CONVERSION

Objection inventory from sales calls

Use when you have sales call transcripts or notes and want a ranked list of objections the site should answer.

objection-inventory-from-sales-calls.md

Download .md

You are extracting buyer objections from sales conversations and mapping them to the website.

CALL TRANSCRIPTS OR NOTES: {{TRANSCRIPTS}}
NUMBER OF CALLS: {{CALL_COUNT}}
DEAL OUTCOMES IF KNOWN: {{OUTCOMES}}
CURRENT SITE PAGES AVAILABLE: {{PAGE_LIST}}

Produce a table: Objection (in the buyer's own words where possible) | How many calls it appeared in | Stage it came up | Currently answered on site? (Yes / Partly / No, with page) | Where it should be answered | Suggested treatment (copy, proof, FAQ, calculator, comparison).

Rank rows by frequency, then by whether the deal was lost.

Then give:
- The three objections that appear most in lost deals and are least addressed on site.
- Any objection that the sales team answers well verbally but the site never mentions.
- Any objection that suggests a product or pricing problem rather than a messaging problem, clearly separated.

Constraints: quote buyers only from the material provided. Do not merge two objections that sound similar but have different root causes - keep them separate and note the distinction. If outcomes were not supplied, say the lost deal ranking is unavailable rather than inferring it. No em dashes.

Fill in before running

Replace each placeholder with your own detail. The more specific you are, the less the model invents.

{{TRANSCRIPTS}}
{{CALL_COUNT}}
{{OUTCOMES}}
{{PAGE_LIST}}

Getting a better result

Include won calls as well as lost ones, so you can see which objections get overcome and how.
The "answered verbally but not on site" row is usually the highest value output here.
Feed the top three objections straight into the trust signal audit prompt.

Questions about this prompt

When do I mine calls rather than survey visitors?

When transcripts are sitting unused and the site is arguing against objections nobody actually raises. Sales conversations are the cheapest objection research available and they are already recorded. Use the exit survey prompt instead when the people you care about never speak to sales, which is most self serve and ecommerce traffic.

What do I need in front of me before running it?

Transcripts or notes, the number of calls, the deal outcomes and the list of pages you have. Include won calls as well as lost ones, so you can see which objections get overcome and how. Without outcomes it will say the lost deal ranking is unavailable rather than inferring it, which is the right behaviour.

What comes back, and which part is worth acting on?

A table of objections in the buyer wording, with frequency, stage, whether the site answers it and where it should be, ranked by frequency and then by lost deals. The row to act on is the objection sales answers well verbally that the site never mentions, because the answer already exists and only needs publishing.

What is the mistake that costs me here?

Merging objections that sound alike. The prompt keeps them separate on purpose, since the price is too high and I cannot justify this internally need entirely different answers. Watch the section separating product and pricing problems from messaging problems as well: those belong to a different team and copy will not shift them.