ACCA APM September 2026: Why "Critically Compare" Isn't Two Lists (The 12-Mark Trap)

"Critically compare" does not mean describe A, then describe B. If your APM answer runs two techniques in two separate blocks, you can score the technical marks and still walk away with nothing from the professional skills allocation.

APM came in at 42% in June 2026 — second lowest of the whole sitting, behind only AAA at 39%. It isn't because students don't know the models. It's because they present them in the wrong shape.

What the examiner keeps saying

Take the 12-mark requirement to critically compare zero based budgeting with an existing incremental approach. The examining team noted that most candidates offered an evaluation of both techniques — often a strong one — but never actually compared them. The reader was left to do the comparing. Those scripts got technical credit and scored none of the professional skills marks in that area.

The same collapse happened on a 6-mark part asking whether kaizen costing would fit alongside ZBB in controlling production costs. Very few candidates spotted the obvious tension: kaizen is incremental and builds on past data, ZBB is fundamental and deliberately ignores past data. Most wrote that both focus on cost reduction — true, and worth almost nothing. Scepticism marks went unawarded across the board.

Behind both is the line the APM team has repeated for years: justify the model against this scenario. Why does this technique matter to this company, right now? Listing a model's advantages and disadvantages is not that. Neither is defining the terms — at APM level, definition and identification earn close to zero.

One more, and it costs students marks every sitting: the word manipulation. The examining team has flagged it repeatedly. Accounting judgement is not manipulation. Unless the scenario shows something genuinely unethical, using the word signals you're reciting a rote-learned criticism rather than reading the case.

Wrong answer vs right answer

What most candidates write: "The difference with ZBB is that it starts from zero. This will help control cost as there will be no budgetary slack carried forward."

The point about slack is fair. But there is no comparison, no mechanism, and no company in it. Half a mark, maybe.

What scores: "The decision package sits at the centre of ZBB — it details every activity requiring funding, and each item must be justified. That contrasts directly with the incremental approach, where last year's budget is the base and adjustments are made for inflation. Under ZBB every cost is scrutinised and justified, which speaks straight to Belivat's stated aim of driving costs down."

Same knowledge. But it names the mechanism, sets it against the alternative in the same breath, and lands on the company's own objective. That is a critical comparison, and it is where the analysis and evaluation marks live.

Three things to do before September

1. Answer by theme, not by technique. When two models are in play, pick 3–4 themes — starting point, time to prepare, treatment of innovation, effect on behaviour — and run both models through each theme in turn. One paragraph per theme, both techniques inside it. Never one section per model.

2. When asked how two techniques interact, lead with the conflict. Your first sentence should name the fundamental incompatibility or fit, not the shared objective. "Both reduce cost" is the answer everyone gives. The marks are in what happens when you try to run them together.

3. Delete "manipulation" from your vocabulary. Replace it with "subjective" or "open to legitimate interpretation" — unless the scenario hands you actual misconduct. Same point, no penalty, and it reads like a professional wrote it.

The numbers

APM: 42% in June 2026. Twenty of the 100 marks are professional skills — communication, commercial acumen, analysis and evaluation, scepticism — and they are awarded largely for doing what the requirement actually asked. An evaluation is not a list. A comparison is not two descriptions. A recommendation is not a balanced discussion.

You almost certainly know enough to pass APM. The question is whether the marker can tell.