Where AI helps and where it gets in the way
Sales reps fall into two extremes: either rejecting AI entirely as generic marketing fluff, or blindly asking LLMs to draft entire cold emails and proposals from scratch. As a result, decision-maker inboxes get flooded with identical emails full of buzzwords like 'revolutionary' or 'synergy', causing reply rates to plummet. Delegating trust-based tasks to AI destroys customer rapport before a conversation even starts.
Implement an operational task matrix based on two dimensions: Cost of Error (Hallucination Risk) vs Degree of Data Structure.
- Green Zone (AI First): High structure, low error cost. Analyzing annual reports, extracting tech stacks from website code, transcribing call notes, populating CRM fields, and generating headline variants for A/B testing.
- Yellow Zone (AI Assisted): High-impact tasks requiring human-in-the-loop review. Drafting first proposal outlines, identifying pain points from prospect quotes, framing discovery questions. AI generates 3 options; the rep selects and refines one.
- Red Zone (Human Only): Live pricing negotiations, building interpersonal rapport, reading vocal hesitation and micro-expressions, making trade-off concessions. AI has no 'skin in the game' and cannot bear financial accountability.
- BAD
- The rep prompts: 'Write an email to a CFO offering a 15% discount so they sign today.' The model creates a desperate message that destroys negotiating leverage and surrenders margin without concessions.
- GOOD
- The rep prompts: 'A logistics CFO raised this objection: Competitor Y is 20% cheaper. Generate 3 probing questions exposing hidden implementation and SLA costs in the competitor offer, so I defend our margin tomorrow without lowering price.' The rep receives sharp commercial ammunition.
Act as: Experienced B2B Sales Enablement Strategist. Analyze the following sales task list from a typical workday: [PASTE TASK LIST AND DURATION] Tasks: 1. Assign each task to one of three zones: - GREEN ZONE: AI handles 80% (Research, Extraction, Data formatting) - YELLOW ZONE: AI assists, human reviews (Drafting, Qualification) - RED ZONE: Human only (Negotiation, Live rapport, Price commitments) 2. For each Green and Yellow task, specify: - Recommended model (Claude 3.5 Sonnet / GPT-4o / Perplexity Pro) - Estimated time saved in minutes - Ready execution prompt with bracketed placeholders [ ] Rules: - No motivational fluff or generic advice. - Every prompt must include negative constraints (what the model MUST NOT do).
Variations & alternative scenarios
SDR Variant: Focus exclusively on account research, buying triggers, and multi-channel prospecting sequences.
AE Variant: Focus on demo transcript synthesis, commercial proposal drafting, and margin defense.
Practical exercise
List 8 repetitive tasks from your past 3 working days. Categorize them into Green, Yellow, and Red zones. Pick 2 Green Zone tasks and build reusable execution prompts for them.