faster, better decisions on stock, fulfilment, and promotions. In the BFCM 1-pager (available to download for free at the end of this article), the A/B/C mix (A%, B%, C%) turns a single orders/day forecast into a realistic split by product importance.
Definitions (practical)
- A items: Small share of SKUs, large share of contribution (e.g., revenue or gross margin). These drive the peak.
- B items: Mid-tier contributors.
- C items: Long tail; low impact individually, useful for breadth.
Tip: Use gross margin contribution if you can (more truthful than revenue or units).
Typical starting thresholds (adjust to your data)
By contribution:
- A: top ~70–80% of contribution (≈10–20% of SKUs)
- B: next ~15–25% of contribution (≈20–30% of SKUs)
- C: remaining ~5–10% (≈50–70% of SKUs)
By mix (for the template defaults):
- A = 60%, B = 25%, C = 15% of orders (adjust per history)
How to derive your A/B/C mix (quick method)
- Choose measure: Prefer Gross Margin £ per SKU over Revenue/Units.
- Data window: Last 6–12 weeks similar to peak (exclude anomalies).
- Clean the data
- Normalise for stockouts (estimate missed sales).
- Mark promo periods to avoid overstating baseline.
- Rank SKUs by chosen measure (descending).
- Cumulative contribution: Assign A/B/C by thresholds above.
- Convert to mix: Calculate the historical share of orders by A/B/C → these are your A/B/C % inputs in the 1-pager.
How it drives decisions in this template
- Forecast split: Orders/Day × (A%, B%, C%) → A/B/C order volumes.
- Capacity & slotting: Wave pick A first; zone fast movers; keep C out of golden locations.
- Safety stock: Higher cover for A, lower for C.
- Channel ring-fence: Reserve A for the most profitable/strategic channels first.
- Promotions: Focus discounts and bundles on A (attach-rate) and B; avoid deep cuts on A below the floor margin.
Example (using the default mix)
If Decision-Ready Forecast = 1,000 orders/day and mix = A 60% / B 25% / C 15%:
- A orders: 600/day
- B orders: 250/day
- C orders: 150/day
Use this to sanity-check picks/hr, carrier caps, on-hand A/B/C, and staffing.
Common pitfalls to avoid
- Using revenue instead of margin → can push low-profit A items.
- Not normalising stockouts → underestimates A.
- Promo distortion → treat promo spikes separately from baseline.
- Static mix → update weekly in peak; brands shift quickly.
- One-size-fits-all → derive mix per channel if behaviours differ (e.g., Marketplace vs DTC).
Maintenance cadence (peak)
- Weekly refresh during BFCM period.
- Re-rank SKUs, re-compute contribution and A/B/C thresholds, and update A/B/C % in the 1-pager.
Optional advanced tweaks
- Dual ABC (Margin × Velocity): Classify on a combined score to avoid “high-revenue/low-profit” A items.
- New product treatment: Seed into B by default, then promote/demote after 2–3 weeks of data.
- Returns-adjusted contribution: Use net margin after returns for categories with high returns.
