Claude Certification Guide CCAO-F

Curriculum › Workflow Integration and Solution Design · 16% of the exam

Research, planning and solution design with Claude

What you’ll be able to do

  • Match a research or planning question to the right Claude feature
  • Map an existing process with Claude and get improvement ideas grounded in that process
  • Run a draft-test-revise loop on a plan, template or prototype

What you need to know

Research: pick the tool that matches the question

Not every question needs the heaviest tool. Claude's own knowledge in a plain chat is fine for explaining a concept, brainstorming options or structuring your thinking. Web search helps when you need something current. Research — where Claude searches across many sources and returns a cited report — fits multi-source questions such as "compare how mid-sized organizations approach hybrid-work policies" or "what are the main options for event-registration platforms and how do they differ?" It takes longer, which is worth it only when breadth matters. (Feature availability depends on your plan and your admin's settings.)

Citations make a report easier to check; they do not make it correct. A cited source may not say what the summary claims, may be outdated, or may be a vendor's marketing. Before a plan goes to decision-makers, open the sources behind the two or three claims the decision actually rests on — a price, a capability, a compliance requirement — and confirm them yourself.

It also helps to ask Claude to separate what the sources actually say from its own synthesis or inference — for example, "mark which points are directly supported by a source and which are your conclusions." That makes it easier to see where checking matters most and where the report is filling gaps with reasoning.

Planning and process optimization

Claude is a strong planning partner when you give it the real constraints: deadline, budget range, team size, dependencies, what has already been decided. Build the plan in stages rather than asking for everything at once:

  • Clarify: ask Claude what it needs to know, or what questions a planner would ask, before drafting.
  • Outline: get the phases and milestones; adjust them.
  • Detail: expand one phase at a time into tasks, owners and risks.
  • Challenge: ask Claude to critique the plan — what could go wrong, what's unrealistic, what's missing.

For process optimization, describe the current process step by step — who does what, how long it takes, where things wait or get redone. Then ask Claude to identify bottlenecks, duplicated effort and handoffs that could be simplified, and to say which step each suggestion applies to. Suggestions that could apply to any organization ("improve communication", "use automation") are a sign Claude didn't have enough detail about yours.

Solution design and iteration

"Solution design" at the Associate level means shaping things you'll actually use: an intake form, a reporting template, a training outline, a customer-reply playbook, a simple interactive checklist or calculator. Claude can draft these as Artifacts — standalone documents or simple interactive pages shown beside the chat — which suit content you'll revise and reuse.

Good design work is iterative. Treat Claude's first draft as a prototype, not an answer:

  • Test it against real cases: run three recent, real examples through the new template or playbook and see where it breaks.
  • Change one thing at a time and say why ("the approver field is missing; add it after budget").
  • Keep the work in one place. A Project with the design brief and constraints in its instructions and knowledge lets every chat start from the same context, so you don't re-explain the goal each time.
  • Get feedback from the people who will use it, and feed that back in.

Who decides

A useful habit before presenting any plan or design: ask Claude to argue the other side. "What would a sceptical finance director object to?" or "What are the strongest reasons not to choose this option?" surfaces weaknesses while you can still address them, and makes your recommendation more credible because you've already tested it.

Claude can compare options, draft the recommendation and lay out the trade-offs. Whether to fund the project, choose the vendor or launch the new process is a human decision, made by whoever is accountable for the outcome. Present Claude's work as input to that decision, with your verification and judgment layered on top.

Key concept

Research broadly, plan in stages and design in short draft-test-revise loops — and verify the facts a decision rests on before anyone acts.

Exam scenarios reward the option that matches the tool to the question (research for multi-source comparisons, not for a quick definition), builds the plan step by step, tests designs against real cases, and checks key cited facts rather than trusting a report because it has citations.

Practice scenario

ScenarioA project manager uses research to compare four project-management tools for her team. The cited report recommends one tool, noting it "includes built-in time tracking at no extra cost" — the deciding factor for her team. She has a meeting with her director in an hour and plans to paste the recommendation into her slides.
Work it through, then open this

The recommendation hinges on one specific claim. Pricing and feature bundling change often, and a cited source might be outdated or might not say exactly what the summary claims. The most appropriate step is to open the source for the time-tracking claim (ideally the vendor’s current pricing page) and confirm it before presenting. Re-running the research, asking Claude how confident it is, or dropping the research altogether don’t address the specific risk. She should also present this as a recommendation for her director to decide on, not a decision already made.

Hands-on exercise — Improve a real process in stages

Beginner · 20 min

What you’ll learn

  • Describing a process so Claude’s suggestions are specific
  • Building a plan in stages
  • Testing a design against real cases

  1. Write out a process you know well — for example, how your team handles new-starter onboarding — as numbered steps with who does each and roughly how long it takes. Ask Claude to identify bottlenecks and redundant steps, naming the step number for each.

    • Why: Step-level detail forces suggestions tied to your process instead of generic advice.
    • You should see: Suggestions that reference specific steps; any that don’t are probably filler.

  2. Pick the best suggestion and ask Claude to draft an improved version of that part of the process as an Artifact — for example, a checklist or template.

    • Why: An Artifact gives you a standalone draft you can revise and reuse.
    • You should see: A first draft you can hold up against reality.

  3. Describe two or three recent real cases (anonymized) and ask Claude to walk each through the new checklist, noting where it fails or is unclear. Revise the Artifact.

    • Why: Real cases reveal gaps that a design that looks good on paper hides.
    • You should see: At least one revision — a missing step, an unclear owner — that came from testing rather than drafting.

Exam traps

Using a cited research report without checking the claims the decision depends on

Citations make checking possible, not unnecessary. Confirm the few claims the decision actually turns on.

Asking for the whole plan in one prompt instead of building it in stages

Staged planning — clarify, outline, detail, challenge — lets you correct direction early instead of untangling a long, flawed draft.

Accepting process-improvement suggestions that are generic rather than tied to your process

Advice that could apply to any team means Claude lacked detail. Give it the real steps and ask it to reference them.

Starting a new chat for each revision and losing the design’s history

Keeping iterations in one chat or Project preserves the brief, constraints and earlier decisions.

Treating Claude’s first design draft as the final solution

A first draft is a prototype. Test it against real cases and users’ feedback before rolling it out.

Letting Claude make the go/no-go decision on a plan

Claude informs the decision; the accountable person makes it.

Sources

Quick check

Select TWO.

After two months of excellent results, a customer-success manager proposes removing the human review step from Claude-drafted renewal emails to clients. Select the two responses that best reflect sound workflow design.