Rubric
Contents — domains, guide and mocks

Adjusting on feedback

CCAO-F 7.214 min read · checked 21 September 2026

Task statementAdjust approach based on feedback and results

Five layers you can change

Change the lowest layer that explains the feedback

  1. Whether to use ClaudeSome tasks should go back to a person entirely
  2. The workflowWho reviews, at which point, against what
  3. The product setupProject, knowledge, instructions, model
  4. The context suppliedDocuments, examples, audience, constraints
  5. The prompt wordingThe layer everyone reaches for first
Work from the bottom up. Most feedback is answered in the lowest two layers, but the layers above are where the durable improvements live.

Three kinds of feedback, and what each is good for

The first kind is your own reading of the output. It arrives instantly and it is the least reliable, because you know what you meant and will read it in charitably. The second is other people: the reviewer whose edits you can count, the colleague who says a draft “doesn’t sound like us”, the manager who sends it back. The third, and the one almost nobody collects, is the result in the world — did the email get replies, did the bid win, did the ticket close first time, did the reader do what the document existed to make them do.

These three often disagree, and when they do the later ones win. An output you were pleased with and a reviewer rewrote is not a good output. A document everyone admired that changed nobody’s behaviour has not done its job. The documentation makes the same point in its own register: before optimising anything, define what success looks like for your use case and have a way to test against it. Most people iterate against a feeling; the discipline is to iterate against a measure, even a rough one such as “how many lines did the reviewer change”.

SignalWhat it can tell youWhat it cannot
Your own reactionObvious errors, wrong shape, wrong lengthWhether it works for the actual reader
Reviewer editsWhere the house voice and standards really areWhether the reader acted on it
Downstream resultWhether the output did its jobWhich specific change caused the difference
Thumbs-down in the productSends the response to Anthropic as unhelpfulAnything back to you about this task
Repeat use over weeksWhich fixes were durable and worth savingMuch, if nobody wrote down what changed

Turning vague feedback into a change you can make

Most feedback arrives unusable. “It’s not quite right”, “too corporate”, “this isn’t how we write”. None of that can be acted on, and the temptation is to guess. The move is to convert the reaction into two things: an example of the failure and an example of what was wanted. A reviewer who cannot articulate the house voice can always point at a document that has it, and that document is worth more than any adjective, because documented practice is to give three to five real, diverse examples rather than describing the target in the abstract.

Feedback made actionable

Unusable feedbacktext

"This doesn't really
sound like us. Can you
have another go and
make it warmer?"

Converted to a changetext

Failing line:
"We regret to inform
you that your request
has been declined."

Wanted instead:
"We can't do this one,
and here's why."

Rule: contractions,
second person, say the
answer in sentence one.

Attach: three past
letters the team liked.
Same reviewer, same objection. The right-hand version names the failing sentence, the wanted alternative and the rule — all three of which can go straight into the next prompt.

Making an adjustment stick

An improvement that lives in one person’s scroll history is not an improvement, it is a rediscovery waiting to happen. Once a change proves itself, it has to move somewhere durable, and there is a natural ladder. A phrasing that reliably works goes into a saved prompt or template. Knowledge and conventions specific to a team’s work go into that team’s project instructions and knowledge — configuring those is 5.1, and keeping them current is 5.4. A rule that should hold across the whole organisation can go into organization instructions, which Owners on Team and Enterprise plans set within a 3,000-character limit, as covered in 6.3.

The ladder matters because each rung costs something. Organization instructions apply to every chat in the business, so a rule that suits one department will annoy five others and their character budget is finite. Project instructions are the right home for most team-level learning. Saved prompts are right for an individual’s recurring task. Putting a fix on the wrong rung is itself a common failure — and a change that helps one team and hinders the rest usually shows up as new complaints rather than as anybody reversing it.

Where should this improvement live?

Who does this improvement apply to?
  • Just my recurring task
    Saved promptA template you reuse and refine
  • My team’s work
    Project setupInstructions and knowledge in the project — see 5.1
  • Everyone, always
    Organization instructionsOwners only; 3,000 characters for the whole business
  • It keeps failing
    Change the workflowOr give the task back to a person
Match the scope of the fix to the scope of the problem. Promoting a team convention to an organisation-wide rule is how instruction files fill up with other people’s preferences.

When the answer is not another adjustment

Some loops should be closed rather than tightened. If the same failure survives three genuine rounds of adjustment, the feedback is telling you something about the task, not about your technique. The documentation acknowledges this in its own way, noting that not every failing result is best solved by prompt engineering — some are better addressed by a different choice of model, which for cost, speed and quality is 3.3. And some tasks simply belong to a person: work where the judgement is the point, where the facts are not available to Claude at all, or where the checking costs more than the drafting saved.

There is one more feedback channel worth knowing because the exam may mention it. Inside the product, the thumbs-down control sends a response to Anthropic as unhelpful, and the help documentation points users to it along with emailing support with suggestions. It is a signal to the people who build the model. It is not a support ticket about your document, and nothing comes back to you — so it complements, rather than replaces, adjusting your own approach.

Traps the wrong answers are built from

Tempting but wrongDo this instead
Answering every complaint with a prompt rewriteIdentify which layer the feedback is about — context, setup, workflow or fit.
Acting on “make it warmer” as if it were an instructionConvert it into a failing example, a wanted example and a stated rule.
Judging improvement by whether colleagues liked the draftPick an outcome measure, however rough, and track the same one over time.
Leaving a proven fix in one person’s chat historyPromote it to a saved prompt, project instructions, or organisation-wide guidance.
Putting a team convention into organisation-wide instructionsMatch the scope of the fix to the scope of the problem.
Adjusting indefinitely on a task that keeps failingAfter three honest rounds, change the model, the workflow, or who does the task.

You should now be able to

  • Distinguish feedback about the output, from reviewers, and from the result in the world.
  • Locate which layer a piece of feedback is really about before changing anything.
  • Turn a vague reaction into a failing example, a target example and a rule.
  • Choose the right home for a proven improvement, from saved prompt to organisation instructions.
  • Recognise when to stop adjusting and change the task, the model or the owner.

Practice questions

Original questions written for this lesson, in the exam’s style. Answer first, then open the reasoning — every option is explained, including why the wrong ones are tempting.

  1. Question 1

    Across a 30-person consultancy, partners report that Claude-drafted client summaries consistently miss the firm’s standard risk paragraph. Each consultant writes their own prompts in their own style.

    What is the most appropriate adjustment?

    1. AAsk each consultant to add a line about the risk paragraph to their prompts.
    2. BPut the requirement and two example summaries into the shared project setup.
    3. CGive the thumbs-down on any summary that omits the paragraph.
    4. DStop using Claude for client summaries until the model improves.
    Show answer and reasoning
    1. AIncorrect. Thirty separately maintained prompt habits are exactly what produced the inconsistency in the first place.
    2. BCorrect. A failure that survives many different phrasings is a context problem, and shared configuration is where such conventions belong.
    3. CIncorrect. That control sends a signal to Anthropic about an unhelpful response and returns nothing that changes this firm's outputs.
    4. DIncorrect. Abandoning a task is reasonable only after adjustments have genuinely been tried, and the obvious one has not been.
  2. Question 2

    A fundraising manager has rewritten her donor-appeal prompt four times. Each version reads better to her and to her colleagues. Donations per appeal have not moved at all.

    What does this pattern most strongly suggest?

    1. AThe prompt needs a fifth revision with more specific instructions.
    2. BShe is iterating against colleague approval rather than the outcome the appeal exists for.
    3. CDonation rates are outside anyone's control, so the appeals are fine.
    4. DThe task should be handed back to an external fundraising agency.
    Show answer and reasoning
    1. AIncorrect. Four rounds against the same impression-based signal is a reason to change approach, not to run a fifth.
    2. BCorrect. When output and result signals disagree, the result wins, and improvement has been drifting toward what is easy to admire.
    3. CIncorrect. Treating the outcome as unknowable removes the only measure that could tell her whether the work is helping.
    4. DIncorrect. Possible eventually, but nothing has yet been adjusted against the right measure, so the option is premature.
  3. Question 3

    An operations lead has found a phrasing that reliably produces good incident write-ups. She wants the improvement to survive her being on leave.

    Which two placements are appropriate for this improvement? (Select 2.)

    1. AThe instructions of the project her team uses for incident work.
    2. BA saved template the team can start from for each incident.
    3. CThe organisation instructions, so every chat in the company follows it.
    4. DA message to Anthropic support describing what worked.
    5. EHer own chat history, which colleagues can search.
    Show answer and reasoning
    1. ACorrect. Team-level conventions belong in the team's project, where everyone's chats inherit them.
    2. BCorrect. A reusable template preserves a specific working phrasing without imposing it on unrelated work.
    3. CIncorrect. An organisation-wide rule for one team's document type would affect everyone and spend a limited character budget badly.
    4. DIncorrect. That is a channel to the model's developers and does nothing to preserve the practice inside her team.
    5. EIncorrect. Relying on someone finding an old conversation is precisely the fragility the question asks her to fix.

Sources

Drafted with AI assistance and checked against the sources above; expert review is in progress. Spotted an error? Tell us and it gets fixed, dated and listed on how this is written.