A 3-layer framework for AI-assisted buyer prep
If you ship enough buyer-facing meetings, you start to notice that "prep" isn't one thing. It's three different things stacked on top of each other, and each layer answers a different question.
Most reps prep at one layer (usually the account) and skip the other two. That's why prep often feels useful in the moment but doesn't translate into stronger calls.
Layer 1: Account context
This is the company. Industry, size, business model, recent news, financial position if public, strategic narrative if available.
The question this layer answers: What world does this buyer live in?
A buyer at a Series A startup operates in a fundamentally different world from one at a public company in slow decline. The same product pitch lands differently. The risks they care about are different. The way they measure success is different.
Account context should take you two minutes to read, not 45 minutes to assemble. What you want:
- One-sentence what they do
- Stage / size (revenue, headcount, funding)
- One or two recent events (last 6 months)
- The strategic narrative (growing fast, restructuring, expanding internationally, etc.)
If your prep tool gives you eight pages of unfiltered facts, that's worse than nothing — you won't read it.
Layer 2: Person context
This is the buyer themselves. Role, tenure in role, career path, public activity (posts, talks, comments).
The question this layer answers: Who is this person, and what are they trying to do here?
A VP of Sales who joined six months ago is in a very different position from one who's been there four years. The new hire is building credibility, often wants quick wins, may be willing to try new things. The veteran has political capital but is also more cautious, has more allies and more rivals internally.
Tenure in role is often the single most useful signal nobody looks up.
What you want: 30 seconds of context that changes how you'd open the call. Not a dossier.
Layer 3: Meeting context
This is the specific meeting. Why are you talking now? What was the last touch? What did they ask for? What did you commit to send? Is this a first call, a follow-up, a redirect to a different champion, a renewal conversation?
The question this layer answers: What does success look like in the next 30 minutes?
This is the layer reps most often skip — and it's the one that actually determines whether the meeting moves the deal forward. A first discovery has one shape; a champion conversion call has another; a procurement-led pricing review has another.
What you want: one or two lines that frame the meeting's purpose and the one thing you most need to learn or land.
What to feed AI vs. write yourself
Layers 1 and 2 are good fits for AI synthesis. The information exists publicly or in your CRM and notes. The work is collection and compression — exactly what language models are good at.
Layer 3 is mostly your job. AI can structure the meeting type and pull last-touch notes, but the strategic framing of "what would make this meeting matter" comes from you. You know the deal context; the model doesn't.
Use AI where the work is mechanical. Keep the strategic judgement.
A simple workflow
Before every external meeting, give yourself five minutes:
- Open the buyer account brief (Layer 1 and 2 — pre-synthesized).
- Skim the last interaction summary (Layer 3 — auto-pulled).
- Write one line: what do I most want to learn or land?
- Note one curiosity: something specific to this buyer you'd ask if you were genuinely curious.
That's the prep. Five minutes. Repeatable. Cheap enough to actually happen.
The goal isn't to be impressive in the call. It's to be present and useful. Five minutes of focused prep does that. Forty-five minutes of unfocused research doesn't.