Supply Chain Metrics and the SCOR Model
Most supply chain metrics arguments are definitional arguments in disguise. Two companies reporting the same perfect order percentage are frequently measuring different things at different levels against different commitments.
The model is a dictionary before it is a benchmark. Its main contribution is shared definitions and a decomposition structure, not a set of target numbers.
Perfect order is an all-conditions test. Drop the documentation or condition test and the figure rises without the operation changing.
On time in full is defined by your customer. Large retailers impose their own windows, measurement levels, and penalties, so the same performance scores differently by account.
The unit of measure changes the answer. Line, order, case, and shipment produce materially different percentages from identical operations. Always state the unit.
External benchmarks are weaker than they look. The available sources are self-reported, sample-dependent, or sold by parties with a commercial interest. Internal trending on a fixed definition is more useful.
Market overview
The short answer
The Supply Chain Operations Reference model, known as SCOR and maintained by the Association for Supply Chain Management as the SCOR Digital Standard, is a shared process and metrics framework for describing and measuring a supply chain. Its value is definitional: it gives commonly used terms a specific meaning, a code, and a place in a hierarchy, so that perfect order fulfillment or cash-to-cash means the same thing to two people who have both read it. What it cannot do is make undisciplined measurement comparable. Most disputes about supply chain performance are not disagreements about the operation; they are disagreements about the unit of measure, the commitment date, and which conditions were counted. Fixing that is what this model is for, and it is worth more than any benchmark.
KEY FACTS
Verified August 2026. Each statement below is complete on its own and cites its source in section 08.
What is SCOR, and what changed in the Digital Standard?
SCOR began in 1996 as a cross-industry reference model, developed under the Supply Chain Council. When that body merged into APICS in 2014, stewardship passed to what is now the Association for Supply Chain Management, which maintains the current SCOR Digital Standard with open access rather than behind a purchase. That access point matters practically: a team can adopt the definitions without buying anything, which removes the usual excuse for local invention.
The Digital Standard is a substantial revision rather than a version increment. Where earlier SCOR described a linear chain of Plan, Source, Make, Deliver, and Return, the Digital Standard describes seven processes and arranges them as a continuous loop rather than a line. Deliver is split into Order and Fulfill, separating the commercial act of taking and managing an order from the physical act of fulfilling it. Make is renamed Transform, which extends the model to services and to businesses that do not manufacture. And a new process, Orchestrate, sits above the others to cover the coordination, governance, and decision layer that the older model left implicit.
Performance in the model is organized by attributes, which are the dimensions along which a supply chain can be good or bad. The long-established set is reliability, responsiveness, agility, cost, and asset management efficiency, with the first three customer-facing and the last two internal. Recent editions of the Digital Standard have extended and relabeled this set, including sustainability considerations, so a page or a scorecard citing attributes should name the specific edition it follows and date it rather than treating the attribute list as fixed.
For a practitioner the useful summary is that SCOR supplies four things: a process vocabulary, a metric hierarchy with codes, definitions for each metric, and a way of decomposing a bad number into its causes. Organizations adopt it well when they take the definitions and the decomposition. They adopt it badly when they take the process diagram, redraw it as a slide, and keep measuring the way they always did.
How does the metric hierarchy actually work?
Metrics in SCOR sit at three levels, and the relationship between them is diagnostic rather than decorative. Level 1 metrics are strategic: a small number of headline measures that describe how the supply chain is performing overall, of which perfect order fulfillment is the best known. Level 2 metrics explain a level 1 result by decomposing it into components. Level 3 metrics explain a level 2 result in turn, and are where the operational causes actually live.
Each metric carries a code that identifies its attribute and level. Perfect order fulfillment is RL.1.1: reliability, level 1, first metric. Its components are RL.2.1 through RL.2.4. The coding is not bureaucratic decoration; it is what allows two organizations, or two systems, to confirm they are discussing the same measure rather than two measures with similar names.
The practical use is root-causing. A level 1 number tells you that something is wrong and nothing about what. Decomposing it into level 2 tells you which condition failed, and level 3 tells you why that condition failed. An organization that reports only level 1 has bought the headline and skipped the diagnostic, which is the most common way of using the model without benefiting from it. The corollary is that a scorecard with a single reliability percentage on it and no decomposition beneath is not a SCOR scorecard, whatever it is called.
What exactly counts as a perfect order?
Perfect order fulfillment is the percentage of orders that meet every delivery performance condition with no error. It is the most overstated metric in supply chain because the conditions are frequently trimmed. Under the model an order qualifies only when four things all hold: it was delivered in full, meaning every line and every unit; it was delivered on time to the customer's commit date, measured on the customer's definition of on time rather than the supplier's; the documentation accompanying it was complete and accurate; and the goods arrived in perfect condition, undamaged and correctly configured.
Figure 1. The decomposition. All four level 2 conditions must hold for a single order to count as perfect, and level 3 diagnostics sit beneath each one to explain failures. Dropping any condition raises the reported number without changing what customers experienced.
The two conditions most often quietly dropped are documentation and condition, because both are harder to capture systematically than in-full and on-time. An organization measuring only in-full and on-time is measuring something legitimate, and it should call that something else. The distinction matters commercially when a customer audits the claim, because a supplier reporting a perfect order percentage that excludes damage and paperwork is making a representation it cannot support.
The associated professional framing, sometimes called the seven rights, expresses the same idea in plainer terms: the right product, in the right quantity, in the right condition, to the right place, at the right time, for the right customer, at the right cost. It is a useful device for explaining to a commercial audience why the composite metric is demanding, and why a high figure is a stronger claim than it first appears.
Why does on time in full mean different things to different customers?
On time in full sounds self-explanatory and is not. Every element is a definitional choice: what counts as on time, measured against which date, at what level of aggregation, and with what tolerance. A supplier can move from strong to weak performance on the same shipments by changing only the measurement basis, which is why cross-company and even cross-customer comparison of this metric is unreliable without stating the definition alongside the number.
Large retailers settle the question by imposing their own definition and attaching money to it. Walmart's program, introduced in 2017, measures deliveries against a must arrive by date and sets a threshold for compliance, with a penalty expressed as a percentage of the cost of goods on non-compliant cases; the retailer moved from monthly to quarterly chargeback billing in February 2024. Other large retailers run equivalents under different names and with different thresholds and measurement points. These specifics change, and any supplier planning against them should verify current terms with the retailer rather than relying on secondary summaries, including this one.
Table 1. The five elements that make on time in full non-comparable. A supplier that has not settled all five in writing with a major customer is measuring something the customer may not recognize.
How is cash-to-cash calculated, and can I benchmark any of this?
Cash-to-cash cycle time measures how long working capital is committed between paying a supplier and collecting from a customer. It has three components: days inventory outstanding, which is how long stock sits; days sales outstanding, which is how long customers take to pay; and days payable outstanding, which is how long the company takes to pay suppliers. The first two are added and the third subtracted, so the metric falls when inventory turns faster, when customers pay sooner, or when the company pays its own suppliers later.
That last lever is where the metric is most often misread. A company can improve cash-to-cash substantially by extending payment terms, which transfers the working capital burden to its suppliers rather than removing it from the chain. Whether that is good practice is a separate question, covered in SCR's guide to supply chain finance, and the point here is narrower: a cash-to-cash improvement should always be decomposed, because the three components carry different meanings and different consequences.
Table 2. The five most misused measures. In every row the misuse is definitional rather than arithmetic, which is precisely the failure a reference model is designed to prevent.
On benchmarking, the honest position is uncomfortable. There is no authoritative public benchmark for perfect order or cash-to-cash by industry that is free of self-reporting or commercial sampling bias. The most methodologically disciplined source is a member-based benchmarking body that standardizes definitions across participants, which is a real strength, and its data is still self-reported by firms that chose to participate. Consultancies and analyst firms sell benchmark data and have a commercial interest in the comparison being purchased. None of this makes external benchmarks worthless; it makes them directional.
The fair case against this skepticism deserves stating, because taken too far it becomes an excuse to measure nothing. Standardization is exactly what SCOR and disciplined benchmarking programs exist to provide, and an organization that adopts the definitions rigorously and participates in a standardized program can obtain real and useful peer comparison. More importantly, internal trending against a fixed definition is valuable regardless of cross-company noise. The problem this page describes is undisciplined use, not the metrics themselves, and the remedy is to fix the definitions rather than to abandon measurement.
Frequently asked questions
What is the SCOR model and who maintains it?
It is the Supply Chain Operations Reference model, a cross-industry process and metrics framework originated in 1996 under the Supply Chain Council and maintained since 2014 by the Association for Supply Chain Management, which publishes it as the SCOR Digital Standard with open access.
How is the Digital Standard different from older SCOR?
It describes seven processes rather than six, arranged as a loop rather than a line: Orchestrate, Plan, Order, Source, Transform, Fulfill, and Return. Deliver was split into Order and Fulfill, Make was renamed Transform to cover services, and Orchestrate was added for the coordination and governance layer.
What are the SCOR performance attributes?
The long-established set is reliability, responsiveness, agility, cost, and asset management efficiency, with the first three customer-facing. Recent Digital Standard editions extend and relabel this set, so cite the specific edition you are following and date it rather than treating the list as fixed.
What are level 1, 2, and 3 metrics?
Level 1 metrics are strategic headline measures. Level 2 metrics decompose a level 1 result into its components. Level 3 metrics explain a level 2 result. The relationship is diagnostic: the hierarchy exists so that a bad headline number can be traced to an operational cause.
What exactly counts as a perfect order?
An order delivered in full, on time to the customer's commit date measured on the customer's definition, with complete and accurate documentation, and in perfect condition. All four must hold. Removing any condition raises the reported figure without improving what the customer experienced.
Why do on time in full numbers differ so much between companies?
Because five things vary: which date is measured against, how wide the on-time window is, whether counting is by line, order, case, or shipment, whether any shortfall is tolerated, and whether a penalty attaches. Change any one and the same performance produces a different score.
How is cash-to-cash cycle time calculated?
Days inventory outstanding plus days sales outstanding minus days payable outstanding. Always decompose it, because the same improvement can come from turning inventory faster, collecting sooner, or simply paying suppliers later, and those three mean different things.
Can I benchmark my perfect order against my industry?
Only directionally. No authoritative public benchmark exists that is free of self-reporting or commercial sampling bias. Standardized member benchmarking programs are the most disciplined option and are still self-reported. Internal trending on a fixed definition is usually more informative.
Why does the unit of measure matter so much?
Because a single failure counts once at order level and many times at line or case level. Two organizations with identical operations can report materially different percentages purely from this choice, which is why a metric without its unit of measure is not a metric.
Is SCOR free to use?
The Digital Standard is published with open access by the Association for Supply Chain Management, which is a membership body that also sells training and certification. The framework can be adopted without purchase, which removes the usual justification for inventing local definitions.
Method, sources, and where to go deeper
Method
Model structure, process names, metric codes, and definitions follow the Association for Supply Chain Management's published SCOR Digital Standard material rather than secondary summaries, several of which still describe the superseded six-process model.
Retailer compliance program details are summarized from published descriptions and are explicitly flagged as subject to change, with readers directed to verify with the retailer.
Benchmarking sources are described by their funding and data collection model so that readers can judge their weight.
Supply Chain Research is independent and vendor-neutral. We accept no payment from the vendors or categories covered, and this page names no products.
Caveats
SCR publishes no benchmark for perfect order, on time in full, or cash-to-cash by industry. No authoritative public source exists that is free of self-reporting or commercial sampling bias, and figures sold as industry benchmarks should be treated as directional.
Metric counts and the exact performance attribute set differ between SCOR versions. Cite one specific edition and date it rather than blending versions, since much published material still describes the earlier model.
Retailer delivery compliance thresholds, penalties, and billing cadence change, in at least one case within the last two years. Verify current terms directly with the customer before planning against them.
The Association for Supply Chain Management maintains the model and also sells training and certification; benchmarking bodies referenced are member-funded and rely on self-reported data. Both facts are context rather than disqualification.
Figure 1, Table 1, and Table 2 are structural and definitional summaries rather than measured research findings.
Where to go deeper
Readers whose question concerns forecast accuracy measurement should read the SCR guide to demand planning and forecasting software, which covers the metric choices there and is deliberately not repeated here. The software ROI method guide covers building a business case, including how to handle benchmark claims. The inventory optimization and MEIO guide covers fill rate and days of supply from the planning side. Readers whose cash-to-cash improvement is coming from extended payment terms should read the SCR guide to supply chain finance and payables finance, and those scoping across categories should start with the supply chain software category map.
Sources
Sources
- Association for Supply Chain Management. SCOR Digital Standard. Standard maintainer. ASCM is a membership body that also sells training and certification.
- Association for Supply Chain Management. SCOR Digital Standard introduction and front matter, 2025 edition. Primary. Source of the current process structure and version.
- Association for Supply Chain Management. Perfect order fulfillment, metric RL. 1.1. Primary. The metric definition and its level 2 components.
- Association for Supply Chain Management. SCOR Digital Standard quick reference guide. Primary.
- APICS. SCOR version 12. 0 quick reference guide. Prior edition, cited for contrast with the Digital Standard.
- APQC. Benchmarking cash-to-cash cycle time. Member-funded benchmarking body. Data is self-reported by participating organizations using standardized instruments.
- Frontiers in Sustainability. Critical review of the SCOR Digital Standard. Peer reviewed. Independent assessment of the model.
- SPS Commerce. Description of retailer delivery compliance chargeback programs. Interested source: a vendor selling retailer compliance software. Cited for the mechanism; verify thresholds with the retailer.
