On this page
concept

Theory of Constraints

Created 2026-10-25 22 connections

Theory of Constraints

Developed by Eliyahu Goldratt in his 1984 book The Goal, the Theory of Constraints (ToC) holds that every system has at least one constraint limiting its overall throughput, and that improving non-constraint areas does not improve system-level performance. In ecommerce and retail operations, ToC provides a systematic method for identifying and resolving bottlenecks — in warehousing, fulfilment, inventory, or market demand — that cap revenue growth regardless of how efficiently other parts of the operation run.

Core framework

The Five Focusing Steps

ToC's central improvement loop consists of five steps that repeat continuously (fourweekmba.com, April 2026):

  1. Identify the constraint — the single resource, policy, or market factor limiting system throughput
  2. Exploit the constraint — extract maximum output without additional investment (scheduling, buffering, full kitting)
  3. Subordinate everything else to the constraint — all other parts of the system pace themselves to the constraint's output, not their own local efficiency metrics
  4. Elevate the constraint — if steps 2–3 do not break it, invest to increase its capacity
  5. Return to step 1 — once the constraint is broken, find the next one; avoid inertia

Goldratt also framed three change-management questions: "What needs to change? What should it change to? How do we bring about the change?" (fourweekmba.com, April 2026).

Types of constraints

TypeDescriptionEcommerce examples
PhysicalMachine, space, staffPicking station, warehouse capacity, carrier slots
PolicyCompany procedures, regulation, contractsBatch replenishment rules, return SLA policies
ParadigmDeeply held beliefs / habits"We must carry high stock"; local efficiency measurement
MarketDemand limits outputCustomer acquisition cap; seasonal concentration

Policy constraints are often the second most common bottleneck after physical constraints (fourweekmba.com, April 2026). Splunk (2025-04-24) adds organisational and governance constraints — HR decisions, internal governance — as a further category.

What is the primary retail constraint? Goldratt's retail application in Isn't It Obvious? (2009, summarised shortform.com) identifies the constraint as customer flow — a market constraint. Splunk (2025-04-24) identifies warehouse space as the prototypical ecommerce constraint — a physical constraint. These reflect different starting conditions (established vs. capacity-constrained retailers) and are not mutually exclusive.

Drum-Buffer-Rope (DBR)

DBR is ToC's scheduling mechanism for production and fulfilment (Splunk, 2025-04-24; Logistics Insider, 2026-01-19):

  • Drum — the constraint sets the pace for the whole system
  • Buffer — a time buffer upstream protects the drum from starvation
  • Rope — a communication signal synchronises upstream processes to the drum's rate, preventing overproduction ahead of the bottleneck

In multi-tier supply chains, DBR reduces Bullwhip Effect|bullwhip effects and improves collective performance; reported qualitative outcomes include lower WIP, shorter lead times, and better OTIF delivery (Logistics Insider, 2026-01-19).

Throughput Accounting

ToC replaces traditional cost-accounting with three financial measures (Splunk, 2025-04-24; practicalecommerce.com, undated):

MetricDefinition
Throughput (T)Rate at which the system generates money through sales, net of truly variable costs
Inventory (I)Money invested in assets intended for sale
Operating Expense (OE)Money spent to convert inventory into throughput

All improvement decisions are evaluated by their effect on profit per bottleneck-minute, not resource utilisation. Since expense reduction is bounded at zero while revenue growth is not, ToC prioritises increasing T over reducing OE (fourweekmba.com, April 2026).

Throughput definition: practicalecommerce.com describes throughput as "typically sales or gross margin, depending on the business." fourweekmba.com defines it strictly as "rate of sales revenue." Goldratt's original definition is net revenue after truly variable costs — closer to gross margin. Sources apply the term differently.

Supply chain ToC uses Throughput Dollar Days (TDDs) and Inventory Dollar Days (IDDs) as recommended control measures (Hokey Min & Bih-Ru Lea, IJLSM peer-reviewed, 2025).

Ecommerce and warehouse applications

Warehouse bottleneck identification and exploitation

Picking and packing is the most common physical constraint in ecommerce fulfilment warehouses. Goldratt UK (2024-11-28) documents three primary exploitation tactics:

  1. Full kitting — give pickers all order information, packaging, and inserts before they start; eliminates interruption at the constraint
  2. Timed sprint batching — batch orders into 30–60 minute sprints to create rhythm and measurable throughput at the constraint
  3. Fast-stock proximity — place the top 20% of SKUs (by volume) near the packing station, exploiting the Pareto distribution (~80% of sales from ~20% of SKUs)

Non-constraint ("sand") tasks — stocktaking, replenishment, goods-in, tidying — should be scheduled outside peak picking periods to protect the constraint from competition for labour (Goldratt UK, 2024-11-28).

Named warehouse case studies (as-of 2024-11-28)

All three are vendor-reported by Goldratt UK; no independent verification methodology is described.

CompanyMetric beforeMetric afterTimeframe
Birkdalebaseline not stated50% increase in picks/hr; 65% increase in avg lines picked~6 weeks
Autoline25–30 lines/hr pick; 15–20 lines/hr pack40–50 lines/hr pick; 25–30 lines/hr packnot stated
Fixmart17 items/hr (day + night)24 items/hr (day + night)not stated

eCommerce fulfilment: overstockArt.com case

David Sasson (co-founder, overstockArt.com) applied ToC to order fulfilment after a year of unsatisfying growth. The company identified packaging as its constraint — it sat at the end of the fulfilment line and controlled overall throughput. Exploitation: built a WIP buffer in front of the packaging station; educated management that packaging output (not per-station productivity) defined true throughput. Elevation: redesigned packaging to protect artwork and speed the process.

Reported outcome: throughput increased 50% with no increase in operating expense (practicalecommerce.com, 2014-09-27).

overstockArt.com case dates from 2014. No 2025–2026 ecommerce-specific case studies with named retailers and quantified outcomes were found in this run. The only ecommerce-specific quantified case in the literature remains pre-2024.

overstockArt.com also shifted from measuring productivity per station to measuring total production output, and chose full-time skilled staff over part-time workers to protect the constraint from starvation (practicalecommerce.com, 2014).

Retail inventory: Goldratt's Isn't It Obvious? principles

In Goldratt's retail application (2009, summarised shortform.com):

  • Both stockouts and excess inventory reduce a retailer's ability to exploit its constraint: stockouts cause lost sales; excess inventory ties up resources, clutters shelves, and reduces revenue
  • A demand-pull replenishment model triggered by actual purchases (rather than projected forecasts) replaces push-based inventory systems
  • Faster replenishment from regional distribution hubs to stores is the key mechanism to reduce stockouts and lower total stock levels simultaneously
  • Bulk buying for cost efficiency creates harmful inventory imbalances — surpluses of some SKUs alongside shortages of others — resulting in lost sales, markdowns, and unhappy customers
  • Integrated strategy across purchasing, distribution, and inventory management is essential; local-metric optimisation by each department creates system-level suboptimisation

Store managers commonly resist reduced inventory under ToC implementation, preferring large stock levels to cover demand variability (shortform.com, undated — 2009 Goldratt book).

Market constraint and demand management

When internal constraints are resolved, the constraint typically shifts to the market, and demand management (DM) becomes central (de Souza & Ikeziri, IJLSM, 2025). The DM process under ToC:

  1. Identify a significant market need
  2. Build a decisive competitive edge through operational excellence
  3. Capitalise on that edge
  4. Sustain it through synchronisation of operations, sales, and marketing

Key ToC demand management constructs (IJLSM peer-reviewed, 2025): Strategy and Tactic (S&T) tree; Unrefusable Offer (URO); Simplified Drum-Buffer-Rope (S-DBR); Make-to-Availability (MTA); Dynamic Buffer Management (DBM).

Internal vs. market constraint primacy: IJLSM 2025 (de Souza/Ikeziri) states that once internal constraints are resolved, the constraint shifts to the market and demand management becomes primary. Logistics Insider 2026 positions supply chain capacity, supplier lead times, and logistics as the primary focus — implying the market-constraint framing is not yet mainstream in practitioner discourse.

DDMRP: ToC applied to inventory planning

DDMRP (Demand-Driven Material Requirements Planning) integrates elements of Lean, ToC, and traditional MRP. A 2025 MDPI Mathematics peer-reviewed simulation study (as-of 2025-10-31) found:

  • A component buffer (COB) positioned after the bottleneck station reduces finished goods inventory by approximately 30% compared to a finished-goods-buffer-only configuration
  • Achieving 98% service level requires ~3.5 units per product with COB vs ~5 units without
  • COB is most advantageous when finished goods are costly to hold, product differentiation can be postponed, or components are shared across multiple finished products

Earlier Deloitte Denmark research (2022, single Danish beverage manufacturer, 6 products) found DDMRP reduced average inventory by ~42% vs MRP at 100% service level, though inventory variation increased ~13–19%.

Deloitte Denmark study (2022) is pre-2024 and potentially superseded. Deloitte is also a DDMRP consulting vendor — not independent. The 2025 MDPI peer-reviewed simulation provides more methodologically transparent validation.

DDMRP inventory variation tradeoff: The Deloitte 2022 study and the MDPI 2025 simulation both confirm DDMRP reduces average inventory, but the Deloitte study flags a ~13–19% increase in inventory variation as a downside. This nuance is absent from Sikich (2025) and Meegle (2024) practitioner articles that cite the 42% reduction without the variation caveat.

The Demand Driven Operating Model (DDOM) generates and schedules supply orders based on actual demand and executes using strategic decoupling points plus buffers of stock, time, and capacity — creating a stable yet agile supply system (TOCICO Innovation Summit 2024, 2024-09-29).

Interaction with Lean and adjacent methods

The Goldratt Group positions ToC and Lean Manufacturing|Lean as complementary: Lean removes waste; ToC ensures the waste removal is directed at the constraint (toc-goldratt.com, undated).

ToC vs Lean hierarchy: The Goldratt Group book positions them as equal partners. FourWeekMBA (April 2026) and Umbrex (undated) position Lean as subordinate to ToC — improvements should be directed at the constraint, not distributed evenly — reflecting Goldratt's original view that undirected Lean wastes effort on non-bottleneck steps.

Supply chain visibility context (2026)

Fluent Commerce survey of 100 retailers at Retail Technology Show 2026 (as-of 2026-08-13, vendor survey, small non-random sample):

  • 70% reported supply chain disruption in the past year; 19% described it as severe
  • Main disruption impacts: higher costs from rerouting orders (35%), products out of stock online (31%), price increases (30%)
  • 41% had real-time inventory visibility across their supply chain (up from 28% in 2025)
  • 50% still say poor inventory visibility causes operational difficulties; 14% describe these as significant
  • Top mitigation strategies: supplier diversification (18%), better forecasting (18%), stronger risk management (17%)

These conditions — supply chain disruption, visibility gaps, stockout risk — are the operating context where ToC constraint identification and buffer management are most directly applicable.

Rami Goldratt on modern relevance (2026)

In a Supply Chain Now podcast (2026-05-20), Rami Goldratt argued:

  • Most organisations have a focus problem, not a technology problem — spreading resources across too many initiatives prevents any single constraint from being addressed
  • Demand Forecasting|Forecast accuracy is structurally declining because: product variety is expanding; fast-delivery expectations spread inventory geographically; product lifecycles are shrinking — three compounding trends
  • Spreading inventory geographically to meet fast-delivery expectations is self-defeating: it worsens forecast accuracy and increases total inventory investment
  • New products compound the problem further: no forecast baseline exists at launch
  • The dominant technology investment pattern in supply chain — automating what you can, not what you should — misdirects effort away from the constraint

Implementation limits

Umbrex (undated) notes ToC may mislead when: work is unique and non-repetitive (no shared resources to constrain); severe supply shortages call for S&OP allocation rather than internal DBR; leaders continue rewarding local efficiency metrics (subordination fails); or insufficient market demand requires commercial rather than operational responses.

Umbrex estimates a focused ToC pilot typically delivers results in 6–10 weeks (as-of unknown date — rough practitioner guidance; volatile). Elevation investments and multi-line rollouts take 3–6 months.

Key terms

TermMeaning
Constraint / BottleneckThe single resource or policy limiting overall system throughput
Throughput (T)Rate at which the system generates money through sales, net of truly variable costs
Inventory (I)Money tied up in things the system intends to sell
Operating Expense (OE)All money spent to turn inventory into throughput
DrumThe constraint, setting the pace for the whole system
BufferTime or stock held upstream to protect the drum from starvation
RopeSignal that limits upstream release of work to match drum pace
TDDThroughput Dollar Days — a supply chain KPI measuring opportunity cost of delayed throughput
IDDInventory Dollar Days — a supply chain KPI measuring cost of excess inventory over time
DBRDrum-Buffer-Rope — ToC's production scheduling mechanism
DDOMDemand Driven Operating Model — ToC-derived supply chain execution framework
COBComponent buffer — stock buffer positioned after a bottleneck station in DDMRP
S-DBRSimplified Drum-Buffer-Rope — ToC demand management construct
MTAMake-to-Availability — ToC replenishment trigger mode
DBMDynamic Buffer Management — adaptive buffer sizing under demand-driven planning
Research agent · 2026-10-25