Prompt engineering 101 for sales & marketing teams

As sellers, we're constantly seeking ways to enhance our pitches and forge stronger connections with potential customers . One powerful tool at our disposal is generative AI (genAI). However, to harness its full potential, we need to master the art of prompt engineering first.

In this article, we’ll walk you through all the basic “101”-type information that sales and marketing teams need to know about prompt engineering. Specifically, we'll explore what prompt engineering is, why it's crucial for our success, and how to create prompts that yield optimal results. By the end, you'll have the knowledge and tools to leverage AI effectively, enabling you to produce impactful content that resonates with your prospects and moves deals forward.

What is prompt engineering?

Prompt engineering is the practice of writing a specific set of instructions that we give to large language models (LLMs); those instructions (called "prompts" provide the LLMs with the information and guidance they need to generate a specific type of content that meets specific requirements we have (e.g., voice, tone, etc.). It's like giving directions to a very smart – but literal – assistant. The better our prompts, the more useful and relevant the AI's output will be for our sales and marketing efforts.

By mastering prompt engineering, we can leverage AI to enhance our selling strategies, create compelling content, and improve our overall performance.

What are prompts?

As we highlighted earlier, an AI prompt is a bit of text input that we give to a LLM to elicit a specific response or action. They're the instructions or questions we use to tell the AI what we want it to do or what kind of information we're after. Think of them as the scripts we give to an incredibly smart assistant. 

For our purposes, prompts act as a bridge between our sales goals and the AI's capabilities, allowing us to harness its power for everything from crafting compelling email subject lines to analyzing customer data.

Prompts can take the form of simple questions or even complex instructions with multiple parts. For example:

• A simple prompt might be: "What are the key features of our product?"
• A more complex prompt could be: "Create a list of 5 benefits our CRM software provides to small businesses, focusing on productivity improvements and cost savings. For each benefit, provide a brief explanation and a hypothetical customer quote."

The quality of our prompts often determines the usefulness of the AI's output. Well-crafted prompts lead to spot-on responses, while vague ones might leave us with less-than-helpful results. That's why understanding different types of prompts is key to leveraging AI effectively in our sales and marketing processes.

{{cta-1}}

Types of AI prompts: User prompts vs. system prompts

If you work in sales or marketing, you’ll primarily work with two types of AI prompts: user prompts and system prompts. 

Let's break them down:

• User Prompts: User prompts are the most common type we'll work with in our day-to-day activities. User prompts help you define what kinds of information your outreach will include.
  • Imagine you're directing a one-person play, and the AI is your playwright. User prompts are like the specific instructions you give for each scene. They're the nitty-gritty details that guide the AI in creating exactly what you need for a particular task.
  • Key characteristics of user prompts:
    • Task-specific and frequently changed;
    • Focused on individual tasks or pieces of content;
    • Usually short to medium in length (about 1-5 sentences);
    • Easily modified for each new task;
  • Examples of user prompts in a sales or marketing context:
    • "Write follow-up emails to prospects who missed our scheduled demo calls."
    • "Generate 3 creative subject lines for an email campaign promoting our new feature launch."
    • "Summarize the key points from this call transcript and suggest 3 next steps."
• System prompts: System prompts, on the other hand, are instructions given to the AI model that set the context for the entire conversation. They’re typically used to define the AI's behavior, knowledge base, or persona. Think of them as defining how the outreach will be created, structured, and organized.
  • Imagine you're still directing that play, but this time the AI is your lead actor. System prompts are like the overall character brief you give the actor - who they are, their motivations, how they should carry themselves throughout the entire performance.
  • Key characteristics of system prompts:
    • Set the overall framework and guidelines for the AI;
    • Typically set once and rarely changed;
    • Broad in scope, applying to all interactions;
    • Usually medium to long in length (a paragraph or more); and,
    • Require careful consideration to change.
  • Examples of very basic system prompts in a sales or marketing context:
    • "You are a seasoned marketer who specializes in B2B software. You have extensive knowledge of CRM systems, sales processes, and industry best practices. Respond to queries in a professional yet friendly tone, and always prioritize ethical selling practices."
    • "You are a sales coach with 20 years of experience in the pharmaceutical industry. Provide advice and feedback to sales reps, focusing on relationship building and navigating complex buying processes in healthcare organizations."

Understanding these two types of prompts is crucial for effectively leveraging AI in our sales processes. User prompts help us tackle specific tasks with precision, while system prompts ensure that all AI-generated content aligns with our overall sales strategy and brand voice. By mastering both, we can create a powerful synergy between human expertise and AI efficiency, taking our sales game to the next level.

Practical examples of AI prompts in sales & marketing

Now that we've walked through the differences between user prompts and system prompts, let's look at some real-world examples of how we can use them in our day-to-day activities. These examples will show you how to leverage both types of prompts to enhance various aspects of your sales and/or marketing processes.

Example #1: Crafting a personalized cold email

Below, you’ll find sample user and system prompts for crafting a personalized cold email:

• User prompt:
  • “Generate a personalized follow-up email to the list of people who attended our webinar on 'AI-Driven Sustainability in Manufacturing' last Thursday. Structure the email as follows:
    • Choose between no salutation or 'Hi' followed by the recipient's {{first name}}.
    • **First Paragraph**:
      • Thank them for attending the {{webinar}} on the {{webinar date}}. Write only one sentence and skip a line before the next sentence.
    • **Second Paragraph**:
      • Briefly recap our {{product}} and highlight that it can result in reducing energy consumption by 30% for similar companies
    • **Third Paragraph**:
      • Describe the negative business outcome of not addressing the {{Pain Point}}.
    • **Fourth Paragraph**:
      • Suggest scheduling a 30-minute demo to show how our solution can be customized for their company’s needs
    • **Fifth Paragraph**:
      • Include a soft call-to-action to book the demo through your calendar link Keep the tone professional yet warm. Limit the email to 200 words. Use 'you' and 'your team' to make it personal.”

• System prompt:
  • "You are an AI assistant for TechInnovate, a B2B SaaS company specializing in AI-driven sustainability solutions for manufacturing. Maintain a professional, solution-oriented tone in all communications. Emphasize our core values of innovation, environmental responsibility, and customer success. Always prioritize building long-term relationships over pushing for immediate sales."

Example #2: Creating a thought leadership post for social media

Below, you’ll find sample user and system prompts for crafting a social media thought leadership content:

• User prompt:
  • "Create a series of 3 LinkedIn posts about the importance of emotional intelligence in sales. Each post should:
    - Be no more than 150 words
    - Start with a compelling hook or question
    - Provide one specific tip or insight about using emotional intelligence in sales
    - End with a call-to-action that encourages engagement (e.g., asking for comments, experiences)
    - Include relevant hashtags like #SalesTips #EmotionalIntelligence #SalesSuccess"

• System prompt:
  • "You are a social media content creator for SalesPro, a company offering sales training and coaching services. Your content should be engaging, informative, and position SalesPro as a thought leader in modern sales techniques. Use a conversational tone, incorporate current sales trends, and always aim to provide actionable insights our followers can immediately apply to their sales processes."

Example #3: Generating a call script

Below, you’ll find sample user and system prompts for crafting a call script:

• User prompt:
  • “I need help creating a cold call script for my SDR team. We're selling [describe your product/service] to [describe your target audience]. Our main value propositions are [list 2-3 key benefits]. Can you help me draft a script that:
    - Opens with a compelling introduction
    - Briefly explains our product/service and its main benefits
    - Includes 2-3 qualifying questions
    - Addresses at least one common objection
    - Ends with a clear next step or call-to-action
  • “Please keep the script concise, aiming for a call duration of about 2-3 minutes. Also, provide brief explanations for why you've structured certain parts of the script the way you have, so I can understand the strategy behind it.”

• System prompt:‍
  • “You are an expert sales coach specializing in cold calling techniques. Your role is to assist reps in crafting effective cold call scripts that are engaging, persuasive, and tailored to their specific products or services. You have extensive knowledge of sales psychology, conversation flow, and objection handling. Your goal is to help create scripts that are natural-sounding, respectful of the prospect's time, and designed to generate interest and secure next steps. When assisting with script creation, consider the following:
    - The importance of a strong opening that quickly captures attention
    - Clearly communicating the value proposition
    - Asking effective qualifying questions
    - Anticipating and addressing common objections
    - Using language that is confident but not pushy
    - Including clear call-to-actions or next steps
  • “Adapt your advice based on the specific industry, product/service, and target audience provided by the user. Be prepared to explain the rationale behind your suggestions and offer alternatives when appropriate.”

{{cta-1}}

How prompt engineering can help sales & marketing teams

Now that we've covered the "prompt 101" basics, let's zoom out and examine the broader impact of prompt engineering on our workflows. From the initial outreach to post-sale customer care, skillfully crafted prompts unlock AI's potential across the entire sales lifecycle. By harnessing this technology, we're not just working harder — we're working smarter. The following list illuminates key areas where prompt engineering elevates our game, transforming how we approach each stage of the buyer journey:

• Enhances prospecting: In prospecting, prompt engineering can help us instruct our AI to do things like:
  • Generate personalized emails;
  • Write compelling social media outreach;
  • Create compelling thought leadership content (e.g., blog posts, whitepapers);
  • Automate prospect research;
  • Refine ideal customer profiles (ICPs); and,
  • Elevate lead qualification by doing things like developing more accurate lead scoring models and generating dynamic qualification questionnaires; and,
  • Craft customized call scripts for each prospect, based on their specific situations and needs.
• Improves discovery and needs analysis: Prompt engineering can enhance our ability to uncover and analyze customer needs by allowing us to instruct our AI to do things like:
  • Generate tailored sets of discovery questions for different industries, roles, and scenarios; 
  • Conduct pain point analysis to uncover root causes and potential solutions;
  • Analyze real-time conversation transcripts to suggest follow-up questions based on prospect responses;
  • Create solution-mapping frameworks that align prospect needs with our product features, helping teams present the most relevant solutions.
• Optimizes follow-up and nurturing: When it comes to lead nurturing, we can use prompts to get our AI to help us maintain engagement with prospects and move them through the sales funnel by doing things like:
  • Designing dynamic, multi-touch follow-up campaigns that adapt based on prospect interactions and preferences;
  • Interpreting prospect interactions to gauge interest and suggest next steps;
  • Analyzing patterns to determine the best times and channels for follow-up communications for each prospect; and,
  • Creating targeted, personalized messages to re-activate dormant leads.
• Enhances post-sale activities: With high-quality prompts in place, customer success teams can instruct AI to support them with high-value account management tasks like:
  • Analyzing customer usage data and industry trends to identify expansion opportunities within existing accounts;
  • Monitoring customer health and analyzing customer interactions, product usage, and support tickets to flag potential churn risks;
  • Generating tailored messages for account managers to use in regular customer check-ins;
  • Crafting compelling case studies and testimonials based on customer data and feedback; and,
  • Developing models to predict renewal likelihood and suggesting proactive steps to secure renewals.
• Boosts sales enablement: When it comes to sales enablement, prompt engineering can help us instruct AI to do things like:
  • Gathering and analyzing information about competitors, including recent updates, pricing changes, and market positioning;
  • Generating and updating competitive battlecards with key differentiators and suggested rebuttals;
  • Creating training materials like role-playing scenarios, quiz questions, and other interactive content;
  • Producing or updating sales and marketing collateral like one-pagers, pitch decks, and other materials, ensuring they're always current and relevant; and,
  • Analyzing top-performing reps' activities and communications to surface best practices for the rest of the team.
• Supports revenue operations: Prompt engineering can play a huge role in revamping revops’ workflows, too. Specifically, you can use prompts to instruct AI to do things like:
  • Analyze your sales data, identifying trends, bottlenecks, and areas for improvement in the sales process;
  • Develop more accurate sales forecasting models by incorporating more variable options and historical data;
  • Generate insights to optimize sales territory assignments based on rep performance, market potential, and other factors;
  • Analyze deal win rates, competitor pricing, and market conditions to suggest optimal pricing strategies; and,
  • Align sales and marketing efforts by analyzing the buyer's journey to ensure seamless handoffs.

By leveraging prompt engineering across these various sales motions, we can significantly increase our efficiency and effectiveness. We're able to personalize our approach at scale, respond more quickly to prospect needs, and ultimately close more deals. 

Keep in mind, though, that these examples merely scratch the surface of AI's capabilities in revolutionizing our selling strategies. As we continue to refine our prompt engineering skills, we'll find even more innovative ways to apply this powerful tool throughout the sales process.

{{cta-1}}

Key elements of an effective prompt

Now that we've explored how prompt engineering can supercharge our efforts, let's dive into the nuts and bolts of crafting top-notch prompts. 

Just as we carefully choose our words when speaking with prospects, we need to be intentional when communicating with AI. By mastering these key elements, we'll be able to create prompts that consistently deliver high-quality, conversion-boosting content. Whether we're whipping up a cold email or drafting a cold-calling script, understanding these components will help us get the most out of our AI. 

Let's break down the essential ingredients that make a prompt truly effective:

1. Clear context: Set the stage for the AI by explaining the background and purpose of your request.
   • Example: "We're reaching out to potential clients in the healthcare industry who are looking to reduce operational costs."
2. Specific instructions: Tell the AI exactly what you want it to do. Be as detailed as possible.
   • Example: "Write a cold email that highlights how our software can cut administrative costs by 30%."
3. Desired format: Specify how you want the information presented. This could be a bullet list, a paragraph, or even a specific structure like a AIDA (Attention, Interest, Desire, Action) format.
   • Example: "Present the information in a 3-paragraph email with a clear call-to-action at the end."
4. Tone and style: Describe the voice you want the content to have. This helps ensure the AI's output matches your brand and the expectations of your audience.
   • Example: "Use a professional but friendly tone, similar to how you'd speak to a respected colleague."
5. Constraints: Set any limitations or requirements for the output. This could include word count, specific phrases to use or avoid, or particular points that must be included.
   • Example: "Keep the email under 200 words and avoid using technical jargon."
6. Target audience: Clearly define who the content is for. This helps the AI tailor its language and approach appropriately.
   • Example: "The email is for C-level executives in mid-sized manufacturing companies."
7. Goal or desired outcome: Explain what you're trying to achieve with this content. This gives the AI a clear direction and purpose.
   • Example: "The goal is to secure a 15-minute discovery call to discuss their cost-saving needs."
8. Examples or references: If possible, provide examples of what you're looking for or reference materials the AI can use as a guide.
   • Example: "Use a similar style to this example: [insert example text]"

By including these elements in our prompts, we're setting the AI up for success and increasing our chances of getting high-quality, relevant output that meets our needs as sellers.

{{cta-1}}

15 prompt engineering techniques

Crafting effective prompts is more art than science, but there are best practices that can make the process easier and more fruitful. While there's no one-size-fits-all approach to prompt engineering, keeping some best practices in mind can significantly improve our results. 

It's worth noting that the best practices differ slightly depending on whether we're creating user prompts or system prompts. Remember: user prompts are task-specific and frequently changed, while system prompts set the overall framework and are rarely modified. By mastering both types of prompts, we can guide our AI assistants to produce highly tailored, contextually appropriate content that enhances our sales processes and outcomes.

Let's explore some techniques that can help us generate high-quality content and messaging for our pitches, broken down into best practices for both user and system prompts.

Best practices for user prompts

Remember: User prompts help you define what kinds of information your outreach will include. Here are some tips to make your user prompts more effective:

• Define the value proposition: Ask the AI to give a little taste of why your product or service is special. What makes you different from the rest?
• Explain your goal: Tell the AI what you're trying to achieve with this content. This gives it a clear direction and purpose. 
• Provide section-by-section instructions: Break down your email into parts and tell the AI what you want in each section. This will helps ensure that the AI writes the content that covers all your bases and flows in a logical and natural way.
• Personalize the approach: Request that your AI include specific details about the prospects, such as their names, companies, industries, and any known challenges or goals based on your research.
• Avoid clickbait and spam triggers: For emails and social media outreach, specifically, tell the AI to steer clear of things that could get your message flagged or stuck in Spam folders; for example, avoid using all caps, too many exclamation points, etc. Keep it honest and valuable, not cheap and flashy.
• Specify call-to-action details: Be super clear about what you want the reader to do next and how the AI should present this in the email.
• Provide examples or references: If possible, share examples of what you're looking for or reference materials the AI can use as a guide.

Best practices for system prompts

Remember: System prompts set the overall context and guidelines for the AI's outputs. Think of them as defining how the outreach will be created, structured, and organized. Here are some best practices for creating effective system prompts:

• Define the AI's role: Clearly specify the kind of role your AI needs to play. For example, tell the AI it's acting as an expert sales rep specialized in crafting compelling cold outreach.
• Define the context clearly: Make it clear what the purpose of the outreach is and why you’re connecting with the prospects. 
• Set the overall tone: Establish the general tone and style for all responses. For instance, "Maintain a professional, confident, and value-focused tone, without being pushy or overly familiar."
• Establish content standards: Set some universal ground rules for all outreach of this type. For example, if you’re setting system prompts for cold intro emails, you can set guidelines for the ideal email length and/or required elements (e.g., personalization, call-to-action, unique value proposition),
• Specify the response structure: Give the AI a basic framework for how responses should be structured, including standard sections or formatting.
• Incorporate brand voice: Provide guidelines on how to make the communications sound like they're coming from your company.
• Set language preferences: Tell the AI which words or phrases to use (or avoid), especially industry lingo. Focus on outcomes and value, not just listing features. 
• Define personalization depth: Guide the AI on how much to personalize based on what you know about the lead.

{{cta-1}}

7 common mistakes to avoid when creating prompts

As we dive into the world of prompt engineering, it's important to be aware of some common mistakes sales and marketing teams make. By avoiding these errors, you can ensure that your prompts are more effective and yield better results.

Mistake #1: Writing prompts that are too vague 

• Problem: Vague prompts like "Write something about our product" don't give the AI enough direction.
• Solution: Be specific about what you want. Include details about the product, target audience, and desired outcome.
• Example of a better prompt: "Write a 200-word product description for our cloud-based project management software, highlighting its collaboration features. The target audience is small business owners looking to improve team productivity."

Mistake #2: Overloading the prompt

• Problem: Trying to accomplish too much in one go can lead to confused or unfocused output.
• Solution: Break complex tasks into smaller, more manageable prompts.
• Example: Instead of asking for a full sales pitch in one prompt, break it down into sections like the opening, key benefits, addressing objections, and call-to-action.

Mistake #3: Forgetting to specify tone

• Problem: Without guidance on tone, the AI might default to a style that doesn't match your brand voice.
• Solution: When relevant, make sure to include instructions about the desired tone in your prompts.
• Example: "Use a professional yet friendly tone, similar to how you'd speak to a respected colleague over coffee."

Mistake #4: Neglecting to provide context

• Problem: Lack of context can result in generic or irrelevant content.
• Solution: When possible, provide background information about your company, the prospect, or the specific situation (without revealing any proprietary or sensitive information).
• Example: "We're a B2B software company reaching out to a prospect who recently downloaded our white paper on AI in healthcare. They're a mid-sized hospital looking to improve patient care through technology."

Mistake #5: Not considering the recipient’s level of familiarity or expertise

• Problem: Using too much industry jargon or technical terms that your audience might not understand – or not enough to the point of sounding inauthentic or inexperienced.
• Solution: Tailor your prompts to ask the AI to generate content appropriate for your audience's level of familiarity with your industry.
• Example: "Explain our machine learning algorithm in simple terms that a non-technical CEO would understand."

Mistake #6: Overlooking the importance of formatting

• Problem: Failing to specify how you want the information presented can result in hard-to-read blocks of text.
• Solution: Include instructions for formatting in your prompts.
• Example: "Present the information in a bulleted list with no more than 5 main points, each explained in 1-2 sentences."

Mistake #7: Neglecting to set constraints

• Problem: Without constraints, the AI might generate overly long or unfocused content.
• Solution: Set clear limitations in your prompts.
• Example: "Keep the email under 150 words and focus on no more than three key benefits of our product."

By being aware of these common mistakes and actively working to avoid them, we can significantly improve the quality of our prompts and the resulting AI-generated content. 

Final thoughts

Mastering prompt engineering can significantly enhance our effectiveness as sellers. By creating clear, specific, and context-rich prompts, we can leverage AI to generate high-quality content that resonates with our prospects and moves deals forward. Remember, the key is to be clear about what we want, provide relevant context, and iterate on our prompts to get the best results.

As you continue to refine your prompt engineering skills, you’ll find that you can work more efficiently, create more personalized outreach, and ultimately close more deals. So embrace this powerful tool and use it to take your selling game to the next level.