Your Agency’s Context Is Your AI Advantage

Insurance agencies get better AI results by providing four layers of context: client information, agency knowledge, professional expertise, and clear task instructions. This background helps AI produce relevant drafts that reflect the agency’s work and give clients a clear next step.

Two agencies use the same AI tool to write a homeowners renewal email.

One gets a generic reminder. The other gets a draft that acknowledges the client’s situation, explains the agency’s renewal review, and invites a useful conversation.

The difference is context.

In her Catalyit session, Angela Ford makes a point: your agency’s relationships, experience, and judgment give AI something useful to work with.

What is context in an insurance agency AI prompt?

Context is the background AI needs to complete a task well. It explains who the client is, how your agency works, what should guide the response, and what the output should accomplish.

“Write a homeowners renewal email” identifies a task. It leaves AI to fill in the rest.

A stronger request includes verified account details, the work your agency completed, the client’s priorities, and the next step you want them to take.

Think about briefing a new employee. You would explain the account before asking them to contact the client. Give AI that same foundation.

What four layers of context improve AI results?

1. Client context: Who are you helping?

A name personalizes the greeting. Relevant background makes the message useful.

Depending on the task, include:

• The client’s relationship with your agency.
• Relevant life or business changes.
• Questions or concerns they have raised.
• Communication preferences.
• Whether they prioritize price, service, or peace of mind.

A client who recently experienced a claim may need a different renewal conversation from someone focused on premium.

Use verified information through agency-approved tools and processes.

2. Agency context: How do you work?

Agency context helps AI reflect your voice and service approach.

Document who you serve, how you communicate, and what your processes include. Explain what clients can expect from your team.

If your agency reviewed alternative carriers before discussing a renewal, give AI that verified fact. The draft can then explain the work behind your recommendation.

Save standing agency information so employees can reuse it across tasks.

3. Expertise context: What should guide the response?

Experienced employees know what deserves attention, which questions to ask, and when a routine request needs a deeper conversation.

Capture that reasoning.

For a renewal email, explain:

• What the employee reviewed.
• Which facts matter to the discussion.
• What still needs confirmation.
• Which conclusions require further review.
• Who approves the message before it goes to the client.

AI can help express your team’s judgment. Your team supplies and verifies it.

4. Task context: What should the output accomplish?

Define the audience, purpose, format, length, and next step.

“Draft a concise email explaining the completed renewal review and asking the client to schedule a call” gives AI a clearer assignment.

Add the details that matter. Specify plain language, a word limit, or one call to action. Tell AI to flag missing information rather than fill gaps.

What does a stronger renewal prompt look like?

This fictional example combines all four layers:

Draft a homeowners renewal email for a fictional client, Maria Garcia.

Client: Maria has been with our agency since 2014. Her household recently added a teen driver. She values clear explanations and personal service.

Agency: We use a warm, direct tone. For this example, the account manager has completed a review with three alternative carriers.

Expertise: Explain that the review is complete. Do not claim the options provide identical coverage or guarantee savings. Invite a conversation about the household change and whether an umbrella coverage review would be useful. Do not invent premiums, limits, or recommendations.

Task: Keep the email under 200 words. Ask Maria to schedule a 15-minute call. Flag missing information for the account manager.

The prompt gives AI facts, direction, and boundaries. The account manager still checks the draft against the account record before sending it.

How can agencies reuse AI context?

Ford offers three steps: tell it, save it, and connect it.

Tell it. Include the background in your request. Use this approach to test what information improves the result.

Save it. Store reusable instructions and reference documents in an agency-approved workspace. Give employees a consistent starting point.

Connect it. Where approved tools support it, connect the information sources the workflow needs.

Start with one recurring task. Refine the context before expanding its use.

What six documents should agencies create?

Build a small reference library around the knowledge your team uses most.

Assign an owner to each document. Review it when the underlying information changes.

Outdated context can produce outdated answers.

How do you capture an experienced employee’s knowledge?

Let AI interview them.

Ford suggests spending about 20 minutes answering questions about familiar work. Choose a process the employee handles regularly.

Try this prompt:

Interview me about how our agency handles renewal reviews. Ask one question at a time about the steps, client concerns, decision points, exceptions, and review responsibilities. Organize my answers into a draft process guide. Flag gaps for follow-up.

Have the employee review and correct the guide before sharing it.

This turns everyday experience into a resource the team can use.

Start with one weekly task

Choose a task your team repeats every week. Write one page covering the client, agency, expertise, and task.

Run the request with and without that page.

Compare the results:

• Which draft needs fewer corrections?
• Which sounds like your agency?
• Which explains your work clearly?
• Which gives the client a useful next step?
• How much editing does each require?

Review the differences with your team. Refine the context page and save it for the next request.

Your agency’s knowledge becomes more useful when your team can capture it, share it, and apply it consistently. Start with one page and one task.

Frequently asked questions

Why does AI write generic insurance emails?

The prompt may lack relevant client facts, agency background, professional guidance, or a clear purpose. Supplying those details gives AI better direction.

What is the difference between a task and context?

The task tells AI what to produce. Context provides the facts and expectations that guide the result.

Should agencies include client information in AI prompts?

Use agency-approved tools and follow your agency’s information-handling requirements. Include only what the task needs. Use fictional or de-identified examples when testing workflows.

How can agencies keep AI drafts consistent?

Use shared agency instructions, approved reference documents, and task templates. Assign owners to maintain them and verify client-specific facts before use.

What is the easiest way to improve an agency’s AI results?

Choose one recurring task, write one page of relevant context, and compare the output with a basic prompt. Refine and save what works.