Benefits Learning
AI Foundations · Lesson 2

Prompting: How to Talk to AI to Get Better Results

This lesson teaches the mechanics of asking. You will learn how to give AI clear instructions, how to steer it when the first answer misses, and how to ask for results in a shape you can actually use.

Structure the askUse context, task, constraints, and deliverable to make prompts clearer.
Improve the draftTreat the first output as a starting point, then refine it through follow-up prompts.
Work in stagesBreak complex tasks into outline, draft, critique, revision, and verification steps.
Stay organizedUse one chat for one project and reset context when the topic changes.
1

The Big Idea

Good prompting is not about magic words. It is clear communication, useful context, and good follow-up questions.

The quality of what you get out of AI is heavily influenced by the quality of what you put in. People who think AI is hit or miss are often giving hit-or-miss instructions. There are no magic words. Good prompting means saying what you want, giving enough context, setting the constraints, and describing the finished product.

When the answer is not right, you do not have to start over. You can tell it what to change. Prompting is a conversation, not a vending machine.

Core mental model

Do not think of a prompt as a one-time command. Think of it as the first turn in a working conversation.

2

The Four-Part Prompt Structure

A strong prompt usually contains four things. You do not need to label them every time, but it helps to know they are there.

1

Context

The situation and who the answer is for. Example: "I'm a benefits advisor preparing for an open enrollment meeting with a 40-person manufacturing company."

2

Task

What you actually want done. Example: "Draft a short explanation of why their premiums went up this year."

3

Constraints

The boundaries, such as length, tone, reading level, what to include, and what to avoid. Example: "Keep it under 200 words and do not promise specific savings."

4

Deliverable

The format of the output. Example: "Give it to me as three short paragraphs I can paste into an email."

Weak prompt
Explain premium increases.
Stronger prompt
I'm preparing for an open enrollment meeting with a 40-person employer. Draft a short explanation of why premiums increased this year. Keep it under 200 words, use a neutral tone, avoid jargon, and do not promise specific savings. Format it as three short paragraphs.
3

Techniques That Do Most of the Work

Once you understand the basic structure, a handful of prompting techniques make AI much more useful.

Iteration and refinement

Treat the first output as a draft. You can say, "Good, but make it warmer and cut the jargon." Each turn gets you closer.

Role prompting

Telling the AI who to be shapes the answer. "Act as a compliance-minded benefits consultant reviewing this for risk" produces a different response than no role at all.

Few-shot prompting

Show, do not just tell. Give it two examples of the style you want and ask for a third in the same style.

Structured outputs

Ask for a table, checklist, numbered steps, or specific headings so the answer is organized instead of a wall of text.

Self-critique

Ask the AI to find weaknesses in its own draft, then rewrite it. The second version is often noticeably better.

Ask what is missing

Before it answers a complex question, ask what information it needs. This catches gaps before they become wrong answers.

4

Prompting in Stages

Beginners often try to get the final answer in one big prompt. More complicated work is usually better when you break it into stages.

A staged workflow

  • Ask what information is needed.
  • Provide the missing details.
  • Ask for an outline.
  • Ask for a first draft.
  • Ask the AI to critique the draft.
  • Ask for a revised version.
  • Ask what should be verified before use.
Simple stage-setting prompt

"Before drafting anything, ask me up to five questions that would help you create a better answer."

5

Keeping the Conversation Organized

AI tools use the current conversation as context. That can help when you are working on one project, but it can create confusion when the topic changes.

Use one conversation for one project or closely related task. If you start with COBRA, then jump to Medicare, then ask about employee contributions, the AI may blend context from earlier parts of the conversation in ways you did not intend.

If you are changing subjects, starting a new chat is often cleaner. If you stay in the same chat, remind the AI what the current task is and what prior context it should ignore.

For this next task, ignore the earlier topic and focus only on the instructions below.
6

Examples from Benefits Work

These examples show the prompting techniques without turning this lesson into the client-communication lesson.

Structured promptTurn "tell me about HSAs" into a one-page internal reference on HSA eligibility rules, grouped by topic.
Role promptingAsk AI to act as a skeptical compliance reviewer and flag anything that could be misread as advice or a guarantee.
Few-shot promptingPaste two summaries you like, then ask AI to summarize a third document in the same style.
Self-critiqueAsk AI to critique a checklist for gaps, then produce an improved version.
Structured outputTurn a benefits topic into a 10-item readiness checklist with a yes/no for each item.
What is missingBefore answering an eligibility question, ask AI what details about the employee and plan it would need.
7

Common Mistakes

Most prompting mistakes are really communication mistakes. The fix is to add context, break the work into steps, and ask for the format you need.

Mistakes to avoid

  • Typing one short prompt and accepting a mediocre answer.
  • Overloading one prompt with five unrelated tasks.
  • Forgetting to set a format.
  • Assuming AI knows your context.

Better habits

  • Refine the instruction after the first draft.
  • Break complicated tasks into stages.
  • Ask for a checklist, table, outline, or other specific deliverable.
  • Give the details that matter.
8

Copy and Paste Prompt Templates

Keep these nearby. They are simple, but they solve many of the problems beginners run into.

Basic prompt skeleton
Context: [who you are, who this is for, and the situation]. Task: [what you want done]. Constraints: [length, tone, reading level, anything to include or avoid]. Deliverable: [the exact format you want].
Refinement follow-up
That's close. Keep [what worked], change [what did not], and make it [the adjustment].
Self-improvement prompt
Before finalizing, list the three weakest parts of this draft and rewrite it to fix them.
Ask questions before answering
Before answering, ask me up to five questions that would help you give a better, more accurate, or more useful response.
Staged workflow prompt
Help me work through this in stages. First, identify what information is needed. Then create an outline. After I approve the outline, draft the answer. After that, critique the draft and revise it.
Conversation reset prompt
For this next task, ignore the earlier topic and focus only on the instructions below.
9

Practice Assignment

The quickest way to understand prompting is to compare a weak prompt with a stronger one.

Take a deliberately weak prompt and rebuild it using the four-part structure. Start with: "Explain COBRA." Then create a stronger version that specifies the audience, scope, format, tone, and what should be verified before use.

Run both the weak version and your rebuilt version, and compare the two outputs side by side. For extra practice, do the same assignment again using a staged approach. Ask the AI what information it needs before answering, answer its questions, ask for an outline, and then ask for the final draft.

What to notice

The improved prompt should not just sound better. It should produce an answer that is easier to use, easier to review, and easier to turn into a real work product.

10

What to Verify

Better prompting improves quality, but it does not guarantee accuracy. A well-written prompt can still produce a confidently wrong answer.

Whenever the result touches compliance, eligibility, coverage, cost, deadlines, or anything a client will act on, verify it against a reliable source. Lesson 5 gives you the method for doing that quickly.

11

Continue the AI Foundations Series

This is the second lesson in the AI Foundations series. Each lesson stands on its own, but the sequence is designed to build from basic use to safer workflows, knowledge systems, simple tools, and agentic AI.

Lesson What It Covers Why It Matters
✓ Start HereAI Crash Course: The Practical Starting Point The orientation resource for the AI Foundations series. It introduces the main concepts, common uses, risks, tools, and next steps. Members get the big picture before moving into focused lessons.
✓ Core SkillsPrompting: How to Talk to AI to Get Better Results Clear instructions, follow-up prompts, structured outputs, examples, rewrites, staged prompting, and revision workflows. Better prompts produce better explanations, summaries, checklists, emails, and training material.
Core SkillsDocument Summaries: Turning Long Files Into Useful Takeaways How to summarize PDFs, carrier updates, compliance notices, transcripts, renewal packets, and long articles without losing important details. Long documents become easier to turn into action items, questions, and items to verify.
Core SkillsClient Communication: Explaining Complicated Topics Clearly Using AI to draft emails, talking points, open enrollment explanations, renewal summaries, and plain-English handouts. Clearer communication saves time and helps readers understand what they need to do next.
Core SkillsUsing AI Safely: Accuracy, Privacy, and Verification Hallucinations, current information, source checking, privacy, sensitive data, consumer vs. business tools, and human review. AI is useful, but benefits work requires careful verification and responsible data handling.
Deeper UnderstandingAI Basics and Python: A Look Under the Hood A beginner-friendly explanation of Python, APIs, tokens, models, transformers, natural language processing, and how modern AI tools became possible. You do not need this knowledge to use AI well, but it helps remove the mystery.
Deeper UnderstandingCapturing Knowledge: Turning Notes, Questions, and Experience Into Reusable Assets How to collect examples, recurring questions, explanations, stories, notes, and workflows so they can become reusable material. Better captured knowledge leads to better training, better prompts, and better future AI tools.
Deeper UnderstandingKnowledge Bases and RAG: How AI Uses Trusted Source Material How curated source material, atomic chunks, embeddings, retrieval, source labels, and answer checks make AI answers more grounded. This is the foundation for more reliable AI-assisted research and knowledge tools.
Building and ApplyingAI-Assisted Tools: How Calculators, Checklists, and Resources Get Built Tool scope, inputs, logic, outputs, warnings, validation, testing, versioning, and user experience. You do not have to become a programmer to understand how useful tools are planned, tested, and improved.
Building and ApplyingAgentic AI: The Future That Is Closer Than You Think How AI is moving from answering questions to helping complete multi-step workflows with source checks, permissions, approval points, and audit trails. This points toward the future while reinforcing guardrails, privacy, source discipline, and human review.
12

Final Takeaway

Prompting is the gateway skill. It is how you turn AI from a generic chat box into a useful assistant.

Clear prompts, useful context, staged work, and good follow-up questions will improve almost every AI task you try. The goal is not to memorize formulas. The goal is to communicate clearly enough that the AI can help you produce something useful.