AI prompt to compare lead sources (internet vs referral vs walk-in)

New business starts with a steady list of people and businesses to talk to, and a fast, personal first touch. AI helps agents research prospects, draft first messages and organize lead lists, so more of the day goes to real conversations.

Below is a ready-to-copy prompt to compare lead sources (internet vs referral vs walk-in), a detailed version that asks you clarifying questions first, and 1 specialized AI tool (HubSpot) with instructions for this task. It also works in ChatGPT, Claude, Gemini or Copilot.

Try it in the search box ›

The prompt

Act as an insurance agency sales analyst. Below is a de-identified table with one row per lead source (e.g., internet/purchased leads, referrals, walk-ins, website form) and these columns: source, monthly spend ($), leads received, quotes issued, policies sold, total written premium ($), average commission rate (%), and estimated producer hours spent. For EACH source calculate: cost per lead, quote rate (quotes/leads), close rate (sales/leads and sales/quotes), average premium per sale, cost per sale (acquisition cost), commission revenue, and ROI = (commission revenue - spend) / spend; if spend is $0 (referrals, walk-ins) show revenue per producer hour instead. Present a ranked table, flag any source with a negative ROI or a close rate under half the agency average, state the assumptions you used, and recommend where to shift budget and producer time next quarter (with one test to run per source). Do not invent numbers that are not in my data; ask me for anything missing. Data: [paste table]

Detailed version (asks clarifying questions first)

Act as an experienced assistant to a licensed insurance agent. Before answering, ask me up to 3 clarifying questions (who you're reaching out to, whether it's an email, call or text, and tone). Then complete this task, and finish with a short checklist of statements I should verify with my compliance team.

Task: Act as an insurance agency sales analyst. Below is a de-identified table with one row per lead source (e.g., internet/purchased leads, referrals, walk-ins, website form) and these columns: source, monthly spend ($), leads received, quotes issued, policies sold, total written premium ($), average commission rate (%), and estimated producer hours spent. For EACH source calculate: cost per lead, quote rate (quotes/leads), close rate (sales/leads and sales/quotes), average premium per sale, cost per sale (acquisition cost), commission revenue, and ROI = (commission revenue - spend) / spend; if spend is $0 (referrals, walk-ins) show revenue per producer hour instead. Present a ranked table, flag any source with a negative ROI or a close rate under half the agency average, state the assumptions you used, and recommend where to shift budget and producer time next quarter (with one test to run per source). Do not invent numbers that are not in my data; ask me for anything missing. Data: [paste table]

What a good result looks like

A good analysis calculates every requested measure for each lead source (cost per lead, quote and close rates, cost per sale, commission revenue, ROI or revenue per producer hour for free sources), ranks them in a table, states its assumptions and flags weak sources.

Common mistakes and watch-outs

Check the math in a spreadsheet; AI arithmetic can slip. Small samples can mislead, and lifetime value (retention, cross-sell) isn't in a one-month table. Make sure the AI didn't fill in numbers you didn't provide. HubSpot can report lead sources if they're tracked consistently.

How to adapt the prompt

Run it quarterly and compare periods, and add a column for retention after 12 months if you have it. Ask for one test per source next quarter.

EXAMPLE output (placeholder data, not a real client)

Lead source comparison - [Quarter]

| Source | Cost/lead | Close rate | Cost/sale | ROI |
| Internet leads | [Amount] | [Percent] | [Amount] | [Percent] |
| Website form | [Amount] | [Percent] | [Amount] | [Percent] |
| Referrals | - | [Percent] | - | [Revenue/hour] |
| Walk-ins | - | [Percent] | - | [Revenue/hour] |

Flag: [Source] has negative ROI.
Assumptions: commission rate [Percent] applied to all sources.
Test next quarter: [Source] - shorten time to first call.

How to use it

Copy the prompt. Copy the short prompt above, or the detailed version if you want the AI to ask clarifying questions first.
Fill in the placeholders. Replace [paste table] with your own details. Use placeholders or de-identified details; never paste client personal information.
Run it in the right tool. Paste it into ChatGPT, Claude, Gemini or Copilot, or use HubSpot as described below.
Review before you use it. Use aggregated numbers, not individual client records, when you analyze results with AI. Check every fact, name and number yourself.
Take the next step. Next, see: Measure marketing ROI by channel.

Best AI tools for this task

Specialized tools first, ordered by fit for this task (never by payout), then any general AI assistant.

HubSpot site ↗

CRM and marketing platform with AI features.

How to use it

In HubSpot, ask Breeze (HubSpot's AI assistant): 'Using contacts created in the last 90 days, group deals by Original Source and show for each source: number of contacts, deals won, close rate, and total/average deal amount. Then rank the sources by close rate and average premium per sale.' Export the result and paste it, with your spend per source, into the AI assistant prompt to get cost per lead and ROI.

Any AI assistant

ChatGPT Claude Google Gemini Microsoft Copilot

Paste the prompt above (short or detailed) into any of these.

Before you use it

Use aggregated numbers, not individual client records, when you analyze results with AI.
Never paste client personal information (names, policy numbers, health or financial details) into AI tools your agency hasn't approved.
Review every AI output for accuracy and compliance before you use it.

Was this page helpful?