Push, Pull, and Postponement

Too much green, not enough blue

Atlas pushed forecast-based production for a spring color palette. Sell-through data showed green hero jackets at 120% of plan but blue at 60%. Markdown on blue ate margin while green stockouts lost DTC revenue. The chain was push-heavy too far upstream with no postponement of final color commitment.

Push systems execute to forecast; pull systems respond to actual demand signals. Postponement delays differentiation (color, label, channel pack) until better information arrives. Apparel is a classic postponement industry if design allows.

Atlas Outdoor Gear is a direct-to-consumer (DTC) and wholesale outdoor apparel brand with global sourcing and the anchor company for OPS 202. Latest annual revenue is approximately $165M across 55% DTC and 45% wholesale, with roughly 2,400 active SKUs and 14-week average production lead time from purchase order release to ex-factory. COO Mei Lin, Logistics Director Carlos Ruiz, and Sourcing VP Priya Shah manage cut-and-sew in Vietnam and Bangladesh, trims in Taiwan, nearshore basics in Guatemala, fulfillment from Reno, Nevada (West DTC and wholesale) and Columbus, Ohio (East wholesale and overflow), and ocean FCL (full container load) from Asia, domestic LTL (less-than-truckload) to wholesale accounts, parcel carriers for DTC.

You met Atlas process fundamentals in OPS 201 (Operations and Process Management) process and capacity work on Atlas fulfillment lines. This course adds the supply network layer: how to design flows from supplier to customer, plan inventory under uncertainty, source ethically at scale, run logistics networks, manage global exposure, and build resilience when ports, weather, or demand surprise you.

Push versus pull boundaries

Push is efficient when demand is stable and setup costs are high (fabric knitting). Pull is efficient when uncertainty is high and lead times are short (replenishing hot SKUs from DC). The push-pull boundary is where forecast-driven production meets demand-driven fulfillment.

Atlas pushes fabric and generic shells; it pulls replenishment to Reno for proven winners via DC-to-DC transfer or expedited PO.

Postponement tactics in apparel

Tactics: delayed dyeing, private label last, channel pack at DC, regional assortment. Atlas tests neutral greige bodies with late color decision for B SKUs while A SKUs commit early for factory capacity.

Postponement requires supplier cooperation and may add unit cost. Mei Lin compares postponement cost to markdown and stockout cost.

Hybrid models and risk

Pure pull rarely works with 14-week Asia lead times. Pure push creates markdown risk. Hybrid models set base push plus pull window for reforecast.

Wholesale pre-books add push pressure; DTC data should feed mid-season pull adjustments.

Managerial signals

Watch forecast error by SKU class, weeks of supply by color, and conversion rate on stockouts. If push error rises, move boundary later or reduce MOQ (minimum order quantity) via supplier deals.

Scenario: COO Mei Lin, Logistics Director Carlos Ruiz, and Sourcing VP Priya Shah must apply push/pull boundary for seasonal jacket this quarter. Wholesale partners want higher fill rates before Q3 pre-fall wholesale bookings and Q4 holiday DTC; DTC marketing is scaling spend on hero jackets; finance caps inventory near $36M at cost.

Part A: Margin impact table

Check: 20,000 blue excess × avg cost $44 ≈ $880K inventory risk before markdown.

Part B: Postponement option

Greige production + late dye for 30% of volume adds $1.20/unit but cuts forecast error cost an estimated $600K on season. Net benefit positive if dye lead adds only 5 days.

Part D: Managerial read

Move push/pull boundary for B/C colors; keep push for A heroes with tighter rolling forecast from DTC weekly.

Wholesale pre-books are a push signal that can stale. Atlas weights DTC velocity at 60% in mid-season reforecast for shared SKUs. Document weights in S&OP to avoid sales vs e-commerce politics.

Atlas pushes 80,000 units at $38 cost. Actual sell-through 70%. Markdown 30% on excess. Compute excess units, markdown cost at retail $95, and compare to postponement premium $0.90/unit on 30% of volume.

Solution

Sold: 56,000; excess: 24,000. Markdown revenue on excess: 24,000 × $95 × 0.7 = $1.596M vs full price $2.28M; loss $684K (simplified). Postponement premium: 24,000 × $0.90 = $21.6K if only postponed portion—full 30% of 80k = 24k units × $0.90 = $21.6K premium vs $684K markdown risk order-of-magnitude. Recommend postponement pilot on B colors.

Key takeaways

• Push/pull boundary should match uncertainty and lead time
• Postponement trades unit cost for forecast error reduction
• Wholesale pre-book and DTC velocity are competing signals
• Measure markdown and stockout together, not separately
• Hybrid models need explicit reforecast cadence in S&OP

1. Re-read the worked examples and verify every check line in your OPS 202 workbook.
2. Apply one concept from Push, Pull, and Postponement to a real SKU or supplier decision at your organization.
3. Preview The Bullwhip Effect and note which Atlas metrics should feed the next analysis.

Applying Push, Pull, and Postponement at Atlas scale

When Atlas Outdoor Gear evaluates push, pull, and postponement, the team starts from operational facts: $165M revenue, 2,400 SKUs, 14-week average factory lead time, and inventory near $36M at cost on the balance sheet. COO Mei Lin, Logistics Director Carlos Ruiz, and Sourcing VP Priya Shah align supply-chain foundations and end-to-end flow design with weekly S&OP cadence, monthly supplier scorecards, and quarterly network reviews. A lesson concept that sounds abstract becomes concrete when tied to purchase order releases, container milestones, and fill-rate dashboards.

Consider how a one-point change in wholesale fill rate affects Atlas. At 45% wholesale mix, a missed key-account delivery can trigger chargebacks and lost floor space for the next season. DTC promises two-day shipping on core sizes; a stockout on hero SKUs shows up in marketing return on ad spend within days. That is why push, pull, and postponement is not an academic exercise for Mei Lin's operations org; it is how the company protects margin while scaling technical shells, midlayers, base layers, packs, and accessories.

The supply-chain foundations and end-to-end flow design workflow at Atlas deliberately separates structural decisions from firefighting. Priya Shah's sourcing team labels supplier risk tiers before PO placement. Carlos Ruiz's logistics team tracks in-transit positions separately from on-hand DC inventory. Mei Lin's S&OP forum forces sales, finance, and operations to reconcile demand plans before factories commit capacity. You should copy that separation habit: name the decision owner, the time horizon, and the metric that proves success before approving spend.

Document definitions alongside every KPI tile. Atlas fill rate specifies eligible lines, cancellation rules, and partial-shipment handling. Inventory turns use average cost inventory and cost of goods sold aligned to fiscal calendar. Lead time clocks start at PO acceptance, not email request. When definitions live in a shared dictionary, the company builds institutional memory instead of re-debating the same report every quarter.

Extended Atlas scenario: cross-functional read

Imagine Atlas's Q3 pre-fall wholesale bookings and Q4 holiday DTC review for push, pull, and postponement. Finance asks whether expedited air freight on delayed containers is worth the margin hit. Merchandising asks whether to cancel a colorway or chase late units for wholesale commitments. IT asks whether a visibility pilot on Tier-1 suppliers should expand before peak. A weak supply-chain foundations and end-to-end flow design answer addresses only one function. A strong answer shows how evidence flows: supplier OTIF (on-time in-full) data explains root cause, inventory simulation quantifies service impact, and network options compare cost versus customer promise.

Work the arithmetic on a conservative example. Suppose Atlas sells roughly $37K at retail value per week across channels. A two-week delay on a container holding $420K at cost on high-velocity fleece SKUs could defer roughly $680K retail sales if substitutes are weak. Expedited split shipment might recover half the lost sales at $95K incremental freight and $18K handling. Mei Lin should compare recovered gross margin to expedite cost, not treat freight as purely operational overhead.

Stakeholder conflict is normal. Priya may push to dual-source a factory to reduce risk. Carlos may resist opening a third DC without volume proof. Wholesale sales may demand 98% fill while finance caps inventory at $36M. Push, Pull, and Postponement gives you language to negotiate those tensions with explicit service-cost tradeoffs rather than charisma. If data is incomplete, the decision is invest in visibility or accept uncertainty, not pretend last year's average lead time still holds.

Translate lessons to your own context by replacing Atlas names while keeping structure. Pick one supply decision you face this quarter. Write the customer promise, supplier constraint, inventory implication, and cash impact before approving a PO or network change. If you cannot write those elements, you are not ready to commit capacity regardless of how urgent the email thread feels.

Technical mechanics and checks (worked patterns)

For push, pull, and postponement, Atlas analysts show work the way finance shows reconciliations. An inventory table prints SKU, on-hand units, average weekly demand, weeks of cover, and a check that extended value equals units times standard cost within rounding. A logistics lane table multiplies transit days, handling days, and order frequency to reconcile total pipeline days to supplier scorecard definitions. A sourcing TCO (total cost of ownership) table sums unit cost, freight, duty, quality fallout, and payment terms into comparable dollars per unit.

Use plain-language decision statements before formulas. Example for safety stock: Atlas targets 96% fill on A SKUs; demand standard deviation over lead time drives buffer size. Still verify seasonality with year-over-year sell-through and document concurrent promotions that could inflate short-term demand. Spreadsheet or ERP replication should state grain first: SKU-location-week for inventory, container-shipment for in-transit, supplier-style for sourcing scorecards.

Common executive questions (and disciplined answers)

Executives ask short questions that require long disciplined answers. "How sure are we on delivery?" maps to OTIF distributions and confidence intervals on lead time, not vendor promises. "What is the dollar impact?" maps to lost margin from stockouts plus expedite cost minus recovery options. "Can we add SKUs?" maps to complexity cost in planning, picking, and supplier minimums. "Why not nearshore everything?" maps to unit economics, capacity, and product quality evidence, not slogans.

Atlas's credible answer format for push, pull, and postponement is three bullets: recommendation, evidence strength (structural data versus anecdote), and next instrumentation step if uncertainty remains. A fourth bullet lists what would falsify the recommendation within sixty days. That discipline prevents the supply chain team from becoming either a bottleneck or a rubber stamp.

Linking Push, Pull, and Postponement to resilience and global exposure

Supply chains fail at interfaces: supplier to factory, factory to port, port to DC, DC to customer. Push, Pull, and Postponement at Atlas must be read alongside global trade and risk lessons later in OPS 202. A sourcing decision that ignores duty exposure or single-port dependence can look efficient on a spreadsheet until a weather event or policy change freezes inventory in transit.

Build a simple interface register for your own organization: node, owner, metric, escalation trigger. Atlas maintains one for Tier-1 cut-and-sew, ocean booking, customs clearance, and wholesale appointment scheduling. When push, pull, and postponement improves one node, update the register and test downstream capacity. Local optimization without system view recreates the bullwhip effect Mei Lin warns about in S&OP.

Practice extension: workbook discipline

Carlos Ruiz requires every supply-chain foundations and end-to-end flow design recommendation to include a one-page workbook tab with four rows: baseline metric, proposed change, reconciliation check, and owner plus review date. Students should mirror that format even when homework uses simplified numbers. The habit trains you to catch unit errors (cartons versus units) and definition drift (calendar days versus business days) before they reach a CFO review.

For push, pull, and postponement, add a fifth row: assumption you would monitor weekly if the recommendation is approved. Atlas examples use in-transit counts, supplier OTIF, DC pick rates, or wholesale cancel rates depending on lesson topic. If you cannot name a weekly monitor, the proposal is not operationalized.

Applying Push, Pull, and Postponement at Atlas scale

When Atlas Outdoor Gear evaluates push, pull, and postponement, the team starts from operational facts: $165M revenue, 2,400 SKUs, 14-week average factory lead time, and inventory near $36M at cost on the balance sheet. COO Mei Lin, Logistics Director Carlos Ruiz, and Sourcing VP Priya Shah align supply-chain foundations and end-to-end flow design with weekly S&OP cadence, monthly supplier scorecards, and quarterly network reviews. A lesson concept that sounds abstract becomes concrete when tied to purchase order releases, container milestones, and fill-rate dashboards.

Consider how a one-point change in wholesale fill rate affects Atlas. At 45% wholesale mix, a missed key-account delivery can trigger chargebacks and lost floor space for the next season. DTC promises two-day shipping on core sizes; a stockout on hero SKUs shows up in marketing return on ad spend within days. That is why push, pull, and postponement is not an academic exercise for Mei Lin's operations org; it is how the company protects margin while scaling technical shells, midlayers, base layers, packs, and accessories.

The supply-chain foundations and end-to-end flow design workflow at Atlas deliberately separates structural decisions from firefighting. Priya Shah's sourcing team labels supplier risk tiers before PO placement. Carlos Ruiz's logistics team tracks in-transit positions separately from on-hand DC inventory. Mei Lin's S&OP forum forces sales, finance, and operations to reconcile demand plans before factories commit capacity. You should copy that separation habit: name the decision owner, the time horizon, and the metric that proves success before approving spend.

Document definitions alongside every KPI tile. Atlas fill rate specifies eligible lines, cancellation rules, and partial-shipment handling. Inventory turns use average cost inventory and cost of goods sold aligned to fiscal calendar. Lead time clocks start at PO acceptance, not email request. When definitions live in a shared dictionary, the company builds institutional memory instead of re-debating the same report every quarter.