Prompt Engineering for Sales: 20 CRM Template Guide
Twenty battle-tested prompt templates for sales teams using AI with CRM platforms. Covers lead scoring, follow-up drafting, objection handling, and reports.
Ready-to-Use Templates
Higher Reply Rates with AI Drafts
Sales Workflow Categories
Weekly Time Saved Per Rep
Key Takeaways
Sales teams have been using AI tools for years, but most reps still struggle to get consistent, usable outputs from the AI integrated into their CRM. The gap is almost never the model. It is the prompt. Vague instructions produce vague results; structured templates with the right context produce outputs that reps can actually send or present with minimal editing.
This guide provides 20 battle-tested prompt templates organized across the core workflows where sales teams spend the most time: lead scoring, follow-up drafting, objection handling, deal summaries, pipeline reporting, discovery preparation, and proposal writing. Each template includes the prompt structure, placeholder fields to customize, and notes on what makes it work. For context on how AI fits into broader CRM and automation strategy, these prompts are most effective when embedded directly into the workflows your team already uses, not treated as a separate tool to open in a browser tab.
Why Prompt Engineering Matters for Sales
The term prompt engineering sounds technical, but for sales teams it means one practical thing: writing instructions that give the AI enough context to produce something useful on the first try. Most reps type a one-line instruction and are disappointed by a generic response. Most managers assume the AI tool is not good enough. The actual problem is that the prompt is missing the three elements every sales AI interaction requires: role context, deal context, and output format.
Tell the model it is acting as a senior sales rep for your specific product category. This sets the vocabulary, formality level, and domain assumptions for every output.
Include company name, deal stage, budget signal, main pain point, and last touchpoint. Without these, the model invents plausible-sounding but wrong details.
Specify whether you want a 150-word email, a five-bullet summary, a numbered objection list, or a scored assessment. Unspecified format produces unpredictably long outputs.
Research from sales enablement platforms consistently shows that reps who use structured AI prompts produce follow-up emails with 40 percent higher reply rates than reps using unstructured prompts or no AI at all. The difference is not intelligence; it is context. The templates below are organized by workflow stage and pre-include the structural elements that produce reliable outputs. Pair them with a deeper understanding of Salesforce AI CRM workflows for SMBs to see how prompt-driven automation fits into a complete pipeline strategy.
How to use these templates: Replace every bracket placeholder with real deal data before running the prompt. The more specific the context you provide, the less editing the output will require. Templates marked with a star are highest-impact for most sales teams.
Lead Scoring and Qualification Prompts
Lead scoring prompts are the highest-leverage application of AI in the sales CRM because they operate at the top of the funnel where volume is highest and rep time is most scarce. A well-designed scoring prompt can triage 50 inbound leads in the time it would take a rep to manually evaluate five.
Prompt template
You are a sales qualification specialist for [Company Name], which sells [Product/Service] to [Target Segment]. Our ideal customer profile requires: [ICP Criteria 1], [ICP Criteria 2], [ICP Criteria 3]. Score this lead on a 1–10 ICP fit scale and list the top two fit signals and top two disqualifying factors. Lead data: [Paste CRM fields here].
Best used when reviewing inbound form submissions or enriched lead records. Fill in your actual ICP criteria — employee count thresholds, tech stack requirements, or revenue ranges — not generic descriptions.
Prompt template
Review this discovery call transcript or meeting notes and assess BANT qualification. For each criterion — Budget, Authority, Need, Timeline — state whether it is confirmed, inferred, unknown, or disqualified, and cite the specific evidence. Notes: [Paste call notes here].
Particularly useful after discovery calls to turn messy notes into a structured qualification assessment that can be logged directly in the CRM deal record.
Prompt template
Analyze the following email thread or LinkedIn conversation for buying signals. List explicit signals (direct statements of intent or urgency), implicit signals (tone, question types, information requests), and red flags. Recommend the next best action. Thread: [Paste conversation here].
Works well for SDRs managing high-volume outbound sequences who need to quickly identify which leads are warming up and deserve priority follow-up.
Follow-Up Email Drafting Prompts
Follow-up emails are where most deals go dark. Reps delay writing them because they do not know what to say, or they send generic check-ins that get ignored. AI-drafted follow-ups that reference the specific conversation dramatically outperform both. These templates require more deal context than any other category, which is why the placeholder fields are longer.
Prompt template
Write a follow-up email to [Prospect Name] at [Company] after a product demo. Key points discussed: [Specific features shown], [Pain points they mentioned], [Questions they asked]. Their biggest concern was [Concern]. Next step agreed: [Next Step]. Write in a [formal/conversational] tone in under 150 words. Do not use "I hope this finds you well."
The explicit instruction to avoid "I hope this finds you well" is not trivial — this phrase appears in a large fraction of AI-drafted emails and immediately signals an automated origin.
Prompt template
Write a re-engagement email for a deal that went dark [X weeks] ago. Last interaction summary: [What was discussed]. Reason they went quiet (if known): [Reason or "unknown"]. New angle to try: [Industry news, product update, relevant trigger event]. Target: one short paragraph with a specific question as the call to action. No "just checking in."
The "new angle" field is critical. Reps who skip it get generic re-engagement emails. Reps who fill it in with a relevant trigger — a funding announcement, regulatory change, or product update — get replies.
Prompt template
Write a follow-up email for [X] days after sending a proposal to [Prospect Name] at [Company]. Proposal value: [Amount]. Main value proposition: [Core benefit]. Add one concrete ROI data point relevant to [their industry]. Include a specific question to uncover any blockers. Maximum 120 words.
Instructing the model to add an ROI data point forces it to produce a substantive email rather than a generic nudge. The data point grounds the follow-up in value rather than urgency.
Style matching tip: Add a sentence to any follow-up prompt: "Match the writing style of this sample email from the rep: [paste 2–3 sentences]." This single addition makes AI drafts read as human-written far more consistently.
Objection Handling Prompts
Objection handling is a category where AI performs especially well because objections are largely predictable and responses can be systematically optimized. The challenge is that reps often paste an objection and ask "how do I respond?" without giving the model any of the context needed to produce a response that fits their product, deal stage, or company positioning.
Prompt template
My prospect said: "[Exact price objection quote]." Our product costs [Price] and the main ROI drivers are [ROI Driver 1] and [ROI Driver 2]. Their company size is [Size] and their stated pain is [Pain]. Write three different objection responses: one focusing on ROI, one on risk reduction, and one that asks a clarifying question to understand whether price or budget is the real blocker.
Generating three variants lets the rep choose the approach that best matches the prospect's personality and the conversation tone, rather than committing to a single AI-generated script.
Prompt template
The prospect said they are also evaluating [Competitor Name]. Our differentiators versus [Competitor] are: [Diff 1], [Diff 2], [Diff 3]. The prospect's primary use case is [Use Case]. Write a response that acknowledges the competitor fairly, pivots to our specific strengths for their use case, and ends with a question that advances the conversation. Avoid disparaging the competitor.
The instruction to avoid disparaging competitors is important. Without it, models sometimes generate responses that undermine credibility by sounding defensive or petty.
Prompt template
The prospect said: "This isn't the right time." Deal stage: [Stage]. They previously said their biggest pain was [Pain]. Write a response that: (1) validates their concern, (2) gently links delay to the cost of their pain continuing, (3) offers a low-commitment next step. Keep it under 80 words for a verbal response or 100 words for email.
Specifying separate word limits for verbal versus email formats gives the rep two usable outputs from a single prompt run, covering both the follow-up call and email scenarios.
Deal Summary and Pipeline Reporting Prompts
Pipeline reporting is the most underutilized AI use case in sales teams. Most reps associate AI with customer-facing communications and overlook the internal documentation and reporting workflows that consume significant time every week. These prompts reclaim that time and, more importantly, produce more consistent and accurate pipeline data for managers and forecasting models.
Prompt template
Summarize the following CRM deal data into a weekly pipeline report for my manager. For each deal, state: current stage, probability change from last week, next action and due date, and main risk. Format as a table. Flag any deals that have not progressed in [X] days. Data: [Paste CRM export or deal field list here].
The stagnation flag — deals not progressed in X days — is the most valuable element. This surfaces hidden risk that manual pipeline reviews frequently miss when managers focus on high-value deals.
Prompt template
Assess the health of this deal and rate it red, amber, or green. Deal data: [Stage], [Value], [Close date], [Days in current stage], [Last activity date], [Number of stakeholders engaged], [Champion identified: yes/no]. List the top three risks and the one action most likely to improve deal health. Be direct, not optimistic.
"Be direct, not optimistic" is a key instruction. Without it, models tend to produce encouraging assessments that systematically overstate deal health — the opposite of what accurate forecasting requires.
Prompt template
Write a monthly sales forecast narrative for [Month] based on the following pipeline data. Include: total committed pipeline, best-case upside, three biggest deals at risk with reasons, and recommended manager actions. Write in first-person as the sales manager. Tone: analytical and concise. Data: [Paste pipeline summary here].
Writing in first-person as the sales manager produces a ready-to-send executive summary rather than a third-party analysis. Sales managers who use this template typically save 45 to 90 minutes per monthly report cycle.
For teams running Salesforce, these prompts map directly to the types of fields and reports available in the platform. See our guide on Salesforce AI CRM workflows for SMBs for the specific integration patterns that make these prompts callable directly from within Salesforce flows, reducing the copy-paste step entirely.
Discovery Call Preparation Prompts
Preparation prompts are run before a call rather than after. They use publicly available information about the prospect — their website, LinkedIn profile, recent news, job posting data — to produce a customized call prep brief. Reps who prepare with AI briefs ask better questions and reference more relevant details during calls, which consistently shortens the average time to close.
Prompt template
I have a discovery call tomorrow with [Prospect Name], [Title] at [Company]. The company is in [Industry], has approximately [Size] employees, and recently [Recent News or Trigger Event]. We sell [Product/Service]. Generate a call prep brief with: five targeted discovery questions, two hypotheses about their likely pain points, and three things to avoid mentioning or assuming.
The "three things to avoid" section often surfaces the most valuable prep insight. For example, if the company recently had layoffs, the model may flag not to discuss headcount expansion use cases.
Prompt template
Based on their LinkedIn profiles and roles, describe likely priorities, concerns, and communication preferences for each stakeholder in my deal. Stakeholders: [Name, Title, Department — repeat for each]. Our product impacts [Business Area]. For each person, give one sentence on what they probably care most about and one sentence on the objection they are most likely to raise.
Most effective when multi-threading into enterprise deals where the rep needs to prepare for a room with three to five different stakeholders, each with different incentives and concerns.
Prompt template
[Company] recently [Trigger Event: e.g., raised a Series B, expanded to a new market, hired a new VP of Sales, announced an acquisition]. Our product helps companies in this situation by [Specific Value Prop]. Write three opening talking points that reference this trigger event and connect it naturally to a pain our product solves. Sound congratulatory not opportunistic.
The tone instruction — "congratulatory not opportunistic" — makes a significant difference. Without it, the model sometimes produces talking points that read as self-serving rather than prospect-focused.
Proposal and Closing Prompts
Late-stage deal prompts focus on proposal customization and closing plan development. These prompts are the highest-stakes in the library because errors or generic content in a proposal directly impacts close rates. They require the most thorough context injection but also produce the highest return when done correctly.
Prompt template
Write a one-page executive summary for a sales proposal to [Company]. Their stated problem: [Problem]. Their success metric: [Metric]. Our recommended solution: [Solution]. Key implementation milestones: [Milestone 1], [Milestone 2]. Investment: [Price]. Expected ROI in [Timeframe]: [ROI]. Write for a C-suite audience. Use specific numbers, not vague claims. Three paragraphs maximum.
"Use specific numbers, not vague claims" is the most important instruction in this prompt. It forces the model to use the ROI and metric data you provided rather than generating plausible-sounding but unsubstantiated claims.
Prompt template
Create a mutual action plan (MAP) for closing a deal with [Company] by [Target Close Date]. Key steps needed on their side: [Their Steps]. Key steps needed on our side: [Our Steps]. Key stakeholders: [Names and Roles]. Format as a table with columns: Action, Owner, Due Date, Status. Include a shared success definition at the top.
A mutual action plan shared with the prospect dramatically improves deal velocity by making the path to close visible and collaborative. AI-generated MAPs structured as tables are easy to paste into Google Docs or share via email.
Prompt template
Write a closing conversation script for a call with [Prospect Name]. Deal context: [Stage, Value, Key Concern]. Their decision-maker is [DM Name/Title]. We are at the point of asking for the business. Script should include: a brief value recap (30 seconds), a transition to the ask, the ask itself, and three responses to the most likely pushbacks. Natural language, not robotic.
Including three pushback responses in the same prompt means the rep can prepare for a branching conversation without needing additional prompts during or between calls.
Prompt template
Analyze this closed deal and produce a win/loss summary for the CRM. Outcome: [Won/Lost]. Final close reason stated by prospect: [Reason]. Key events in the deal timeline: [List 3–5 pivotal moments]. Competing solutions evaluated: [List]. Write: one paragraph on what worked, one on what failed or could improve, and three lessons for future similar deals.
Win/loss analysis is consistently underdone by sales teams because it feels low-priority post-close. This prompt makes it fast enough that reps actually complete it, building a historical pattern library for coaching.
Prompt template
Review this existing customer's usage data and CRM notes to identify expansion opportunities. Customer: [Company]. Current product usage: [Metrics]. CRM notes from last 90 days: [Paste relevant notes]. Identify: the most natural upsell or cross-sell opportunity with a rationale, the best-fit contact to approach, and a suggested opening message angle. Prioritize organic growth over hard selling.
Expansion revenue from existing customers costs three to five times less to acquire than new-logo revenue. This prompt systematizes a workflow that most account managers skip because it requires connecting disparate data sources.
Implementing Prompts in Your CRM
Having 20 templates in a document is a starting point, not a solution. The prompts that teams actually use are the ones embedded in the tools reps open every day. Depending on your CRM, there are three implementation approaches in increasing order of adoption effectiveness.
Store templates in Notion, Confluence, or Google Docs with a clear category structure. Reps copy, fill in context, and run in their preferred AI tool. Zero-friction setup but depends on rep discipline to use consistently.
Configure custom prompts inside Salesforce Einstein, HubSpot AI, or Pipedrive AI. Templates surface in the CRM interface with deal fields auto-populated. Higher adoption because reps never leave the tool.
Trigger prompts automatically when deal stages change — a follow-up draft generated when a deal moves to proposal, a health assessment flagged when a deal stalls. This requires CRM workflow automation but removes all manual friction.
Level 3 implementation requires connecting your CRM to an AI API, which is where CRM and automation services become relevant. For teams comparing platform capabilities before investing in automation, the HubSpot vs Salesforce 2026 comparison covers the AI feature differences in detail, which directly affects which Level 2 and Level 3 integrations are possible without custom development.
Adoption recommendation: Start with Level 1 for three weeks to identify which prompt categories your team uses most. Then invest in Level 2 integration only for those categories. Teams that try to implement Level 3 without usage data frequently automate the wrong workflows.
Related Articles
Continue exploring with these related guides