Everything is a
flow.
A can of detergent on a Walmart shelf is the visible end of a vast, invisible network — product, information, and money moving in both directions. This chapter is the map. Let's route through it.
What a supply chain actually is
Not a chain. A web.
A supply chain is every party involved, directly or indirectly, in filling a customer request — not just the manufacturer and suppliers, but transporters, warehouses, retailers, and the customer too.
Walk into Walmart for detergent and the chain is already humming: the store is stocked from a distribution center, which is stocked by the manufacturer (P&G), which buys raw materials and packaging (from firms like Pactiv) — which themselves have lower-tier suppliers. The customer isn't outside this system. The customer starts it.
The word "chain" is misleading. A maker takes material from many suppliers and ships to many distributors, so most chains are really networks — a supply web. And three things flow through it, in both directions: product, information, and funds.
The flows are now mostly digital and instant. When you tap "buy" on Amazon, point-of-sale data races upstream through ~1,200+ logistics facilities worldwide, and inside the warehouse over 1 million robots now help fulfil roughly three-quarters of US orders — the same product/information/funds flows the textbook describes, just automated end to end.
Where it came from
Three rivers, one delta.
Modern supply chain management is the confluence of three streams of practice that evolved separately over ~60 years, then fused under the pressure of intense competition.
Grew out of purchasing. Cutting procurement price boosts profit far more easily than chasing equivalent extra sales. Today: sourcing, supplier relationships, inbound logistics.
Forecasting, inventory, warehousing, scheduling → production planning. Materials are ~60% of most products' cost, so managing their flow became the route to cost reduction.
From military origins: moving goods to the right place at the right time. Transport is up to 50% of logistics cost. Drucker called distribution "today's frontier."
That "early avatar" framing looks prophetic. The 2020–24 era — COVID, the Ever Given blocking the Suez for six days (≈$10bn/day of trade stalled), Red Sea attacks cutting Suez traffic ~75%, a Panama Canal drought slashing capacity ~36% — turned logistics from a back-office function into a boardroom and front-page topic, exactly the integrative, strategic discipline this section anticipated.
The one equation
Grow the pie, don't fight over slices.
Every supply chain has one objective: maximise the total value it generates — the supply chain surplus.
Buy a wireless router for $60 at Best Buy. That $60 is the only money entering the chain — the customer is the single source of revenue. Everything else (Best Buy paying suppliers, suppliers paying makers) is just funds moving around inside the chain. The gap between the $60 and the total cost to produce and deliver the router is supply chain profitability — the profit shared across all stages.
Measure success by total chain profitability, not the profit of any single stage. Optimising one stage in isolation can shrink the whole pie. Grow the surplus and every member can benefit.
The consumer-surplus engine the chapter describes is now a national export story. India's pharmaceutical exports reached about $30.5bn in FY2024–25 — up from the ~$11bn the textbook cites — supplying roughly a fifth of the world's generics and nearly half of all US generic prescriptions by volume. The price gap between a patented drug and its generic equivalent is the surplus, and it scaled into an industry.
Win or die by the chain
Five companies. The chain decided their fate.
Supply chain design, planning, and operation directly shape whether a firm survives. Tap each to see how its chain made or broke it.
Invested early in transport and information infrastructure. Built clusters of stores around distribution centers for cheap, frequent replenishment, and pioneered information-sharing with suppliers. Result: matching supply to demand better than rivals.
The textbook's figure: ~$17bn net income on ~$469bn revenue (2013), from just $1bn of sales in 1980 — a 20%+ compounded growth rate.
A very responsive replenishment system plus an outstanding information system meant products were available exactly when and where customers wanted them — even changing the merchandise mix by time of day to match demand.
Grew from ¥1bn (1974) to ~¥1.9 trillion (2013), ¥222bn profit.
Big warehouses in major cities, delivering groceries to homes. Traditional supermarkets bring full truckloads to a store near you and let you do the picking — very low cost. Webvan turned inventory only marginally faster but bore high transport + labour (picking) costs.
Folded in 2001, two years after a successful IPO.
Dominated the 1990s with the superstore concept — ~100,000 titles vs <10,000 at local shops, with higher inventory turns. Then Amazon offered even more variety at lower cost from a few DCs, selling online. Borders couldn't adapt its chain.
Sales fell from ~$4bn (2004, $132m profit) to $2.8bn by 2009 with a $109m loss.
1993–2006: spectacular growth from selling direct and postponing final assembly until the order arrived — huge variety, low component inventory. By 2006: $3.5bn+ net income on $56bn+ revenue.
Then the market shifted to less customization. Dell adapted — selling through Walmart and GOME, outsourcing assembly, building to stock. Unlike Borders, it kept reshaping the chain.
Design, planning, and operation play a significant role in a firm's success or failure. To stay competitive, supply chains must adapt to changing technology and customer expectations.
The adapt-or-die pattern keeps running. Dell reinvented itself again, riding the AI-server wave to revenues around $95–113bn with infrastructure-group sales surging on AI demand. Seven & i Holdings (Seven-Eleven's parent) went through a major 2025 restructuring, spinning off weaker units to defend its convenience-store core. The chain you build is never "done."
Three phases of every decision
Years, then months, then minutes.
Chain decisions sort into three phases by how often they're made and how long their impact lasts. Each phase sets the constraints for the next.
Design decisions constrain or enable good planning, which in turn constrains or enables effective operation. They're a hierarchy, not a list.
Two lenses on the same chain
Cycles for operators. Push/pull for designers.
There are two ways to view chain processes. The cycle view groups them into cycles between stages — good for operations. The push/pull view splits them by timing relative to the customer order — good for strategy.
The cycle view: four handoffs
The push/pull view: drag the boundary
Push = made in anticipation of orders (a forecast — speculative). Pull = made in response to a real order (reactive). The push/pull boundary is where forecast hands off to actual demand. Move it and you trade inventory against responsiveness:
Shein pushed the boundary further toward pull than the textbook's examples imagined. It makes tiny test batches, watches what sells online, then produces winners on demand — reportedly launching on the order of 1.5 million products in a year (dwarfing Zara's volumes). Almost pure pull. The flip side — overconsumption and waste — prompted France to pass a per-item ultra-fast-fashion tax in June 2025, showing the boundary choice now carries regulatory weight too.
Three macro processes
Supplier-side, inside, customer-side.
Zoom out and every chain process inside a firm falls into one of three macro processes — defined by which interface they sit at.
- source
- negotiate
- buy
- design collab
- supply collab
- strategic planning
- demand planning
- supply planning
- fulfilment
- field service
- market
- price
- sell
- call center
- order mgmt
All three serve the same customer, so integration is crucial. The classic failure: marketing owns CRM, manufacturing owns ISCM, purchasing owns SRM — with little communication, and even different forecasts. That mismatch hurts supply-demand matching, frustrates customers, and raises cost. Structure the org to mirror the macro processes.
The case files
Eight real chains, and the questions they force.
The chapter closes with worked example chains. Each raises the strategic, planning, and operational questions you'd have to answer to design it. Tap to open the file — the original discussion questions are kept verbatim.
Gateway opened stores in the late 1990s that carried no finished-goods inventory — all PCs built to order and shipped. Stock peaked above $80 (1999), fell below $4 (2002); all retail closed by 2004; bought by Acer for $710m in 2007. Apple, opening stores from 2001, always carried inventory with limited variety — by 2013, 415+ stores, $20bn+ sales, $51.5m average revenue per store.
- Why did Gateway carry no finished-goods inventory while Apple did?
- Should a firm with retail stores carry finished-goods inventory? Which products suit inventory vs. build-to-order?
- How does product variety affect required store inventory?
- Is direct selling without stores always cheaper than selling with stores?
- What explains Apple's retail success and Gateway's failure?
2012: ~€16bn sales, 6,000+ stores in ~86 countries. Cut the design-to-sales cycle from the industry's 6+ months to 4–6 weeks, refreshing 75% of displays every 3–4 weeks. Uncertain-demand items made in Europe (Portugal/Spain), predictable ones in Asia; ~40% capacity in-house. 40%+ of finished-goods purchases happen after the season starts (vs <20% typical). Eight Spanish DCs, 24–48h delivery.
- What advantage does a very responsive chain give Zara?
- Why both in-house and outsourced manufacturing? Why keep European capacity despite cheaper Asia?
- Why source uncertain-demand items locally and predictable ones from Asia?
- What's the advantage of replenishing stores multiple times a week?
- Is Zara's responsive model better suited to online or retail sales?
Inditex hit ~€38.6bn sales and €5.9bn net income in 2024 — more than double its textbook figure. But the "fast" crown moved on: Shein and Temu now out-iterate even Zara, and European fast-fashion has slowed enough that those rivals are taking share.
Both sell maintenance-repair-operations supplies via catalog and web; neither manufactures. Grainger: several hundred US stores, 9 DCs, ~300,000 SKUs stocked. McMaster-Carr: ships almost everything, 5 DCs, ~500,000 SKUs. Pure distribution success built on supply chain skill.
- How many DCs, and where?
- Should all DCs carry all products?
- What to stock vs. ship direct from the supplier?
- What should Grainger carry in a store?
- How to allocate markets to DCs? What if an order can't be fully filled — backup locations?
Strategy: a factory in every market served. The core question is each plant's capability — local-only at one extreme, able-to-supply-everywhere at the other. After the 1996–97 Asian crisis, Toyota adopted "global complementation" so strong markets could export to weak ones. Parts commonality cut costs but bit back: a 2009 recall of ~12 million cars over common parts hurt brand and finances.
- Where to locate plants, with what flexibility and capacity?
- Produce for all markets, or only specific contingency markets?
- How to allocate markets to plants, and how often to revise?
- How should the investment in flexibility be valued?
Began filling orders from a distributor's books; added warehouses to react faster. By 2013: ~40 US + ~40 international warehouses; 2012 outbound shipping cost $5bn+. Pushed digital (Kindle), then expanded categories — acquiring Zappos in 2009 (huge variety: 121,000 product descriptions, 2.2m+ images) and diapers.com in 2010 (little variety, big shipping volume).
- Why build more warehouses as it grows? How many, and where?
- Should Amazon stock every product it sells?
- What can bricks-and-mortar players gain from an online channel?
- Online channel advantages/disadvantages for shoes vs. diapers?
- For which products does online beat retail — and what characterises them?
The "~40 + 40 warehouses" of 2013 is now 1,200+ logistics facilities and 1M+ robots. Amazon answered its own "how many warehouses, where" question by pushing inventory ever closer to customers for same/next-day delivery.
A seamless online-and-store experience: browse online, see it in store, or order online after seeing it. Crucially it's about fulfilment too — any channel can reach the whole assortment. By 2012, 292 stores could fulfil online orders or cover stockouts at other stores; buy-online-pickup-in-store and in-store returns of online buys.
- Fill online orders from stores or fulfilment centers?
- How to manage store inventory in an omni-channel setting?
- Keep returns at the store or send to a fulfilment center?
An indigenous model for India's fragmented, small-scale sector. Modeled on GCMMF (Amul), Gopaljee pools the tiny output of small dairy farmers into organised scale. From a 1989 first plant to ~1 million litres/day, ~125,000 farmers, 4,500 village societies. Grows by direct procurement — eliminating middlemen — trains farmers, uses UHT tech, and has expanded overseas to SE Asia-Pacific, the Middle East and Europe.
- How can the model extend to other Indian/South-Asian business lines?
- Which socio-economic features of South-Asian societies support sustainable chains?
- How can existing distribution channels be transformed to maximise customer value?
- How can indigenous SCM models integrate with global supply chains?
A medium-scale rug/carpet maker that links remote rural artisans to world markets, weaving corporate social responsibility into the chain's design. Sourcing/production spans eight Indian states; markets include the USA, Australia, Canada, Greece, Germany, Spain — product moving from thousands of village looms through leading retail chains to global buyers.
- How can SMEs in the developing world use chains to deliver value and raise artisans' living standards?
- How can middlemen be minimised to free artisans from exploitation and underpayment?
The case study's core claim — that consumer surplus drives chain performance — held up dramatically. Generic Gleevec from Indian makers sold for a fraction of Novartis' price; that surplus helped build an industry now exporting ~$30bn+/year, on which global health bodies (MSF, the Global Fund, PEPFAR, UNICEF) heavily rely. India still supplies roughly a fifth of the world's generics.
You've mapped the whole chain.
Here's the chapter in eight ideas. Tap any unit in the bar logic below to jump back.