Business Paper 1 — Marketing and People

Level Threshold Guide

#WBS11 — Level/Threshold Guide (Unit 1: Marketing and People)

Provenance (2026-09-28). Built from a full read of research/veridian/WBS11-verified-facts.md §2 (the verbatim June 2019 level-descriptor tables for the 8-, 10- and 20-mark questions, the only series with a fully published mark scheme in the reviewed set, cross-series consistency independently confirmed against examiner-report language from June 2022, January 2023, June 2023, January 2024, June 2024, October 2021, January 2025 and October 2024) and §§3 and 10 for the specific graded real-exam anchors used as worked boundary illustrations below (Superdry profit share, Vietjet Air performance-related pay, Virgin Group leadership, Unilever organisational restructuring, Deutsche Bank restructuring, and the Define/Calculate/Analyse internal-mark traps). Cross-checked against forge-exam-technique-guide.md (read first, so nothing below repeats its tariff-by-tariff overview) and content/veridian/question-bank/wbs11.ts (read first, so the three full level-exemplars already built there — Skechers PLC, the electric-scooter B2B/B2C question, Lehman's — are cross-referenced, not rebuilt). Not affiliated with or endorsed by Pearson Edexcel. Every quotation below is a short, attributed fragment (kept under roughly fifteen words); no full mark-scheme passage, question or extract is reproduced. Illustrative "near-miss" and "clears" snippets in this document are authored for teaching purposes, built strictly from the real credited indicative-content points cited alongside them — they are not verbatim candidate scripts, and are labelled as such throughout, the same convention wbs11.ts's own level-exemplars already use.

What this is, and isn't. forge-exam-technique-guide.md §2 already states, tariff by tariff, what a Discuss/Assess/Evaluate answer needs in general terms. This document does the narrower, deeper thing that guide's own summary doesn't: it takes the real, verbatim level-descriptor language for each ladder, diffs it level by level to show exactly which words change at each boundary, and pins that language to real graded exam answers wherever the archive has one — naming the specific thing that kept a real response on one side of a boundary rather than the other. It is not a grade-boundaries document: signal-frequency-and-boundaries.md §2 already establishes that no raw-mark-to-grade boundary figures exist in anything reviewed for this paper, and that is a different kind of threshold from the per-question level boundaries below. Nothing here states or implies a grade boundary, a pass mark, or a likelihood percentage.

#1. Two different kinds of threshold on this paper

WBS11's marking structure runs two genuinely different mechanisms, and treating them as one "more detail = more marks" continuum is itself a source of lost marks:

Discrete marks (Define 2, Calculate/Construct 4, Explain 4, Analyse 6): each mark is either awarded or not, against a fixed Knowledge/Application/Analysis split. There is no judgement call — a candidate either wrote the second creditable idea or didn't. §4 below works through these.
Levels of response (Discuss 8, Assess 10, Evaluate 20): the whole answer is read and placed into a band first, then a mark is set within that band by how fully it meets the band's own descriptor. This is genuinely holistic — two answers that are both "Level 3" can score differently within it — and it is where a "just misses vs just clears" boundary actually exists in the way the word implies. §§2–3 below cover this in depth, because it is the harder of the two thresholds to see from the mark scheme's own printed table alone.

#2. The three level-of-response ladders, and how their bottoms and tops compare

Three ladders exist on this paper: Discuss's 3-level, 8-mark ladder; Assess's 4-level, 10-mark ladder; Evaluate's 4-level, 20-mark ladder. All three verified verbatim from the June 2019 full mark scheme [WBS11-verified-facts.md §2], with the same three-tier/four-tier shape and closely matching descriptor language independently confirmed still in force as late as June 2024 and January 2025's examiner reports.

	Discuss (8)	Assess (10)	Evaluate (20)
L1	1–2	1–2	1–4
L2	3–5	3–4	5–8
L3	6–8 (ceiling)	5–7	9–14
L4	—	8–10 (ceiling)	15–20 (ceiling)

The single most useful fact in this table is the shape, not the numbers. Discuss has no fourth level — its own top band is worded, almost word for word, the way Assess's third band is worded. Concretely: 8-mark Discuss's Level 3 (ceiling) asks for an assessment that is "balanced… using quantitative and/or qualitative information, and shows an awareness of competing arguments/factors." 10-mark Assess's Level 3 (not its ceiling) asks for much the same thing, but its own descriptor adds the qualifier that this awareness is "unlikely to show the significance of competing arguments" — a deliberately lower bar than Discuss's ceiling sets, even though it sits one rung lower on Assess's own four-rung ladder. The one thing that separates Assess's actual ceiling (Level 4, 8–10) from that Level 3 is a single added clause: "leading to a supported judgement." Practically, this means the same quality of counterbalanced reasoning that would max out an 8-mark Discuss question is not automatically enough to reach the top of a 10-mark Assess question on the same content — Assess additionally requires the judgement step Discuss never asks for at all. This is a real, textually-grounded difference between two command words that this paper's own tariff table lists as adjacent (4→6→8→10), not an invented distinction.

The same relationship holds, at greater length, between Assess and Evaluate: Evaluate's Level 3 (9–14) reads almost identically to Assess's own Level 3 language — developed chains of reasoning, and only "a partial awareness of… competing arguments" that "may" (not must) reach a conclusion — and its Level 4 (15–20) adds exactly the ingredient Assess's Level 4 adds — except Evaluate's version is stricter still: not just "a supported judgement" but "an effective conclusion that proposes a solution and/or recommendations." Evaluate's own top band is Assess's top band, plus a second, separate requirement — the conclusion has to propose something, not just judge between what's already been said.

Level 1 is worded almost identically across all three ladders ("isolated elements of knowledge and understanding… weak or no relevant application to business examples"), with one small but real difference: Evaluate's own Level 1 descriptor explicitly names "an argument," where Discuss's and Assess's shared Level 1 language instead says "generic assertions may be presented" — the 20-mark ladder expects some attempt at an argument even at its lowest band, where the 8- and 10-mark ladders don't ask for one yet.

#3. Threshold zoom-ins: what specifically separates one band from the next

Each entry below states the verbatim (short-quoted) language the boundary turns on, then a real, cited graded response that sits on one side of it, and an authored illustration of what moving it across the line would need to look like. Three questions on this paper already have a full, authored Level 1→Level 4 progression built and published — Skechers' product life cycle (8-mark Discuss), the electric-scooter B2B/B2C question (8-mark Discuss), and Lehman's niche-market question (10-mark Assess), all in content/veridian/question-bank/wbs11.ts. This section deliberately does not rebuild any of those three; it uses different, real anchors the archive has but no lesson file has yet turned into a full exemplar.

#3.1 Discuss, Level 2 → Level 3 — the stakeholder-accuracy boundary

June 2019 Q1(d), 8 marks (Superdry, a three-year profit-share scheme). The mark scheme's own credited chain runs: profit share is a financial incentive tied to firm performance, and signals to staff that the employer values their contribution; but lower-hierarchy staff may feel less motivated by a comparatively small allocated share, and the whole mechanism delivers nothing if Superdry simply doesn't turn a profit [WBS11-verified-facts.md §10.1]. The confirmed real trap: most candidates "focussed on the impact… on Superdry, rather than the impact on the employees" [PEF-Jun19] — the question asks what profit share does to staff, and a chain of reasoning that is accurate but aimed at the business's own finances cannot satisfy Level 3's requirement that knowledge be "applied accurately to the business and its context," because the context this question sets is the employees, not Superdry's balance sheet.

Illustrative near-miss (would sit at Level 2, 3–5/8). A response that explains, correctly and at length, that a profitable year lets Superdry afford larger payouts, and an unprofitable one means smaller ones — accurate knowledge, but the chain of reasoning never turns to what any of that does to a member of staff's motivation. This satisfies Level 2's "chains of reasoning… showing cause(s) and/or effect(s)" but not Level 3's requirement that those effects be shown "throughout" for the actual named context (employees), because the effects shown are all on the business.
What crosses the line. Adding the credited employee-facing half of the same mechanism — the scheme's own role-dependent payout range means staff lower in the hierarchy receive a far smaller share than senior staff, which the real mark scheme credits as a reason they may feel comparatively undervalued even in a profitable year (the lesson's own convention, followed here, is not to restate Superdry's exact published figures) — turns the same knowledge into an answer that is "supported throughout by relevant and effective use of the business behaviour/context," because the context now being used is the one the question set.

#3.2 Discuss, Level 2 → Level 3 — the counterbalance-depth boundary

January 2024 Q2(d), 8 marks (a fast-food business, emotional branding). Credited counterbalance content genuinely exists and is substantial — cost of ethical commitments raising prices in a price-sensitive market; the risk that a shared commitment (recyclable packaging, say) fails to differentiate a business from competitors making the same pledge [WBS11-verified-facts.md §3, 1.3.3.3]. The examiner's report states the specific failure mode directly: merely asserting that emotional branding "may not work is likely to be insufficient to reach level 3" [PEF-Jan24].

Illustrative near-miss (Level 2, 3–5/8). A response that develops the benefit side fully — emotional branding builds a felt relationship with the brand, which can raise loyalty and differentiate the business — then closes with one unlinked sentence: "however, emotional branding might not work for every customer." This is exactly the pattern the examiner names: a counterbalance clause with no developed reasoning behind it.
What crosses the line. Developing that same sentence into its own chain — the target audience for fast food skews younger and more price-sensitive, so the cost of the ethical commitment itself (not just "might not work") could push up prices in a way that specific audience is unusually likely to notice and resist — turns an asserted counterbalance into a developed one, satisfying Level 3's "logical chains of reasoning, showing cause(s) and/or effect(s)" on both sides of the argument, not just the supporting side.

#3.3 Assess, Level 2 ceiling — knowledge present, causal chain absent

June 2022 Q1(e), 10 marks (an airline, performance-related pay for pilots). A response scoring 3 of 10 marks — solidly inside Level 2's 3–4 band, not near Level 3's 5-mark floor — is described by the examiner's report as having "no clear links as to why the incentive will attract pilots" [PEF-Jun22]. This is the clearest available real illustration of Level 2's own descriptor language: "elements of knowledge and understanding, which are applied to the business example," but the "chains of reasoning" that follow are themselves the thing missing, not the underlying facts.

What the 3-mark response likely did. Named performance-related pay accurately, applied it to pilots specifically (not a generic workforce), but stopped there — the why this attracts or retains a pilot specifically step, the actual chain of reasoning Level 3 needs, was never built.
What crosses the line into Level 3 (5–7/10). One developed chain: pilots are a scarce, highly-trained workforce with genuine outside options, so a bonus tied to performance gives this specific group of staff a reason to stay with this specific airline rather than a competitor offering flat pay — the same underlying fact (PRP attracts pilots), but now with the causal mechanism the mark scheme actually credits, not just the label.

#3.4 Assess, Level 3 → Level 4 — the unlinked-clause boundary, again, at a higher tariff

June 2023 Q1(e), 10 marks (a sportswear business, market segmentation). Real, confirmed trap, worded almost identically to the Discuss-tariff version in §3.2 above but now capping a response at Level 3 rather than Level 2: some students' entire counterbalance was the single unlinked clause "segmentation is expensive and risky," which the report says "gains no marks due to lack of reasoning, development or context" [PEF-Jun23]. The recurrence of the identical failure mode at a different tariff, five questions apart in this archive, is itself evidence this is the single highest-value habit to drill across every double-digit-mark question on this paper, not a quirk of one question.

Illustrative near-miss (top of Level 3, around 7/10). A fully developed case for why segmentation helps the business target its marketing more precisely, closed with the exact unlinked clause the examiner quotes — present, but doing no work, because it names a cost without saying to whom, in what amount, or with what consequence.
What crosses the line into Level 4 (8–10/10). Naming which cost — market research and running separate campaigns for each segment raises marketing spend specifically, which for a business already competing on price could erode the margin segmentation was meant to protect — and then closing on a conditional judgement (segmentation is worth it if the segments are large enough to justify the extra spend, less so for a business with a narrow customer base) — satisfies Level 4's added requirement, "leading to a supported judgement," that Level 3 never asks for.

#3.5 Evaluate, Level 3 (near its own ceiling) — a real, numbered example

June 2023 Q3, 20 marks (Unilever, a move from matrix to flat structure). This is the one boundary in this whole document with a real, published numeric mark attached to a described response: 13 out of 20, explicitly identified as Level 3 [PEF-Jun23] — inside the 9–14 band, one mark short of its own top. The same report explains what the response did well — making "reference to both" the old and new structures "to give context" [PEF-Jun23], rather than describing flat structures generically — and, by naming the paper-wide failure this response avoided but didn't fully exceed, implies what the missing marks were for: linking the structural change specifically to Unilever's "efficiency and motivation" — the exact two outcomes the question named — rather than a general survey of flat structures' pros and cons.

What kept 13/20 inside Level 3, not at Level 4. Level 3's own descriptor caps at "a partial awareness of… competing arguments" and a chain of reasoning that "may" (not must) reach a conclusion — contextualised, developed, genuinely two-sided reasoning, but without Level 4's stronger, unconditional requirement for "a full awareness of the validity and significance" of those arguments, and a conclusion that goes further still — proposing "a solution and/or recommendations," not just judging between the sides already raised.
What the last two marks would need. Not more description of either structure, but a closing judgement that actually weighs the redundancy-driven efficiency gain (fewer staff, lower cost, a flatter chain of command) against the concrete motivation risk on the other side (workload redistributed onto remaining staff; a possible negative reaction to colleagues' job losses, the mirror image of the "survivor relief" effect this paper's own Deutsche Bank question separately credits, [WBS11-verified-facts.md §3, 1.3.4.3]) and states, specifically for Unilever's case, which effect is likely to dominate and why — the "solution and/or recommendations" step Level 4 is built around.

#3.6 Evaluate, Level 3-top → Level 4 — a named alternative is not a judgement

June 2022 Q3, 20 marks (Virgin Group, democratic leadership). A response the examiner's report calls a "high" Level 3 example is quoted attempting exactly the move Level 4 requires, and falling short of it: it "suggest[s] other types of leadership styles" but there is "little justification for this recommendation," and the response ultimately "fails to conclude if democratic leadership is the best style" for the business [PEF-Jun22]. This is a precise, real illustration of the difference between naming a competing argument (which Level 3 already credits — "a partial awareness of… competing arguments") and using it to reach the "balanced comparisons, judgements" Level 4's descriptor specifically asks for.

What the described response did (Level 3-top). Developed the case for democratic leadership fitting Virgin Group's culture, then gestured at an alternative style existing, without stating which one, why it might suit the business better or worse, or under what circumstance either would be preferable.
What crosses the line into Level 4. Naming the specific alternative (say, a more autocratic style) and the specific condition under which it would actually be preferable — a fast, unified decision during a genuine crisis, where consultation costs time the business doesn't have — then stating, for Virgin Group's actual described situation, which condition holds and therefore which style the evidence supports. The mark scheme rewards the comparison being resolved, not merely raised.

#4. Discrete-mark thresholds: real, numbered boundaries below the levels

These aren't "levels" — no holistic judgement is made — but several carry a real, specific, verified cap that behaves like a threshold in its own right, and belongs in a document about where this paper's marking structure actually draws lines.

Define (2 marks): one idea, however well written, does not reach 2. The recurring, verified trap across at least seven confirmed instances (qualitative research, market share, product differentiation, demand, empowerment, mass market, competitive advantage — see forge-exam-technique-guide.md §4.1) is restating the term rather than adding a second, separately creditable idea. Product differentiation is the most explicit confirmed case: a candidate who wrote only that it means "making the product different" was told this is "insufficient for 2 marks," because Pearson credits two distinct elements — the differentiating feature itself, and the fact that it stands out from competitors [PEF-Jun23]. The threshold is structural, not a matter of degree — a definition either contains two distinct creditable clauses or it doesn't, and no amount of elaborating the single clause crosses it. (No series reviewed states the exact partial-mark total a one-idea answer receives — treat the rule as verified and the resulting single-idea mark as unconfirmed, not assumed to be 1 of 2.)
Calculate (4 marks): a verified, exact numeric split on the same figure. Price elasticity of demand's sign is the one place in this whole archive where two different mark totals for effectively the same final number are printed verbatim: "Award 4 marks for correct answer -2.5 / Award 3 marks for 2.5" [WBS11-verified-facts.md §3, 1.3.2, citing EXM-Jun19]. The entire boundary between full marks and one-mark-short, on an otherwise perfectly worked calculation, is a single minus sign.
Analyse (6 marks): a confirmed hard cap for an incomplete pair. The Primark mass-market question — "analyse two possible difficulties" — is explicitly marked "capped at 3 marks if only one difficulty is given" [WBS11-verified-facts.md §3, 1.3.1.1, citing MS-Jan19]. Because Analyse's internal shape is two independent Knowledge→Application→Analysis chains (K2/App2/An2, confirmed structurally across this whole paper), a single, however well-developed, chain can only ever earn its own half — the cap isn't a penalty on top of a weaker answer, it's the direct arithmetic consequence of the mark scheme only crediting one of the two required chains.
Explain (4 marks): the same two-chain shape, one tariff down, with no confirmed exact cap. Confirmed independently five years apart that this tariff needs "two points of application," not one point developed further [PEF-Jan23, PEF-Jun19, both cited in forge-exam-technique-guide.md §2]. Unlike the Primark and PED cases above, no series in the reviewed archive states an exact numeric cap for a one-point Explain answer — treat the rule itself (two distinct points required) as verified, and the precise resulting mark as unconfirmed rather than invented here.

#5. What the pattern across every boundary above actually says

Read together, §§3–4 point at one repeated mechanism, not several unrelated ones: almost every real boundary this archive documents is crossed by adding a second, distinct, developed element to an answer that already has one — a second stage of a model (Skechers), a second credited benefit (Lehman's), a second application point (Explain), a second chain (Analyse), or the second half of a counterbalance that was only asserted, not developed (Discuss/Assess/Evaluate, §§3.2–3.6). The recurring failure is not thin knowledge — every anchor cited above shows real, accurate content on the page — it is stopping at one worked point when the mark scheme's own architecture, at every tariff on this paper, is built around crediting (at least) two.

#6. Gaps flagged, not filled

Discuss, Level 1 → Level 2 has no real, cited graded example in the archive reviewed for this document — every Level-boundary anchor found and verified sits at the Level 2/3 boundary or higher, because examiner reports overwhelmingly discuss what stopped a developed answer reaching the top band, not what a bare Level 1 script looked like. The verbatim descriptor language for this boundary is accurate (§2); no real "just misses it" illustration for the very bottom rung is claimed here.
Evaluate, Level 1 → Level 2 is similarly undocumented by a real graded example in the sources checked for this pass — flagged rather than filled with an invented one.
This document's Assess and Evaluate illustrations lean on five real anchors (Superdry, Vietjet Air, Unilever, Virgin Group, market segmentation) drawn from the facts bank's reviewed set — primarily June 2019 through June 2024, plus the two series independently checked in forge-exam-technique-guide.md. It does not claim these five boundaries hold identically in every series this paper has ever run; WBS11-verified-facts.md §0's own denominator (6–7 series for most spec points, up to 19 for a few specifically re-audited ones) is this document's denominator too.