AI Sales Call Prep in 2026: 15 Minutes of Research Per Prospect

What you get

• Cut prep time from 45-60 minutes to 15 minutes per prospect
• Produce a consistent one-page brief: company, person, trigger, three discovery questions
• Reference a specific, current fact in the first two minutes of every call
• Lift discovery-to-next-step conversion by opening on the prospect's actual situation, not a generic pitch
• Prep 4-6 calls in the time manual prep took for one

Step-by-Step Workflow

5 steps · 1 hour to set up · 15 min/call ongoing

Workflow at a glance

5 steps · 1 hour setup

13 min

Pull firmographics

Apollo.io

24 min

Research pass

Perplexity

35 min

Build brief

Claude

42 min

Review prior call

Fathom

51 min

Final skim

1

Pull firmographics

3 minApollo.io

2

Research pass

4 minPerplexity

3

Build brief

5 minClaude

4

Review prior call

2 minFathom

5

Final skim

1 min

1. 1

   Pull the firmographic layer from Apollo

   Open the account in Apollo and grab the structured facts a brief needs: employee count and growth trend, funding stage and last round date, industry, tech stack, and where your prospect sits in the org chart. Note who reports to them and who they report to; that tells you whether you are talking to the decision maker, a champion, or a blocker.

   This is the layer AI research is worst at (it hallucinates headcount and funding), so pull it from a real data source first and treat it as ground truth for the rest of the brief.

   

   3 minOutput: Firmographic block: size, funding, stack, org positionTools: Apollo.io

   Tip: If Apollo shows the prospect changed roles in the last 90 days, lead with it. New-in-seat buyers are the most receptive to change and the most likely to have budget to prove themselves.

2. 2

   Run a sourced research pass in Perplexity

   Ask Perplexity a tight, sourced question rather than a vague one. Prompt: 'What has {company} announced or been in the news for in the last 90 days? Include product launches, funding, executive hires, and any stated strategic priorities. Cite sources with dates.'

   Then a second pass on the person: 'What has {prospect name}, {title} at {company}, posted or been quoted on publicly in the last 60 days?' Perplexity's sourcing matters here; you want facts you can verify and quote, not a confident summary that might be stale.

   Copy the cited findings into a scratch doc. You are gathering raw material for Claude to synthesize, so do not edit yet.

   

   4 minOutput: Dated, cited research on company and personTools: Perplexity

   Tip: Discard anything Perplexity returns without a source or date. Uncited claims in call prep are how you confidently reference a fact that stopped being true six months ago.

3. 3

   Synthesize a one-page brief with Claude

   Paste the Apollo firmographics and the Perplexity findings into Claude with a fixed prompt (see the sample below) that outputs a one-page brief in the same structure every time: two-line company snapshot, the single most relevant trigger, likely priorities for this role, three discovery questions tailored to the trigger, and one risk or objection to expect.

   Because the structure is fixed, you learn to skim it in 60 seconds. Because Claude is synthesizing your ground-truth data rather than researching from scratch, it does not hallucinate the facts; it organizes them and proposes questions.

   Save the prompt as a template so every brief comes out identical in shape.

   

   5 minOutput: One-page brief in a fixed, skimmable structureTools: Claude

   Tip: Tell Claude to write discovery questions that assume the trigger is true, e.g. 'Now that you have raised the Series A, how are you thinking about...'. Questions that reference the trigger prove you did the work and open real conversation.

4. 4

   Review the prior call if this is a follow-up

   For a second or third meeting, spend two minutes on what actually happened last time. If your team records calls, open the prior recording in Fathom or Gong and read the auto-summary: what they committed to, what objection surfaced, what you promised to send.

   Nothing loses a follow-up faster than re-asking a question they already answered or forgetting the objection they raised. The summary gives you continuity without re-watching 40 minutes of video.

   Add the two or three carry-over items to the top of the brief so they are the first thing you see.

   2 minOutput: Carry-over items and open objections from the last callTools: Fathom

   Tip: If you do not record calls yet, at minimum keep two lines of notes per call in the CRM. The follow-up prep is only as good as what you captured last time.

5. 5

   Skim the brief in the 60 seconds before the call

   Right before the call connects, skim the one page: the trigger, the three questions, the one objection to expect, and any carry-over items. That is the whole point of a fixed structure; you internalize it in a minute and walk in able to reference something specific and current.

   Do not read from the brief on the call. Use it to open on their actual situation ('Saw you just launched X, congrats, how is early adoption?') and to steer discovery, then let the conversation go where it needs to.

   After the call, drop two lines of outcome back into the CRM so the next prep is fast.

   1 minOutput: A warm, specific open and three ready discovery questions

Why 15 minutes beats both zero and sixty

Zero prep loses the call: you open generic, the prospect mentally files you as another rep who did not do their homework, and discovery turns into an interrogation. Sixty minutes of prep does not scale past a couple of calls a day and most of that hour is spent gathering, not thinking. The 15-minute version uses AI to do the gathering in two or three minutes, spends the rest on the part only you can do (deciding what the trigger means for this buyer and what to ask), and produces a consistent artifact you can skim in the 60 seconds before the call connects. The goal is not more research. It is the right one page, every time, fast enough that you never skip it.