Build a Comprehensive Account Brief with Deep Research
Deep-research account brief
Promptifi · Curated prompt map
A focused map of 50 reusable prompts for the sales jobs that matter most—from account research and outreach to deal strategy, renewal and forecast review.
Every card opens one canonical Promptifi record. This page explains what to choose; the full library remains the exhaustive browser, and each prompt keeps its own Free or Pro access.
Canonical links verified against the live Promptifi Prompts collection on 2 September 2026.
An AI sales prompt is a structured instruction that turns the context you already have—an account, transcript, opportunity, buyer, pipeline file or goal—into a specific sales output you can review and use.
The useful part is not the clever wording. It is the job definition, evidence, constraints and expected output.
Strong prompts reduce ambiguity before the model starts. They make it clearer what the AI should use, what it should not assume, and what a usable answer looks like.
Salespeople rarely think “I need prompt #742.” They think “I need to prepare for tomorrow's CISO meeting,” “this deal is stuck,” or “I need more pipeline.” Start there.
Turn account, market and buyer evidence into a defensible plan for where to spend selling time.
Use verified context to earn attention across email, phone, voicemail and social outreach.
Enter meetings with relevant evidence, sharper questions and an explicit outcome to secure.
Convert transcripts and practice sessions into evidence, coaching and reusable learning.
Expose missing evidence, strengthen champions and turn buyer commitments into forward motion.
Handle commercial, procurement and technical pressure without inventing proof or overcommitting.
Carry value evidence into handoffs, QBRs, renewals and account growth.
Inspect pipeline and performance patterns, then translate them into accountable action.
8 prompts
Turn account, market and buyer evidence into a defensible plan for where to spend selling time.
Deep-research account brief
Monitor account trigger events
Identify buying signals
Synthesize an executive's public statements
Score and prioritize target accounts
Build an ICP from win data
Map a buying committee
Build a current competitive battlecard
6 prompts
Use verified context to earn attention across email, phone, voicemail and social outreach.
Research an account and draft a cold email
Write a trigger-plus-empathy cold email
Prepare a contextual cold-call opener
Leave a signal-led voicemail
Extract outreach hooks from an executive interview
Use a prospect's public quote in an opener
7 prompts
Enter meetings with relevant evidence, sharper questions and an explicit outcome to secure.
Build an evidence-based meeting brief
Prepare for a C-suite meeting
Create MEDDIC-aligned discovery questions
Plan technical discovery
Plan executive-sponsor discovery
Anticipate objections before discovery
Draft an action-oriented meeting follow-up
7 prompts
Convert transcripts and practice sessions into evidence, coaching and reusable learning.
Score and coach a sales call
Map discovery evidence to MEDDPICC
Convert a call transcript into CRM notes
Find ICP and messaging patterns across calls
Build an objection library from transcripts
Run scored discovery role-play
Diagnose rep skill gaps and build a coaching plan
7 prompts
Expose missing evidence, strengthen champions and turn buyer commitments into forward motion.
Identify and test a champion
Create a mutual action plan
Build a deal risk register
Assess and revive a dormant deal
Draft a proposal from call evidence
Build a champion-ready business case
Map approved proof points to persona and stage
6 prompts
Handle commercial, procurement and technical pressure without inventing proof or overcommitting.
Plan concessions with give-get logic
Reframe a price objection around cost of inaction
Analyze a procurement thread and unblock delays
Answer technical due diligence without overcommitting
Manage an RFP response from approved knowledge
Extract pilot proof for the close deck
4 prompts
Carry value evidence into handoffs, QBRs, renewals and account growth.
Build a renewal pitch from value evidence
Prepare a customer QBR
Identify expansion opportunities
Create an AE-to-CSM handoff
5 prompts
Inspect pipeline and performance patterns, then translate them into accountable action.
Build a pipeline-health report
Analyze forecast risk and slippage
Find pipeline conversion bottlenecks
Find recurring win/loss patterns
Build a territory action plan
A prompt should produce one useful result. Use a connected resource when the work needs several steps, persistent context or automation.
Every job-specific prompt collection in the library, grouped by where you would reach for it.
Good prompt design matters, but the evidence you provide matters more. These checks keep the output grounded enough to use in real selling work.
Name the decision, deliverable or next action—not just the topic.
Use account facts, call notes, pipeline data or approved source material whenever possible.
Tell the model what the work should help you decide, prepare or produce.
A brief, table, talk track, questions or risk register is easier to review than “ideas.”
Require unknowns and missing evidence to be called out instead of silently filled in.
Validate anything customer-facing, technical, financial, legal or commercially material.
A prompt is the right tool for one bounded output. As the work becomes multi-step, recurring or tool-driven, move up the execution ladder instead of forcing everything into one giant prompt.
A prompt your assistant runs on a cadence for recurring work.
See scheduled promptsThe prompt itself starts with the job. If your role or model choice is the stronger signal, Promptifi already has deeper routes for those contexts.
Role pages narrow the library to the work your seat actually owns.
This page stays tool-agnostic. Model-specific pages explain when ChatGPT or Claude changes the setup, capabilities or prompt structure.
Quality shows up in concrete ways, not in words like “expert-crafted”. What matters is whether the prompt has a defined sales job, usable inputs, a clear output, guardrails and a canonical record that can be maintained.
The prompt is attached to a specific task or decision, not a generic topic.
The seller can tell what context or evidence the prompt expects.
The result has a useful shape: brief, table, questions, scorecard or action plan.
The prompt should distinguish supported facts from missing information and assumptions.
Prompt records can connect to stage, seat, task and other Promptifi taxonomy.
Each entry links to the maintained Promptifi prompt record rather than duplicating the full prompt body.
They are structured instructions that use account, buyer, call, pipeline or deal context to produce a defined sales output—such as a research brief, outreach draft, discovery plan, risk review or renewal narrative.
A strong prompt defines the job, provides relevant context and evidence, sets constraints, specifies the output and tells the model how to handle missing or uncertain information.
Not exactly. This page is platform-agnostic. Promptifi's ChatGPT-specific page goes deeper on ChatGPT capabilities and prompting patterns, while the canonical prompts here can be evaluated for the AI tool you use.
Use the Browse Library experience when you know what you need and want exhaustive search, filtering and task-level discovery across the full 2,900+ prompt library.
Give it the minimum evidence needed for the job: account facts, approved source material, call notes, opportunity context, pipeline data or customer evidence. More text is not automatically better; relevant evidence is.
Use a prompt chain when the work has several dependent steps and the output of one prompt needs to become the input to the next.
Use a workflow when the task is a repeatable multi-step process. Use automation when a tool can reliably execute that process, and scheduled prompts when the useful behavior is primarily recurring on a cadence.
Start with the curated 50 when you need direction. Move to the full library when you know exactly what you want.