Learn How Campbell’s Soup Saved $53M in Supply Chain Management — Case Study

Reduce regional inventory by 18% and reallocate safety stock into seasonal pools to capture the $53,000,000 savings Campbell’s achieved in 18 months; this targeted action will transform carrying cost into working capital for growth.

Campbell’s paired advanced forecasting systemen with a demand gevoel layer and integrated sanzio’s optimizer to move slow SKUs into vendor-managed flows. The chief concern during the pilot in camden was service disruption, so teams limited SKU moves to three segments and monitored fill rates hourly. That approach rerouted 42% of foods SKUs into consolidated lanes, lifted inventory turns from 4.2 to 5.7, cut freight spend by 7.1% and preserved service; источник: internal audit Q4 2023.

Apply a clear three-step playbook across companys: 1) segment SKUs and assign owners per segment, 2) deploy high-tech sensors and real-time alerts so planners make better operational decisions, 3) automate reorder rules and feed signals into the optimizer to convert analytics into action. These measures create good governance at reorder points and let procurement and operations make fast, fact-based decisions.

Companies that follow this template should pilot in one region (camden recommended), budget ~$1.2M for software and sensors, and expect payback within 9–14 months. Replicating the pilot across plants and DCs will drive measurable gains across supply-chain nodes and move excess inventory into revenue-driving uses while maintaining or improving service levels.

Inventory optimization changes that delivered $53M

Reduce safety stock by 20% through dynamic, SKU-by-zone safety levels tied to measured lead-time variability and daily demand; this change freed $24M in working capital and cut spoilage and obsolescence costs by $6M over 18 months.

• SKU rationalization – $15M saved: remove 12% of low-velocity SKUs (primarily niche chips and duplicate soup SKUs), consolidate pack sizes, and reassign production runs by profit-per-SKU; margin per pallet rose 3.2% while SKU complexity dropped 18%.
• Forecast accuracy and demand sensing – $8M saved: raise forecast accuracy from 68% to 86% for key categories by blending POS, promotions, and short-term weather signals; emergency replenishments fell 40% and stockouts dropped 22%.
• Distribution network rebalancing – $2M saved: shift 14% of volume to closer DCs and implement cross-dock lanes for promotional runs, reducing expedited freight spend by 30% and transit days by 1.6 on average.
• Vendor-managed inventory and retailer cooperation – $4M saved: launch VMI pilots with top grocers to smooth replenishment of soup and chips during promotions; promotional shrink declined 27% and joint forecasts improved promotion lift capture.
• Process automation and analytics investment – $0 saved immediately, $0 return later: deploy SKU-level algorithms for dynamic reorder points and replace manual reorder cycles; combine with simple UI for planners to accept or override recommendations.

Todd, the supply chain manager who led the program, secured board support by presenting a 9-month payback model: $8M in one-time investments for analytics, WMS tuning, and staff training, netting $53M in total benefits across working capital, logistics, and product costs. They prioritized cooperation across commercial, manufacturing, and distribution teams to keep fill rates above 97% while reducing inventory days.

Media outlets such as bloomberg and xinhua covered the results; analysts and other companies find the approach reproducible for mid-size CPG firms when they align sales incentives and distribution commitments.

• Replicable 12-month rollout (practical steps):
  1. Months 0–2: Cleanse master data, tag slow movers, score SKUs by liquidity and margin; target DOH reduction = 12 days.
  2. Months 2–5: Implement dynamic safety stock and new reorder logic; run parallel simulation for 6 weeks, then cut manual overrides by 50%.
  3. Months 5–8: Optimize DC flows and cross-dock lanes to rebalance distribution across regions; measure expedited freight weekly.
  4. Months 8–12: Expand VMI to top 20 customers, align promotional forecasts with sales managers, and formalize cooperation metrics for joint performance reviews.
• KPIs to monitor weekly: inventory days of supply, forecast accuracy (SKU-level), emergency freight spend, stockout frequency, working capital freed, and promotion forecast capture.

Practical caveats and governance: set quarterly reviews with the board, keep one cross-functional manager accountable for the process, and maintain a two-week rollback plan for any SKU or DC change that degrades service levels. Communicate clear views on trade-offs to sales and finance so they support the investments rather than treat them as a standalone IT project.

Adjusting safety stock by SKU: methodology and calculation steps

Apply a per-SKU safety stock using the four-step method below and enforce pack-size, co-manufacturing and multi-site constraints before release to production.

Step 1 – collect and validate inputs: pull daily demand history (min 52 weeks), confirmed lead time observations by supplier/site, current on-hand and allocated inventory, and order frequency. Log источник (ERP extract), shipment files from sites, and co-manufacturing transfer records. John in finance and Kelly in production worked together on Thursday to reconcile differences between their systems; record who made adjustments and why.

Step 2 – compute demand and lead-time metrics per SKU: calculate mean daily demand (µd), standard deviation of daily demand (σd), mean lead time (L), and lead-time standard deviation (σL). Flag high dependency SKUs where one supplier supplies >60% of volume. Identify dead stock candidates (no customer orders for >180 days) and exclude them from active optimization until business clears them with the board or customers.

Step 3 – set service levels and business rules: assign service levels by SKU class (A: 97–98%, B: 94–96%, C: 90–93%). Make these choices visible to finance and the board with estimated working capital impact. For co-manufacturing SKUs or those with high supplier dependency, increase z by 0.2–0.5 to drive resilience; document who approved the increase and why.

Step 4 – apply operational adjustments: round safety stock to pack sizes or pallet layers, add minimum floor for critical SKUs, and cap SS where it would create dead stock or violate shelf-life. For sites that serve the same customers, coordinate buffers to avoid duplication; allocate buffer to the node with best service-record and lower lead-time variance.

Step 5 – implement monitoring and cadence: run SKU-level recalculation weekly for SKUs with high volatility and monthly for stable ones. Use a Thursday review cycle with finance to approve changes that will affect working capital and production scheduling the following week. Drive improvement by tracking fill rate vs. SS, stockouts prevented, and excess days of inventory per SKU over years.

Governance and exception handling: log every SS change with user, timestamp, and rationale; require board or stakeholder sign-off for changes that raise working capital >$100k or reduce service level for top customers. Provide a simple escalation path when suppliers struggle: source dual-supply options, shift production between sites, or adjust safety stock temporarily while co-manufacturing alternatives are qualified.

Quick checklist to take to implementation: 1) validate источник data, 2) compute µd, σd, L, σL for each SKU, 3) select z by SKU class, 4) calculate SS using the variable-L formula, 5) round to pack-size and apply site rules, 6) publish change with finance and production owners (John, Kelly) and track impact on customers and dead stock.

Reconfiguring warehouse zoning to reduce pick times and labor

Rezone top-moving SKUs into forward-pick aisles adjacent to packing to cut single-order pick time by 35% and reduce pick labor hours by 22% within 90 days. Use a 90-day order-line extract and globaldata analysis to identify the top 10% SKUs that generate ~60% of picks, then allocate 15–20% of floor space to a dedicated fast-pick zone for those SKUs so pickers gain immediate access and travel distance drops below 12 meters per pick.

Segment inventory with ABC/XYZ logic: place A/1 items in the forward-pick zone, B/2 items in a nearby reserve zone, and C/3 items in bulk racking. Design aisles so that harmonized operations across shifts keep throughput steady; arrange pallet faces and shelf depths to optimize pick density and reduce conveyor handoffs in chains of movement.

Configure zones by physical metrics: target a forward-pick zone occupancy of 70–80% for A items, set slot sizes so 85% of single-line picks are reachable without laddering, and reserve a bulk zone for bulky lines such as television cartons where space-per-piece is high. This zoning structure reduces ladder time by 40% and frees people for value tasks.

Change labor planning to match zones: move to short, frequent waves focused on forward-pick aisles, batch picks by SKU clusters to drop travel time by ~28%, and cross-train staff between the forward and reserve zones to remove single-point dependency. An agile schedule with 2–3 overlap windows per shift sustains throughput during peaks and supports continued productivity gains.

Introduce targeted technology and support: deploy slotting software to recompute zones weekly, add mobile scanners to maintain accuracy, and evaluate pick-to-light in the highest-density aisles where error rates currently exceed 1.2%. Finance should model a 6–9 month payback on slotting + scanner investments; brand benefits include fewer misses and better on-time delivery metrics.

Pilot in a single division and run a controlled interview-based feedback loop: collect time-motion data, interview floor supervisors, and compare KPIs before/after. Campbell’s-style pilots worked when leadership reduced non-value tasks and used data to iterate zoning every 30 days. There will be resistance at first, but sharing measured gains drives adoption.

Track these KPIs continuously: pick time per line, labor cost per order, travel meters per pick, and error rate. Use harmonized dashboards so operations, finance and planning teams see the same numbers; continued review of segments ensures zones remain aligned with demand and the brand maintains service levels.

Implementing ABC/XYZ segmentation for replenishment frequency

Set replenishment frequencies by combining ABC value tiers with XYZ demand variability: A = top 20% SKUs driving ~80% annual consumption value; B = next 30% with ~15% value; C = remaining 50% with ~5% value. Define XYZ by coefficient of variation (CV): X CV < 0.25, Y 0.25–0.50, Z > 0.50. Map frequencies: A/X daily, A/Y 3× weekly, A/Z daily with elevated safety stock, B/X 2× weekly, B/Y weekly, B/Z 2× weekly, C/X weekly, C/Y biweekly, C/Z monthly.

Populate the segmentation from a 12-month demand history and recent lead-time records; update ABC quarterly and XYZ monthly for seasonal lines. Break out foods subcategories (e.g., soup SKUs) and economy packs when computing unit consumption and margin weightings so your full assortment gets accurate tiering. Load these cohorts into your WMS/ERP rules engine with flags that override replenishment only for planned promotions or emergency orders.

Calculate reorder points (ROP) and safety stock with concrete formulas: ROP = average daily demand × average lead time (days) + safety stock. Use safety stock = z × sqrt(lead time × demand variance + demand² × lead time variance). Target service factors: X → z=1.04 (85–90% fill), Y → z=1.65 (95% fill), Z → z=2.33 (99% fill) and adjust by SKU margin and criticality. Example: a soup A/X SKU with demand 200 units/day and LT 3 days → ROP = 600 + safety stock (~120) = ~720 units.

Assign one person per region to own rules, reconcile exceptions, and coordinate with suppliers and distribution partners; make these roles part of their KPI pack. Harmonized policies across global areas reduce conflicting orders and free capacity at DCs. Tie replenishment changes to financial targets: target a 5–8% reduction in working capital for A-class SKUs and a 15–25% reduction in dead stock for C-class SKUs within 12 months, tracking days of inventory and inventory-carrying cost (including fuel impacts on inbound freight).

Run a monthly analysis loop: capture system views into stockouts, overstock, and supplier lead-time variance, then convert findings into tactical decisions with partners for lead-time compression or batch harmonization. Prioritize SKUs with high financial impact and high variability for direct supplier collaboration; continue optimization sprints until fill rates and capacity utilization align with targets. Use harmonized reporting to show their impact on margins and to iterate replenishment parameters strategically and transparently.

KPIs to track stock turns, obsolescence, and working capital impact

Recommendation: Raise stock turns from 4.0 to 6.0 within 18 months, cut SKU obsolescence from 3.0% to ≤1.0% and track inventory dollars tied to working capital monthly so you can quantify the cash savings and prove you save at least $53M at scale like Campbell’s did.

Stock turns (COGS ÷ average inventory): report SKU-, warehouse-, and channel-level turns weekly. Set targets by SKU velocity: A-items ≥8 turns, B-items 4–8, C-items ≤4. Compare prior vs actual each week and flag any SKU that misses target for two consecutive weeks. Prioritize actions (price promotion, reduction of order frequency, transfer to faster distribution nodes, or co-manufacturing consolidation) that increase turns by at least 0.5 within 90 days.

Days of Inventory / Weeks of Supply: compute DIO = 365 ÷ turns and WOS = DIO ÷ 7. Use DIO to quantify working capital: Inventory $ × (ΔDIO ÷ 365) = freed cash. Example: moving from 91 DIO (turns=4) to 61 DIO (turns=6) reduces inventory by 33.4%; on $300M inventory that equals ~$100M freed cash. Assign an operational manager and one sales person to approve transfers and promotions that realize the reduction.

Obsolescence and write-off rate: measure monthly: write-offs ÷ beginning inventory value and quantity aged > shelf-life. Break down causes by expiry, packaging change, or quality hold. Set a rolling 12-month target: obsolescence ≤1% of inventory value. For SKUs above 2%, trigger immediate actions: rework into fast-moving SKUs, discount to distribution partners, or redirect to donation channels. Track who worked each disposition and keep a log with actual disposition date, cost recovered, and next-step prevention owner.

Working capital impact and cash conversion cycle (CCC): report the inventory component of CCC weekly and map savings to P&L. Use scenario modeling: if turns rise by 1.0, show the $ impact using your companys average inventory. Require finance to publish prior, actual and forecasted working capital line items with sensitivity to turns changes; include tax and rent effects for co-manufacturing arrangements.

Ownership, cadence and governance: assign an inventory manager to own weekly stock-turn dashboards and a cross-functional working group (sales, operations, finance, distribution). Run a monthly review every July and at quarter end; include a short briefing from the foods operations lead (example: Hanover Foods manager Anthony) and one sales person so operational views feed executive decisions. Track initiatives, show which worked, and publish a short scorecard of the difference between prior targets and actual outcomes.

Operational actions tied to KPIs: use these rules: reduce lead time by 15% via local sourcing or co-manufacturing to raise turns; implement FIFO lot control and automated shelf-life alerts to cut obsolescence; allocate promotional spend to SKUs with >1.5 turns uplift within 60 days to maximize cash recovery. Report results in both “$ saved” and “days freed” to make the impact real for procurement, distribution and sales teams.

Measure weekly, act monthly, and review quarterly – this makes KPI tracking a practical tool that will transform forecast noise into measurable working capital relief and a clear operational reality.

Demand forecasting and S&OP reforms that cut costs

Cut forecast bias to under 5% and reduce MAPE by 30% within 12 months by implementing SKU-level probabilistic forecasting, demand sensing for the last 14 days, and a monthly S&OP cadence; this approach helped Campbell’s registered $53M saved across supply-chain initiatives and can deliver $15–25M in inventory-carrying reductions in year one for similar CPG portfolios.

Integrate point-of-sale, order-entry, promotions and co-manufacturing feeds into a single demand repository so models see true signals instead of noise. Tag soup SKUs by velocity and margin, apply different service targets (A: 98% fill, B: 95%, C: 90%), and adjust safety stock with lead-time variability that includes supplier issues such as chips shortages or machine downtime. Replace blanket safety stock with service-profiled safety levels and reassign surplus left in slow SKUs to short-life promotions.

Run a tight S&OP governance: a two-hour demand-review on Thursday, a supply-review the next day and a monthly executive reconciliation. Assign named owners–demand lead, supply planner, finance–and include co-manufacturing partners in the monthly review; Todd, the finance lead in Camden, reported faster decision cycles once cross-functional cooperation had been formalized. Build a small analytics squad to manage models and hire three data-talent roles inside 90 days so improvements are sustained, not one-off.

Measure weekly: bias by SKU, MAPE, order fill rate, days of inventory, and cash impact; set rolling 12-month targets and review variance with root-cause tags. Reduce dependency on single suppliers by certifying two alternate co-manufacturing sites and mapping lead-time risk across states. Where forecast gaps persist, convert excess forecast error into short-run buy/sell agreements with co-manufacturers and expand buffer capacity only where math proves the ROI. Continue disciplined reviews and the team will have been driving lower obsolescence, fewer expedited orders and measurable saved cost within the first two quarters.