Technique Masterclasses

Business Paper 2 — Managing Business Activities

Examiner Intelligence

*Provenance (written 2026-09-28): built primarily from direct, fresh extraction of the primary-source Pearson examiner-report PDFs themselves, not only from summaries of them. 15 examiner reports read in full this pass, text-extracted with pdftotext -layout from two local archives — /Users/neosteinhoff/ Documents/SOL/Old/P2 Materials/Business/ (October 2019 → October 2024, 13 readable reports; June 2019 and June 2024 both fail extraction — a corrupted xref table and a broken font-encoding layer respectively, confirmed independently in this pass, matching research/veridian/WBS12-verified-facts.md's own note) and ~/Downloads/PAPERI/IAL_Unit2_Complete_2024_2025/IAL_Materials/Business/ (January 2025 and June 2025, both readable — the same two series forge-exam-technique-guide.md in this folder independently opened for its own, differently-scoped pass). 15 of 17 series on file are readable; June 2019 and June 2024 are not, for either this document or any other WBS12 document in this repository. research/veridian/WBS12-verified-facts.md (1,222 lines), forge-exam-technique-guide.md, forge-error-pattern-tracker.md, signal-frequency-and-boundaries.md and codex-definitions-and-application-bank.md were all read in full first to establish what is already covered, so this document does not restate it — see "What this adds" below. October 2025's question paper is on file locally but its examiner report is not yet published; not used here.

What this document adds that nothing else in this folder does: every existing WBS12 document analyses this paper's content — what topics get confused, what technique earns marks, how often a spec section has been examined. This document instead treats the 15 examiner reports as a body of writing in their own right and asks what their own recurring language reveals about how they are produced and what that implies for how scripts are actually read. Concretely: which sentences Pearson reuses almost word-for-word series after series (not just the same underlying point — the same sentence), what changed between the reports that do and the ones that don't reuse a given formula, and what a report's frozen boilerplate says about where examiners have already decided the real dividing line between a strong and a weak answer lies, before a single script from that series was marked. No trap, content gap, or command-word rule already documented in this folder's other four files is repeated here as new content — each is cross-referenced by file name only where it supports a language-pattern finding.

Citation key: quotations are kept under ~15 words, attributed to a named series, and checked against the extracted primary-source text directly (not against this course's own prior summaries of it). Recurring sentences that run longer are described and paraphrased rather than reproduced in full — reproducing a whole Pearson paragraph, even once, is the kind of large verbatim copying this course's standing rule exists to prevent. "Confirmed in N of 15 readable series" is this document's own denominator, distinct from the verified-facts bank's "13 of 15" (that file's window stops at October 2024; this document's window runs through June 2025). A pattern confirmed in only one series is labelled a single finding below, not generalised as a pattern — see the "Single findings, not patterns" section near the end.

WBS12 SIGNAL — Examiner Intelligence

Managing Business Activities (Unit 2). Not a technique guide and not a trap list — both already exist in this folder. This is a study of the examiner reports themselves as a recurring, semi-templated document, and of what that template reveals about how the people writing it are actually reading scripts.

1. The reports have a frozen skeleton, and it splits into two families

Every one of the 15 readable reports follows the same shape: an opening general-commentary section, a "Report on individual questions" section working through Q1 and Q2 part by part and Q3 as a whole, and a closing bulleted advice list. That much is unsurprising for a templated exam-board document. What's more useful is that the template itself comes in two distinct, stable versions that never mix, split cleanly along the January/October vs June line:

The practical implication is simple and checkable by anyone reading a report cold: if it's a June report and you're scanning for the boxed technique tips the January/October format trains you to expect, you'll miss them — they're there, just written as ordinary sentences inside the question-by-question commentary instead of set apart. The two formats carry the same kind of advice; they package it differently, consistently, and the packaging has not crossed over between the two sittings in any of the six series checked from each side.

2. The opening formula: "well prepared," then "generic" — confirmed in all 15 readable series

Every single readable report — without exception, checked directly rather than assumed — opens its general commentary with some variant of the same two-part claim: candidates were broadly "well prepared" for the paper's topic areas, but "some" (January/October's usual word) or the equivalent June phrasing named areas where that held less true. The sentence is not merely making the same point series after series; in the January/October family specifically it is close to the same sentence, re-typed with small substitutions each time (June 2019 and June 2024 can't be checked; the other 12 in that family all carry it).

What follows immediately after that opening line, in every one of those 12 January/October reports checked, is a second sentence contrasting the strongest candidates — who related their knowledge to the evidence in front of them — against weaker ones, who "typically answered questions with a more generic approach." That exact four-word phrase, "a more generic approach," appears in some form in 12 of the 12 January/October reports and, worded slightly differently but making the identical contrast, in the June reports too. Across all 15 readable reports combined, some form of the word "generic" appears 70 times — an average of roughly 4–5 uses per report, or about once per 500 words of report text.

The implication worth taking seriously: examiners are not describing "generic" as one mistake among many — it is the paper's single most-repeated diagnostic word, deployed at every tariff from the 8-mark Discuss up to the 20-mark Evaluate, and it is the very first substantive judgement every report makes about the whole candidate cohort, before any individual question is discussed. The dividing line the examiners have already decided to report on, before reading a single script that series, is specific-and-applied versus generic-and-unapplied — not "correct vs incorrect content." A student whose answer is accurate but written at the level of a textbook definition, never anchored to the extract's own numbers, business or context, is describing precisely the failure mode this recurring word names.

3. The single most stable sentence in the whole corpus: the Discuss "two-sided argument" tip

If one sentence had to represent "what Pearson keeps telling candidates, in almost the same words, for six years running," it is the standing tip on the Discuss command word. Checked directly in every one of the 15 readable reports (the only rule in this entire document confirmed literally 15 for 15): a boxed or in-prose reminder that "the command word 'discuss' requires a two-sided argument" and that a candidate who does not provide one, or who gives a "generic answer," will not reach the higher levels.

The opening clause of that sentence — command word, two-sided argument requirement — is close to word-for-word identical across October 2019, January 2020 and October 2020 specifically (independently checked side by side this pass); only the closing clause describing the consequence of failing to meet it varies release to release (worded across different series as restricting the mark, not reaching the higher levels, not meeting the full level descriptors, or not receiving full marks — four slightly different closing clauses attached to the same opening diagnosis, not four different rules). The "no conclusion required for an 8-mark Discuss" line — already documented as a cross-series constant in WBS12-verified-facts.md — is standing content inside this exact same recurring sentence, not a separate finding.

What this implies: of everything Pearson could choose to repeat almost verbatim series after series, this is what they chose. A command-word rule this stable, this often reinforced, in a report otherwise rewritten substantially every series, is Pearson's own signal of where the highest volume of marks is realistically being lost across the whole paper — not a minor footnote.

4. Two eras of "lacked understanding": a real, dated shift in how content gaps get reported

The clearest evidence in this whole corpus of the report-writing formula itself changing, not just being reused, sits in how the 20-mark Question 3 commentary describes a genuine content gap (as opposed to a technique failure). Two distinct templates were found, cleanly split by date, each confirmed in at least two series and neither found outside its own window:

Template A — confirmed in five series, October 2021 through October 2023 (Oct 2021, Jan 2022, Oct 2022, Jan 2023, Oct 2023): a fixed three-part sentence structure — candidates struggled to apply the extract or balance their argument; "some candidates lacked understanding of" the series' specific weak topic (quality assurance/control, efficiency, economic influences, liquidity and working capital, or business-failure causes, depending on the series); and a closing line that a general discussion or a rewritten extract, rather than an answer to the actual question, did not let the candidate progress through the levels.

Template B — confirmed in three series, January 2024 through January 2025 (Jan 2024, Oct 2024, Jan 2025): a different, more specific two-part structure — "some candidates lacked understanding of" the weak topic (short product lead-in times, factoring, or the business cycle) "and so ignored that part of the question" or guessed at it; then, separately, a second group of candidates who did know the topic but listed its advantages and disadvantages without ever linking them back to the actual evaluative ask.

The genuinely interesting fact is not that both templates exist — it's what changed between them. Template A collapses "didn't know it" and "knew it but didn't use it for this question" into one diagnosis. Template B explicitly separates the two into different candidate groups with different fixes: one group needed the content itself; the other group had the content and needed to connect it to the question's specific evaluative demand. That is a real, dateable sharpening of diagnostic precision in how Pearson describes this exact failure mode, sitting on the paper's highest-value question, and it happened at a specific, identifiable point (between the October 2023 and January 2024 reports) rather than gradually. For a student: a report from 2024 onward that says you "lacked understanding" of something is making a narrower, more specific claim than the identical-sounding phrase in a 2021–2023 report would have made about the same failure.

5. "Especially pleasing to see": examiners started explicitly rewarding exam-technique preparation, from October 2021 onward

A second dated, genuine shift, distinct from Template A/B above and confirmed independently: starting with October 2021 and January 2022, and recurring in six further series through January 2025 (Jan 2023, Oct 2023, Jan 2024, Oct 2024 all confirmed; October 2022 confirmed with "practise" spelled the British way, a minor copy-edit variant of the same sentence) — eight series total from October 2021 onward — the opening general commentary adds a new observation that is genuinely absent from every earlier report checked (Oct 2019, Jan 2020, Oct 2020, Jan 2021 all lack it): that it was "especially pleasing" to see evidence candidates had used past papers to practise, and that "note had been taken... of previous examiner reports."

This is worth stating plainly rather than just logging: from October 2021 onward, Pearson's own examiner reports explicitly and repeatedly single out, as a positive marker of candidate preparation, the specific behaviour this course's own resources exist to support — working through real past papers and reading what past reports actually said about them. That is not a claim this document can extend into a promise about any individual candidate's outcome, and it says nothing about how many marks that preparation is worth — but it is a real, six-times-repeated observation from the exam board's own writing about what "well prepared" has increasingly come to mean in its own account of the phrase, not an inference this document is making on Pearson's behalf.

6. The closing advice list: frozen since October 2019, edited only by addition

Every readable report closes with a short bulleted list of standing reminders — headed "Summary" in the earlier reports on file (Oct 2019, Jan 2020) and "Paper Summary" in later ones (Oct 2022, Jun 2022, Jun 2023, Jun 2025) — and the list itself has stayed remarkably stable. At least seven bullets recur, in substance and largely in wording, across every readable series checked at both ends of the six-year window: read the whole question and follow every stated requirement; understand what the command word itself requires before answering; Quantitative Skills are tested throughout, not confined to calculation questions; application marks require the extract's evidence to be used, not repeated; Discuss specifically needs a two-sided argument (the same standing rule as §3 above, restated here as the closing list's own version of it); the answer booklet may provide more space than is needed; and the extracts exist to be used, not merely read.

Two small, checkable facts about how this list has actually changed over time, not assumed from its stability: (1) October 2019 and January 2020 — the two earliest readable reports — share the literal same typo, "offered te following" for "offered the following," fixed by January 2021 onward; a small, almost certainly meaningless detail on its own, but a clean confirmation that this list really is a copy-edited template carried forward from one series' document into the next, not independently retyped each time. (2) The bullet stating that WBS12 questions may draw on WBS11 (Unit 1) content is a later addition: checked directly across every January/October report on file, it is present, worded near-identically, in all seven consecutive series from January 2022 through January 2025, and absent from all five January/October reports checked before it (Oct 2019, Jan 2020, Jan 2021, Oct 2020, Oct 2021 — zero mentions of WBS11 or Unit 1 in any of the five) — a clean, dated addition to the closing list, not a rule that was always there. Consistent with WBS12-verified-facts.md's own independent dating of this exact rule's wording shift "from Jan 2023 onward" for the phrasing used elsewhere in the same reports (this document's finding is about when the closing-list bullet itself first appears, a related but distinct fact, not a restatement of that file's finding).

For a student, the honest reading of a list this stable for this long is: these seven-plus items are Pearson's own version of "if a candidate only internalises one page from six years of reports, this is the page" — and the fact that new items get added to it (never removed) means it is also the most conservative, lowest-risk single list in this whole corpus to actually memorise.

7. What doesn't recur: single findings, not patterns

Three things surfaced during this pass that looked, on a first read, like they might be recurring patterns and turned out, on checking every series rather than a sample, not to be. Flagging them here is itself part of this document's standing discipline — a pattern seen once is a single finding, not a trend, and should be labelled as such rather than written up as if it generalises:

8. Corpus gaps, stated plainly

June 2019 and June 2024 remain unreadable in this pass, exactly as in WBS12-verified-facts.md — confirmed independently here rather than taken on trust: June 2019's PDF has a broken cross-reference table and June 2024's has a font-encoding layer that garbles all extracted text to symbol noise. Neither gap is filled by this document. October 2025's question paper is present in the local archive but its examiner report is not yet published as of this writing (2026-09-28) and is not used here. Every claim above is stated against its own real denominator — "of the 15 readable series," not "of all 17" — for exactly this reason.

Sources

Primary: the 15 examiner-report PDFs themselves, text-extracted directly for this pass with pdftotext -layout from /Users/neosteinhoff/Documents/SOL/Old/P2 Materials/Business/ (Oct 2019–Oct 2024) and ~/Downloads/PAPERI/IAL_Unit2_Complete_2024_2025/IAL_Materials/Business/ (Jan 2025, Jun 2025). Cross- checked against, and confirmed not duplicating: research/veridian/WBS12-verified-facts.md, forge-exam-technique-guide.md, forge-error-pattern-tracker.md, signal-frequency-and-boundaries.md and codex-definitions-and-application-bank.md, all in this folder or its parent, all read in full before writing. No content, quote or claim in this document is carried over from the pre-audit old Veridian build; none of that material was consulted for this pass.