Nonprofit Techies

7 MIN READ

Building An AI Tool - Marketer's Guide by Nia Joseph

Former Employee

Jun 12, 2025

Prompt Engineering Worksheet 

This document serves as a comprehensive guide and template for Marketing professionals, detailing a systematic approach to prompt engineering. It includes instructional notes and methodologies specifically designed for the field of Marketing.

What is Prompt Engineering?

Prompt engineering is the practice of crafting and refining input queries to guide machine learning models, particularly language models like Copilot and ChatGPT, to produce optimal and relevant responses.

The 3 Main Types of Prompts

Generic (No Context)

Minimal (Some Context)

Data-Fed (Most Context)

Broad queries with minimal guidance.

Example: What is email marketing?

Usage for a Marketer: When looking for a general overview or introduction to a topic.

Benefits:

Offers a wide scope of information.

Good for brainstorming or exploring a new concept.

More pointed prompts that give some direction.

Example: What strategies are effective for email marketing?

Usage for a Marketer: When they know the domain they're interested in but want various insights or options.

Benefits:

Provides targeted information within a chosen domain.

Allows for deeper exploration of a particular subject.

Highly specific queries that incorporate particular data or conditions.

Example: Considering an open rate of only 10% for my last email campaign selling nostalgic bucket hats, targeting millennials, how can I improve this? I’ll include the email subject line and contents here.

Usage for a Marketer: When they have specific data or conditions, they want advice or information on.

Benefits:

Offers tailored recommendations or insights based on given data.

Best for optimization and refinement of existing campaigns or strategies.

Choosing Your Prompt

Generic: For broad overviews or exploring topics.

Minimal: When you have a domain in mind and want diverse insights.

Detailed: For more precise advice based on specific data or conditions.

Examples of Each Prompt

Generic Prompt - Equivalent to asking: Recommend a book

Minimal Prompt - Equivalent to asking: Recommend a science fiction book.

Detailed Prompt - Equivalent to asking: Recommend a science fiction book set in space, with strong female protagonists, written in the last decade.

A Simple Framework for Prompt Engineering

ARC(E) - Act, Request, Clarify. Engage.

Benefits of ARC(E)

Structures interactions to produce precise outputs.

Reduces ambiguity in prompts. Can be used across teams.

Adaptable for various domains and requirements.

A	Definition: Define the role or perspective you want the model to adopt. Example: Act as a LinkedIn Ads specialist. Usage: To set the tone and perspective for the response, ensuring it's tailored to a specific viewpoint or expertise.
R	Definition: Clearly state the task or information you are seeking from the model. Example: Provide audience insights and actionable performance tips. Usage: Gives the AI clear instructions on the type of output desired; advice, information, or a specific task.
C	Definition: Offer specific data, background, or situational details to inform the model's response. Example: Campaign targeting Data Scientists to sell a new AI Data management tool. Currently has a CTR of 2% and no conversions after being live for 30 days. Usage: Provides the model with essential background information, ensuring the response is relevant, data-informed, and accurate.
E	Definition: Invite the model to ask for any additional information it might need to optimize its response. Example: If you require more details to assist me better, please inquire. Usage: Promotes a dynamic interaction, allowing the model to clarify ambiguities or gather more context.

Examples of How to Use Engagement Prompts

Progressive Interaction

Start broad and refine based on AI feedback, ensuring the AI knows it can ask for details.

User: Help me design a marketing strategy. Ask for details if needed.

AI: For which product or industry?

Collaborative Inquiry

Proactively invite AI to ask clarifying questions to guide the conversation.

User: I want to boost user engagement. What details do you need from me?

AI: Can you share your platform's current engagement metrics and any patterns you've noticed?

Challenge Mode

Encourage AI to counter or critique ideas for a holistic perspective.

User: I'm leaning towards influencer marketing for our launch. Can you challenge that approach?

AI: Certainly. Have you considered potential risks, such as brand misalignment or the fleeting nature of influencer popularity?

Use Cases for Marketers

Leveraging AI can prove beneficial in many areas of your career as Marketing professionals. The focus should be on user-centric AI, where you are strategically guiding the AI and editing the results. In this way it becomes human-powered AI, helping to enhance your expertise.

Content Creation:

- Crafting social media posts.

- Generating blog topics or outlines.

- Producing creative ad copy.

Market Research:

- Analyzing customer feedback and sentiments.

- Identifying emerging market trends.

- Generating questions for customer surveys.

Strategy Development:

- SWOT analysis generation.

- Competitive landscape assessment.

- Crafting marketing plan outlines.

Brand Management:

- Creating brand taglines or slogans.

- Generating brand stories or narratives.

- Ideating brand campaign concepts.

SEO and SEM:

- Keyword research and suggestions.

- Meta description and title generation.

- AdWords campaign optimization suggestions.

Email Marketing:

- Crafting engaging email subject lines.

- Personalizing email content for different segments.

- Automating responses to common customer queries.

Customer Interaction:

- Responding to customer feedback or reviews.

- Generating templates for common customer communications.

- Crafting scripts for customer support teams.

Product Positioning:

- Generating product descriptions or USPs.

- Ideating names for new products.

- Crafting product launch announcements.

https://techfren.xyz/6-ai-marketing-tools-in-2025/

Image Prompt:
I want an image of An astronaut with his pet jellyfish in an underwater world. Style: claymation that is layered. Colors: shades of Deep Ocean Blue: #00578E Luminescent Purple: #800080 Neon Green: #39FF14 Glimmering Pink: #FF69B4. Background: underwater utopia that is futuristic.

Content Prompt:

Act as [Role]. [Request], I will provide you the details: [Context]

Building AI Tools from a Marketing Perspective

Here’s a doc I created to explain my process of building an AI tool while working in a Marketing org and what I think might work for someone with a marketing or traditionally non-technical background looking to build or develop an internal tool or process. While this example focuses on AI it can be adapted for any technology stack.

Before You Build:

Stakeholder alignment: Ensure that any relevant stakeholders or managers are aligned for this tool build. Be sure to keep them updated on progress and small wins.
Unique need: the problem you’re trying to solve and is not duplicative,
Business impact: this tool will reduce/improve/streamline xyz by x amount of time.
Collaboration request: getting your engineering team onboard; creating a v-team and pitching your product for volunteers.
1. Can be seen as a stretch project for engineering and shows collaboration.
2. Post your v-team in different channels and seek out spaces with engineering to understand more of their process and needs.
Understand AI Types:
1. Generative AI (Gen AI): Creates new, unique content. Useful for content creation, design, etc.
2. Large Language Models (LLM AI): Understands and generates human-like text based on vast data. Useful for chatbots, customer support, etc.
3. Others (e.g., Image Recognition AI, Predictive AI) Decide on Features & Inputs:
What information will the AI need? (Tone, product details, user data)
1. What outputs are expected?
  1. Example: For a content creation tool: tone (formal/informal), target audience, content type (blog, social media post).
Design the User Experience:
1. How will users interact with the tool?
2. How will errors be handled?
  1. Example: For a chatbot: Welcome message, typing indicators, error messages.
Continuous Improvement:
1. AI tools benefit from ongoing training and refinement.
2. Stay updated with AI advancements and user needs.

Product Requirements Worksheet for AI-powered Marketing Tools

1. Project Overview:

Project Name: [___________]

Objective: (Briefly describe the goal of this tool) [___________]

Target Audience: (Who is this tool intended for?) [___________]

2. Problem Statement:

Define the specific problem you aim to solve with this tool.

[____________________________________________________________________________________________]

3. Desired Outcomes:

List the expected results from using this tool. E.g., Increased conversion rates, improved customer engagement, etc.

[___________]

(Add more as needed)

4. Core Features & Functionality:

Describe the main functionalities the tool should have.

Feature: [___________]

Description: [________________________________________________________________]

Importance: (Critical, High, Medium, Low) [___________]

(Repeat for each feature)

5. Data Inputs:

List the types of data the tool will need access to. E.g., Customer data, sales data, web analytics, etc.

[___________]

(Add more as needed)

6. User Journey & Experience:

Describe a typical user's interaction with the tool, from start to finish.

Step: [___________]

Description: [________________________________________________________________]

Expected Outcome: [________________________________________________________________]

(Repeat for each step)

7. Integration & Compatibility:

List any platforms or systems the tool needs to integrate with. E.g., CRM systems, e-commerce platforms, social media, etc.

[___________]

(Add more as needed)

8. Constraints & Limitations:

Identify any potential obstacles, e.g., data privacy concerns, technical constraints, budget limitations, etc. (API internal vs external)

[___________]

(Add more as needed)

9. Stakeholders & Teams Involved:

List all departments or individuals who have a stake in this tool's development and usage. (This quantifies the benefits to this tool and how it can be used to help the team).

Team/Individual: [___________]

Role: [___________]

Responsibilities: [________________________________________________________________]

(Repeat for each stakeholder)

10. Milestones & Timeline:

Define the key phases of the project and the estimated completion dates.

Milestone: [___________]

Description: [________________________________________________________________]

Deadline: [DD/MM/YYYY]

(Repeat for each milestone)

11. Budget (Optional):

Outline the estimated costs associated with developing and maintaining the tool. (Sourcing a contractor, external tools needed, etc.)

Development Costs: [___________]

Maintenance Costs: [___________]

Other Costs (e.g., training, integration, etc.): [___________]

Total Estimated Budget: [___________]

A simple tracker to keep items organized (spec sheets, product notes, etc). It’s editable and can link to files from your Google Drive automatically, keeping sharing seamless.