The common drift pattern
- Counts live in three places (POS, e-comm, warehouse).
- Store teams do infrequent, manual cycle counts.
- “Phantom stock” creates BOPIS cancellations and walkouts.
- Teams stop trusting on-hand, so they over-order or spend hours recounting.
A practical fix (no rip-and-replace)
- Mobile cycle count app: scan shelf and backroom; flag deltas in real time.
- Reconciliation rules: prioritize the most reliable source by SKU class; auto-open exceptions for big variances.
- Event hooks: whenever items move (transfer, return, BOPIS pick), write back an adjustment.
- Nightly sync: push a “truth set” back to POS/e-comm; mark confidence levels per SKU.
What to measure
On-hand accuracy Target 95%+ for A SKUs.
BOPIS cancel rate Should fall quickly as accuracy rises.
Recount labor hours Should drop week over week.
Out-of-stocks & substitutions Track by store and category.
ROI sketch
If one store does 60 BOPIS orders/day and cancels 6% (3–4 orders), reclaiming even half is meaningful. Multiply by stores × days × AOV and the payback window gets short, even before labor savings on recounts.
Takeaway
Treat inventory accuracy like a data pipeline problem. Scan more often, reconcile automatically, and write back clean numbers. Reliability at the shelf beats a new ERP on a slide.
References
- National Retail Federation — Innovative technologies transforming retail asset protection — Link
- National Retail Federation — Fulfillment and delivery — Link
- GS1 — General Specifications — Link
- GS1 — Guidelines on the use of EPC/RFID — Link
- McKinsey — Future of retail operations: Winning in a digital era — Link
- McKinsey — Deconstructing silos to discover savings: The end-to-end excellence playbook for retailers — Link
- Harvard Business Review — 4 Organizational Design Issues That Most Leaders Misdiagnose — Link
- Harvard Business Review — The Case for Good Jobs — Link