ABC analysis is the Pareto principle applied to your catalog: rank every SKU by its revenue (or margin) contribution, and you will usually find the top slice, the A-items, carries most of the business. The payoff is permission to be deliberately unequal: A-items earn rigorous forecasting and tight monitoring; C-items earn simple rules and benign neglect. Equal attention to unequal SKUs is how small teams burn out.
Running the analysis
- Pull per-SKU revenue for a representative window, 90 days for stable catalogs, a full year if seasonality distorts quarters. Multichannel sellers: combined revenue across every channel, a SKU quiet on your store may be an A-item on Amazon.
- Sort descending, compute cumulative share.
- Cut the classes. Common boundaries: A = SKUs covering the first ~80 percent of revenue (often ~10-20 percent of SKUs), B = the next ~15 percent, C = the remainder. The exact cuts matter less than having them.
- Sanity-check with margin and strategy. A low-revenue SKU that anchors a bundle or brings repeat buyers can deserve honorary-A treatment; rank informs judgment, not replaces it.
What each class earns
A-items: real forecasting, tight reorder points watched continuously, generous safety buffers because their stockouts are the expensive ones, frequent cycle counts, and first claim on your attention when anything looks odd. An A-item stockout during a promotion is the single most avoidable expensive event in small-seller operations.
B-items: standard rules, weighted-average forecasts, moderate buffers, monthly review. The middle class runs on autopilot with periodic supervision.
C-items: minimum viable policy, simple averages, small or no buffers, order in economical batches, count rarely. And ask the honest question annually: which C-items should stop existing? The tail eats storage, attention, and catalog hygiene out of proportion to its revenue.
The multichannel corrections
- Rank on combined data. Per-channel rankings mislead: the pool is shared, so contribution is cross-channel by definition. Unifystock’s channel-mix analytics show exactly where each SKU earns, which both feeds the ranking and tells you where an A-item’s demand actually lives.
- Classes drift, re-rank quarterly. Seasonality, trend, and channel shifts promote and demote SKUs. A stale ranking quietly gives A-item stockouts C-item attention.
- Match protection to class. A-items are where oversell risk costs most: thin-stock A-items selling on multiple channels deserve buffers and real-time sync before anything else in the catalog does.
Common questions
Is ABC analysis worth it under 100 SKUs?
Yes, informally. Even at 40 SKUs, knowing your top 8 changes where buffers, counts, and reorder vigilance go. The spreadsheet takes an hour.
Revenue or profit for the ranking?
Start with revenue (cleanest data), then run a margin-ranked pass, the differences between the two lists are your pricing and sourcing homework.
How do new products fit in?
Unclassified until data arrives; treat launches as provisional B with deliberate buffers, then let the first quarter’s numbers assign the class.
Can this be automated?
The ranking is arithmetic on per-SKU sales data, exactly what a platform computing cross-channel velocity already holds. The judgment layer, strategic exceptions, kill decisions on the tail, stays yours.
Give the A-items what they deserve
Cross-channel sales in one view from $49/month with unlimited orders, and the per-SKU concentration, days-of-cover forecasts and oversell-risk flags that show where the money is on the plans above. See pricing.