06-reference/research

apqc pcf function motion mapping

2026-08-17·research-brief·source: deep-research·by Ray Data Co (deep-research synthesis)·! medium
apqc-pcforganizational-platformtaxonomykwik-tripphdata

APQC PCF vs the Organizational Platform's function/motion tree — what to borrow, what to build

The question

Verbatim from the Notion Research Backlog: "What established frameworks cover function/motion decomposition of organizations (APQC PCF, BIZBOK capability mapping, operating-model design, GTM-motion literature, industry reference models), and what should the organizational-platform's function/motion tree borrow vs build?"

Scope narrowed 2026-08-16 to APQC's Process Classification Framework only. BIZBOK/operating-model design and GTM-motion/industry-reference-models (eTOM, BIAN, SCOR) are covered by two sibling briefs filed the same night. Consumer: the Organizational Platform build, phase 3 — the 10 function pages and 36 motion pages are frontmatter-and-spine until this framework-research round lands the hand-section template.

Terminology in this brief: our levels are function and motion. PCF's own level vocabulary (Category / Process Group / Process / Activity / Task) is named as PCF's and mapped onto ours, never merged with it.

What we already know (from the vault)

What the web says

1. What the cross-industry PCF actually is

Read directly. I downloaded and read the full APQC-authored PDF of APQC's Process Classification Framework® — Cross-Industry, Version 7.4 (footer: "Version 7.4 • August 2024"), mirrored publicly at solutions.ifrc.org (APQC catalog code K014750). Everything in this subsection is from that document's own text unless flagged otherwise.

APQC's self-description, verbatim from the overview page: "a taxonomy of cross-functional business processes intended to allow the objective comparison of organizational performance within and among organizations... developed by APQC and its member companies as an open standard to facilitate improvement through process management and benchmarking, regardless of industry, size, or location." It "organizes operating and management processes into 13 enterprise-level categories, including process groups and more than 1,000 processes and associated activities." Origin: 1992, initial design involving APQC plus 80+ organizations. It is the backbone of APQC's Open Standards Benchmarking database.

The five levels, verbatim definitions from the "PCF Levels Explained" graphic on page 3 (I read the rendered page, since the graphic does not extract as text):

PCF level Name APQC's definition (verbatim) Worked example from the PDF
Level 1 Category "Represents the highest level of process in the enterprise in general groupings such as manage customer service, supply chain, finance, and human capital." 11.0 Manage Enterprise Risk, Compliance, Remediation and Resiliency (16437)
Level 2 Process Group "Groups of processes that are part of executing a category. Examples include perform after sales repairs, procurement, accounts payable, recruit/source, and develop sales strategy." 11.1 Manage Enterprise Risk (17060)
Level 3 Process "A single process that may include elements related to variants and rework in addition to the core elements needed to accomplish the process." 11.1.3 Manage Business Unit and Function Risk (17462)
Level 4 Activity "A key step performed to execute a process." 11.1.3.3 Develop Mitigation Plans for Risks (16458)
Level 5 Task "An element of work that goes into executing an activity. They are generally much more fine-grained and may vary widely across industries." 11.1.3.3.1 Assess Adequacy of Insurance Coverage (18129)

Numbering scheme, verbatim: "The PCF identifies each process element using a unique 5-digit reference number following the name of the process element... This number will always refer to the conceptual definition of the process element. The actual process elements and actual definition may change, but conceptually the decomposition will remain consistent... A new 5-digit number will be assigned to a process element if its definition substantially changes." So there are two identifiers per element: a positional dotted path (11.1.3.3.1) that is unstable across versions, and a stable 5-digit concept id (18129) that is the real key. This is the single most transferable idea in PCF and we should copy it outright — see the synthesis.

The 13 categories, with the operating vs management-and-support split exactly as the PDF draws it:

Operating processes:

  1. 1.0 Develop Vision and Strategy
  2. 2.0 Develop and Manage Products and Services
  3. 3.0 Market and Sell Products and Services
  4. 4.0 Manage Supply Chain for Physical Products
  5. 5.0 Deliver Services
  6. 6.0 Manage Customer Service

Management and support services: 7. 7.0 Develop and Manage Human Capital 8. 8.0 Manage Information Technology (IT) 9. 9.0 Manage Financial Resources 10. 10.0 Acquire, Construct, and Manage Assets 11. 11.0 Manage Enterprise Risk, Compliance, Remediation, and Resiliency 12. 12.0 Manage External Relationships 13. 13.0 Develop and Manage Business Capabilities

The full level-2 process-group list (extracted from the same PDF; this is the layer that actually competes with our motions):

Current version. Version 7.4 is dated August 2024. APQC's resource library also lists a Cross-Industry PCF Version 8.0 in PDF and Excel, and APQC's own resource-collection page states v8.0 is the latest, released 2025-05-30 (secondary: search-surfaced APQC page text at apqc.org resource listing; APQC's HTML pages return HTTP 403 to automated fetch, so I could not read the page myself). Descriptions of 8.0 still say "13 enterprise-level categories" and "over 1,000 processes." I could not obtain the v8.0 full text — no public mirror found, and the apqc.org static-asset path requires the exact filename. Treat every level-2 name above as verified-for-7.4 and presumed-stable-in-8.0, not verified for 8.0.

Addendum (added at close-out, 2026-08-17): partial v8.0 text obtained — one correction, two upgrades. A background sub-agent reached the first 5 of 38 pages of the v8.0 cross-industry document (APQC item K016809). Sourcing chain, stated plainly because it is not a clean primary read: the document's listing on Scribd embeds OCR'd text of its first ~5 pages as a free SEO preview, reachable through a readability proxy; pages 6–38 remain paywalled. This is APQC's own document text at one remove, with OCR risk, and it is better evidence than the search-snippet secondary above but weaker than the direct 7.4 PDF read.

2. The retail PCF — and why it under-serves a convenience/fuel operator

Read directly. APQC's static-asset path is not behind the 403 that blocks its HTML pages, so the retail PDF is fetchable: K09368_Retail Process_v721_011519.pdf. Title: "RETAIL PROCESS CLASSIFICATION FRAMEWORK® — Version 7.2.1", January 2019, 30 pages, developed by APQC in conjunction with Microsoft and stated as compliant with Cross-Industry PCF v7.2.1. Originally announced November 2012 (APQC press release).

It keeps 13 categories and the same management-and-support block (7.0–13.0 identical to cross-industry), but rebuilds the operating half:

Cross-industry 7.4 Retail 7.2.1
1.0 Develop Vision and Strategy Develop Vision and Strategy (unchanged)
2.0 Develop and Manage Products and Services Develop and Manage Customer Experience (16477)
3.0 Market and Sell Products and Services Market Products and Services (16610)
4.0 Manage Supply Chain for Physical Products Merchandise Products and Services (16634)
5.0 Deliver Services Deliver Products (16679)
6.0 Manage Customer Service Deliver Services (20025)

Retail operating process groups: 2.1 Operate direct channels · 2.2 Develop and manage retail store channel · 2.3 Operate retail store · 2.4 Develop and manage omni channel · 2.5 Manage customer relationship || 3.1 Perform customer and market analysis · 3.2 Market to customers || 4.1 Develop product plan · 4.2 Manage product development lifecycle · 4.3 Source products || 5.1 Operate warehouse · 5.2 Manage transportation and logistics · 5.3 Manage enterprise inventory · 5.4 Manage direct-to-store-delivery || 6.1–6.3 service delivery governance / resources / delivery.

The retail-native content that cross-industry genuinely lacks, with PCF numbers: 2.2.2.2 Design seasonal layouts (planograms) · 2.3.1 Operate customer checkout · 2.3.2.6 Manage store pick-up orders · 2.3.4 Manage store inventory (2.3.4.2 Restock shelves; 2.3.4.3 Monitor and remedy out of stocks) · 2.3.5 Manage store cash · 2.3.6 Plan and manage retail store workforce · 2.3.7 Execute floor merchandising and promotions (2.3.7.2 Maintain shelf tagging) · 2.4.4 Prevent loss (shrink) · 2.4.6 Manage omni-channel experience · 3.2.6 Develop and manage customer loyalty programs · 4.1.1 Develop category plan · 4.1.2 Define assortment plan · 5.3.1 Replenish retail stores · 5.4 Manage direct-to-store-delivery.

The Kwik Trip verdict is a hard one. A case-insensitive grep of the full extracted text for fuel, convenience, c-store, grocery, foodservice, perishable, private label, pharmacy returns zero matches in all 30 pages. (I ran this myself against my own download, independently of the sub-agent that first located the file; matches for "deli" are all substrings of "delivery"/"deliver".) The Retail PCF is a general-merchandise / apparel / omni-channel model built with Microsoft's retail SMEs. It has no process content for fuel retailing, forecourt operations, prepared food/foodservice, fresh and perishable handling, private label manufacturing, or the convenience format. For Kwik Trip, which pumps fuel, runs a commissary and bakery, makes its own dairy, and self-distributes, roughly the entire value-making core of the business is outside this framework's scope. Named plainly: there is no APQC industry PCF for convenience stores, grocery, or fuel retail.

3. Licensing — the one unambiguous good news

Read directly, verbatim from the v7.4 cross-industry PDF, copyright page:

©2024 APQC. ALL RIGHTS RESERVED. The Process Classification Framework is the copyrighted intellectual property of APQC. APQC encourages the wide distribution, discussion, and use of the PCF for classifying and defining organizational processes. Accordingly, APQC hereby grants you a perpetual, worldwide, royalty-free license to use, copy, publish, modify, and create derivative works of the PCF, provided that all copies of the PCF and any derivative works contain the following attribution:

APQC's Process Classification Framework® (PCF) is an open standard developed by APQC, a nonprofit that promotes benchmarking and best practices worldwide. The PCF is intended to facilitate organizational improvement through process management and benchmarking, regardless of industry, size, or geography. To download the full PCF or industry-specific versions of the PCF, as well as associated measures and benchmarking, please visit www.apqc.org/pcf.

The Retail PCF v7.2.1 carries a materially identical grant (©2019, opening "This Process Classification Framework® ('PCF') is the copyrighted intellectual property of APQC"). Page footers on both read "Permission granted to photocopy for personal use." — that line sits alongside, and is narrower than, the derivative-works grant on the copyright page; the grant is the operative term.

What this means for a consulting firm, stated carefully. The license text is unqualified as to who may exercise it — no non-commercial carve-out, no consultant exclusion, no field-of-use limit, no fee. Derivative works are expressly permitted. The only stated condition is the attribution paragraph, reproduced in all copies of the PCF and any derivative works. My reading: phData may derive a client-facing taxonomy from the PCF, including in paid engagement deliverables, provided the attribution paragraph rides along on the derivative. That reading is mine, not APQC's — it is an inference from the license text, not a statement APQC makes about consultancies. If the Organizational Platform ships PCF-derived structure to a client, get it past phData legal on the one question the text does not answer: whether attribution must appear on the client deliverable itself or whether crediting it in the platform's own provenance suffices.

Three honest gaps. (a) The PCF is also distributed as Excel, TreePad, and via APQC's benchmarking portal; I could not read APQC's website terms of usewww.apqc.org returns HTTP 403 to every automated fetch I attempted, including with a browser user-agent. A site-wide ToU could impose conditions the PDF does not. (b) The framework is downloadable at no charge per the PDF's own overview ("available for download at no charge at www.apqc.org/pcf"), but whether that requires free account registration I could not verify — the download pages are behind the same 403. (c) The measures and benchmarking data attached to the PCF are an APQC membership product and are not covered by the framework's derivative-works grant as far as I can tell; the grant text names the PCF, not the Open Standards Benchmarking dataset. Do not assume benchmark numbers are free to redistribute. PCF® is a registered trademark of APQC; use the ® on first mention.

Convergences and contradictions

Synthesis for RDCO

Bottom line: borrow PCF's identifier discipline and its coverage checklist; do not borrow its shape. PCF is the wrong axis for what our tree is for, but it is the best available auditor of whether our tree has holes. Treat it as a completeness oracle and an id-design precedent, not as a source of structure. The license is permissive enough that borrowing is a design question rather than a legal one, which is the unusual and lucky part of this answer.

The mapping table

Our function/motion tree against PCF 7.4 cross-industry (CI) and Retail 7.2.1 (R). "Buries" means PCF has the content but at a depth or under a parent that makes it invisible at our altitude.

Our function → motions Where PCF puts it The mismatch
FN.SALES — DTC/ecommerce, Field, Inside, Partner/channel, Retail/store CI 3.0 (fused with Marketing); channels appear as childless leaves CI 3.5.6/3.5.7/3.5.8 + 3.2.3 Define and manage channel strategy; partner at 3.4.2 / 3.5.5. R 2.1 Operate direct channels, 2.2/2.3 retail store channel + store ops PCF fuses Sales into Marketing at category level, then splits Sales by artifact (leads, accounts, proposals, orders) rather than by channel. Our five motions are PCF's three undeveloped leaves plus two PCF doesn't name. Inside sales has no PCF node at all.
FN.MARKETING — Brand & content, Comms, Performance/demand gen, Product marketing CI 3.1, 3.2, 3.3 + 3.3.9 Manage product marketing material; PR lives elsewhere, at CI 12.5 PCF has no brand/content vs performance split — that distinction is a modern channel-economics distinction, not a process one. PCF exiles PR to External Relationships; we put it in Marketing comms. Genuine disagreement about what belongs together.
FN.SERVICE — Contact/call center, Field service, Success/account mgmt CI 6.0 (contact center = 6.2); field service buried at 6.3.2.3.2, a level-5 task under warranty-claim investigation; account management sits at CI 3.5.2 under Sales, not Service PCF buries field service four levels deep under warranty and puts account management in Sales. Both are motions we surface as first-class. PCF adds recalls (6.4) and satisfaction measurement (6.5), which we have nowhere.
FN.OPERATIONS — Core production, Maintenance & facilities, Quality control, Real estate & facilities dev Core production = CI 4.3 Produce/Assemble/Test product, i.e. inside Supply Chain. Quality = CI 4.1.8 + 4.3.3 + 13.3 Manage enterprise quality. Maintenance + real estate = the whole of CI 10.0 Acquire, Construct, and Manage Assets (10.1–10.4) The sharpest structural disagreement. PCF has no Operations category — production is a supply-chain sub-process and asset work is its own top-level category. We invert both. Note our REAL_ESTATE motion maps cleanly to CI 10.2 Design and construct assets, which validates the 2026-08-08 KT-driven addition.
FN.SUPPLY_CHAIN — Distribution/fulfillment, Inventory & planning, Procurement CI 4.1 (planning), 4.2 (procurement), 4.4 (logistics/warehousing). R 5.1 warehouse, 5.2 transportation, 5.3 enterprise inventory, 5.4 direct-to-store-delivery Near-clean 1:1 at the level-2 layer — the best-fitting function in the whole tree. Retail PCF adds DSD and store replenishment (R 5.3.1), which a convenience operator runs constantly and we have no node for.
FN.FINANCE — AP/AR & billing, FP&A, Tax & treasury CI 9.1–9.11 (eleven groups) Same territory, 11 PCF groups vs our 3 motions. PCF surfaces internal controls (9.8) and payroll (9.5); we file payroll under HR. Resolution gap, not an axis conflict.
FN.HR — Recruiting, Payroll & benefits, L&D, Talent ops CI 7.1–7.8 (eight groups); payroll processing is CI 9.5, under Finance PCF splits the employee lifecycle finer (onboarding, relations, reward, redeploy/retire, comms) and routes payroll processing to Finance while keeping reward design in HR. Our single payroll motion straddles that line.
FN.IT — Applications, Data & analytics, Infrastructure, Security CI 8.1–8.7; security is CI 8.3 IT resilience and risk; analytics is split across CI 8.4 Manage information and 13.7 Develop, manage, and deliver analytics PCF's IT is organized by service lifecycle (strategy → build → deploy → support), not by asset class. Our four motions are asset classes. Real axis conflict, second only to Sales. PCF putting analytics partly outside IT is a signal worth taking seriously for a data consultancy's own taxonomy.
FN.PRODUCT — Product mgmt, Design, Engineering/R&D CI 2.1–2.3. R 4.2 Manage product development lifecycle PCF's 2.0 is a stage-gate development pipeline (govern → ideate → develop), not a discipline split. Our three motions are disciplines. Superficially similar, structurally different.
FN.LEGAL_RISK — Contracts, Regulatory & audit, Risk mgmt Risk/compliance = CI 11.0 entirely (11.1–11.4). Legal = CI 12.4 under External Relationships. Contracts are fragmented: 12.4.9 (legal agreements), 4.2.2 (supplier contracts), 5.3.1.1 (service terms), 8.1.5.5 (IT customer contracts) PCF refuses to co-locate legal and risk, and refuses to treat contracts as one thing. Our CONTRACTS motion is a deliberate re-aggregation of four PCF locations — defensible, but we should know we are overruling PCF here, not accidentally diverging.
(no function) CI 1.0 Develop Vision and Strategy; CI 12.0 External Relationships (investor, government, board, PR); CI 13.0 Business Capabilities (BPM, PMO, change, KM, benchmarking, EHS, sustainability) The real coverage gap. Three PCF categories have no home in our tree. Our EXEC pseudo-motion holds executive roles but there is no function for strategy, corporate affairs, or enterprise capability management. For Kwik Trip, EHS (13.8) alone is a staffed, regulated, budget-owning function.

Borrow-vs-build verdict, per function

Function Verdict Why, in one line
FN.SALES Build PCF names our channel motions and then leaves them as empty leaves; there is nothing under them to borrow.
FN.MARKETING Build PCF fuses Marketing into Sales and exiles PR; the modern brand-vs-performance split has no PCF analogue.
FN.SERVICE Build, borrow one node Keep our three motions; steal CI 6.4 Manage product recalls as a candidate motion — regulated food retail runs recalls as a standing capability.
FN.OPERATIONS Build, but re-check against CI 10.0 Our inversion of PCF (Operations owns production and assets) is correct for a routing tree; use CI 10.1–10.4 as the checklist for whether MAINTENANCE and REAL_ESTATE together cover plan/design/maintain/retire.
FN.SUPPLY_CHAIN Borrow The only near-1:1 fit; adopt CI 4.1/4.2/4.4 boundaries as-is and add a direct-store-delivery motion from R 5.4.
FN.FINANCE Borrow the boundaries, keep our resolution CI 9.x is the mature, uncontroversial cut; use it to sanity-check that AP/AR, FP&A, and Tax/Treasury leave nothing stranded (internal controls currently are).
FN.HR Borrow selectively CI 7.x is finer than we need, but its payroll-processing-is-Finance boundary is a real design question our single PAYROLL motion currently papers over.
FN.IT Build, with one borrow PCF's lifecycle axis conflicts with our asset-class axis; the one idea to steal is CI 13.7 — analytics as an enterprise capability, not an IT sub-team.
FN.PRODUCT Build PCF models a stage-gate pipeline; we model disciplines that staff seats. Different questions.
FN.LEGAL_RISK Build, knowingly Our re-aggregation of PCF 11.0 + 12.4 + four scattered contract nodes is a deliberate override; document it so it reads as a decision, not drift.
Coverage gap: Strategy / Corporate affairs / Enterprise capabilities Borrow as a gap list, decide separately CI 1.0, 12.0, 13.0 are three staffed function-shaped things our tree cannot express; EHS and PMO are the two most likely to bite on a real engagement.

The structural critique — where PCF simply cannot be borrowed

PCF decomposes on process: what is done, in what order, to produce an output. Our tree decomposes on function → motion: who owns a domain, and which operating pattern they run it in. These are orthogonal, and the orthogonality is not a defect in either.

The proof is at CI 3.5.6/3.5.7/3.5.8. APQC put physical outlets, field sales, and digital sales into the framework as three sibling processes — and gave them no children. In a process taxonomy those three are near-identical: qualify, quote, close, book. The differences that matter are cost structure, headcount shape, cycle time, tooling, and who you hire — every one of which is a staffing and economics fact, not a process fact. PCF cannot decompose them because there is nothing process-shaped to decompose. That is exactly the material our motion layer exists to carry, and it is why the motion layer has to be built rather than borrowed. Any attempt to derive our 36 motions from PCF's level 2 will produce a process taxonomy wearing our labels, and the 80 role pages will not hang off it.

The mismatch is superficial, though, wherever a function's work is genuinely one operating pattern. Finance, HR, and Supply Chain are like this: there is no meaningful "field AP" vs "inside AP." For those functions our motions are effectively PCF process groups at lower resolution, and borrowing PCF's boundaries costs nothing and buys coverage. The honest split is: the motion layer is real and un-borrowable in the customer-facing functions (Sales, Marketing, Service, and IT), and is a thin relabeling of process groups in the back-office functions. That asymmetry is itself a finding the platform should state out loud rather than hide behind a uniform-looking tree.

Two things to borrow outright regardless:

  1. The dual-identifier scheme. A stable 5-digit concept id decoupled from the positional path, with a new id minted when a definition substantially changes. Our current ids (FN.SALES.INSIDE) are positional and semantic and stable — three jobs one string cannot hold. The moment a motion moves between functions, every client YAML that cites it breaks. PCF solved this in 1992; we should copy the solution before the first re-parenting, not after.
  2. The completeness oracle. Run every client's org through the 13 PCF categories as a checklist. If a category has no home in our tree and the client staffs it, that is a coverage gap. For Kwik Trip, that test currently fires on EHS, PMO/change management, government relations, and internal controls.

And one thing to stop expecting: there is no APQC framework for convenience, grocery, or fuel retail. The Retail PCF is a general-merchandise omni-channel model with zero content on forecourt, foodservice, fresh, or private-label manufacturing. Kwik Trip's core value-making work — pump fuel, make food, make dairy, self-distribute — is the part APQC does not cover. Our FN.OPERATIONS.CORE_PRODUCTION motion, defined as "the industry's value-making work," is doing the job PCF's retail cut cannot, and that is a point in favor of the design we already have.

Why this is in the vault

This brief unblocks phase 3 of the Organizational Platform build: the README states the hand-section template locks after the framework-research round, and this is that round for the PCF leg. It settles three concrete build decisions — that the motion layer is built not borrowed, that supply chain and finance boundaries can be lifted from PCF 7.4, and that PCF's stable-concept-id scheme should be adopted before any motion is re-parented — plus one client-facing fact for the Kwik Trip engagement: there is no APQC framework covering convenience, fuel, or grocery operations, so the core-production motion carries that weight alone.

Open follow-ups

Related

Filed the same night, same research round — and they do not fully agree. [[2026-08-17-gtm-motion-origins-industry-reference-models]] · [[2026-08-17-bizbok-operating-model-function-motion-mapping]]. Two disagreements the founder should see before the content template locks. (1) This brief concludes the motion layer is real and un-borrowable in the customer-facing functions including Marketing and IT; the GTM sibling grades Marketing and IT as sub-functions rather than true motions. Opposite verdicts on the same two functions. (2) The GTM sibling proposes an axis: substitutable | sub_function frontmatter amendment; the BIZBOK sibling concludes the founder's original prior survives and proposes a different addition (a configured-as block: location / systems / org shape). Note also that the GTM sibling carries a VERIFICATION STATUS: ITERATE (hard) banner — weigh its verdict accordingly against this one.

Sources

Primary, read directly:

Secondary, not read directly (apqc.org HTML returns HTTP 403 to automated fetch):

Vault:

Working tree (phData-internal, not published):