As generative AI continues to gain momentum, you might find yourself bombarded with courses and articles offering "plug-and-play" prompts—short, ready-to-use instructions for various AI tools. While these prompts can work for simple tasks, they typically lack the depth to tackle more complex product management challenges.
Sometimes, a brief prompt seems like it’ll work, but it can lead to surface-level or incomplete answers. You might find yourself asking for revisions or clarifications, which wastes time.
Generic prompts usually lack relevance to your specific situation, produce answers that need a lot of refinement, and lead to more back-and-forth iterations.
For product managers who want to make the most of AI, it’s essential to move beyond quick prompts. Instead, focus on crafting detailed, context-rich prompts that guide AI to give you more insightful and relevant results.
Why Does Prompt Quality Matter in Product Management?
In simple terms, a prompt is what you give to an AI to get something back. But it’s not just about asking a question—it's about giving the AI enough information to deliver an answer that actually helps you with your product management work. Whether defining user personas, prioritizing features, or creating product roadmaps, AI works best when given clear context.
For instance, asking AI to “generate a user persona” might give you something useful, but a more detailed prompt that outlines the desired format, persona’s role, goals, and challenges will provide a much more useful output.
The Importance of Context in Prompts
AI doesn’t automatically know the nuances of your specific product or business context. This is where you, as a product manager who has interviewed customers, come in. By adding that context to your prompts, you help AI understand the bigger picture and provide outputs that align with your real goals. Here's why context-rich prompts matter.
Accuracy
AI can only be as accurate as the information it’s given. Without the right context, you might get generic responses that don’t really solve your problem. A well-structured prompt gives AI enough direction to deliver results that are directly relevant to your business needs.
Avoiding Misunderstandings
Sometimes, AI might "hallucinate"—that is, generate information that sounds plausible but isn’t quite right—if the prompt is too vague. Providing clear context reduces the chance of this happening and keeps your outputs aligned with your actual needs.
Facilitating Team Collaboration
When you craft clear prompts, the outputs become useful discussion starters. We use the popular Quartz Open Framework to define the phases of product management from idea to market. Whether you're brainstorming in the Quartz “Define” phase or refining solutions in “Create,” these outputs will be structured in a way that’s easy for your team to engage with and act upon.
Leverage your Team
Evaluate the accuracy of the AI output with your personal experience, then share the result with some respected team members. They’ll often see something you missed, providing an insight that may lead your thinking in a new direction. Just as a team retrospective identifies process challenges, a peer review is a quick way to validate and improve your work.
The Building Blocks of a Great Prompt
To craft a great prompt, focus on being clear, specific, and structured. Here’s a simple breakdown of the key components:
Role Framing
Start by defining the role AI will take on. This helps guide the context of the response. Example: “Imagine you’re a product manager working on a SaaS product for tech startups.” The word ‘imagine’ isn’t actually necessary, but it makes more sense to some people.
Task Description
Tell AI precisely what you want it to do. Be specific about the task and the outcome you're hoping for. Example: “Create a detailed user persona for a mid-level marketing manager who uses our analytics platform to optimize campaigns.”
Provide Background Context
Share relevant details about your product or audience so AI can better understand the task at hand. Example: “The platform helps marketing teams in tech startups visualize real-time data and integrate with common CRM tools.”
Desired Output Format
Explain how you want the information to be presented. Example: “Please format the persona with: Name, Job Title, Goals, Pain Points, Tools Used, and Daily Workflow.”
Include Examples
Offering examples can clarify your expectations and guide the AI’s output. Example: “For pain points, one might be: ‘Struggles to consolidate reports from multiple tools.’”
Define Success Metrics
Let AI know how you’ll evaluate its output. Example: “The persona should be realistic and useful in a product design workshop.”
Example: A Persona Using These Prompts
ChatGPT 4o provided the following response for a persona using the following prompt:
Create a detailed user persona for a mid-level marketing manager in the banking industry who uses our analytics platform to optimize campaigns. Please include the following:
Name: A typical name for the persona
Titles: Some typical titles they may hold
Description: General description of the persona and relevant context
Typical Day: What do they do all day? What are they responsible for? How do they interact with co-workers and customers?
Skills: The abilities and typical products they know well
Technology: The type of devices and applications they can access
Segments: Target market segment
Goals and Aspirations: Describe what will improve their performance or benefit the business.
Friction: Problems that cause the most trouble in their daily work
How to reach them: As Marketing and Sales versus as Product and Support.
Be as brief as possible.
Response:
User Persona: Rebecca Simmons
1. Name:
Rebecca Simmons
2. Titles:
Marketing Campaign Manager, Digital Marketing Analyst, Marketing Operations Specialist
3. Description:
Rebecca is a 34-year-old mid-level marketing manager at a regional bank. She manages multi-channel campaigns aimed at increasing customer engagement, acquiring new accounts, and driving cross-sell opportunities. She works closely with data teams to analyze campaign performance and optimize strategies based on customer insights.
4. Typical Day:
Morning: Reviews campaign KPIs, checks in on current campaign performance, and meets with her team to align on priorities.
Midday: Collaborates with data analysts to review customer segmentation and performance dashboards. Works with compliance teams to ensure marketing messages meet regulatory standards.
Afternoon: Develops briefs for upcoming campaigns, optimizes live campaigns (e.g., adjusting budgets or creative), and collaborates with sales teams on co-branded materials.
Interactions: Works cross-functionally with data, compliance, creative, and sales teams. Occasionally interfaces with customers for focus groups or surveys.
5. Skills:
Campaign management, A/B testing, and customer segmentation.
Strong understanding of banking products (e.g., loans, credit cards).
Tools: Adobe Analytics, Google Analytics, Salesforce Marketing Cloud, and Tableau.
Familiar with GDPR, CCPA, and banking industry regulations.
6. Technology:
Devices: Windows laptop, dual monitors, iPhone.
Applications: Marketing automation platforms (e.g., HubSpot, Marketo), business intelligence tools, and collaborative apps (e.g., Slack, Microsoft Teams).
7. Segments:
Retail banking, wealth management, small business banking.
8. Goals and Aspirations:
Deliver more personalized, data-driven campaigns to improve conversion rates.
Increase ROI by identifying underperforming segments and reallocating resources.
Position herself for a senior role by mastering analytics and campaign optimization.
9. Friction:
Inconsistent or siloed customer data across systems.
Slow response times from compliance or IT teams when launching campaigns.
Difficulty demonstrating ROI for upper management due to lack of unified reporting tools.
10. How to Reach Them:
Marketing and Sales: Target LinkedIn ads, host webinars on campaign analytics in banking, attend industry conferences (e.g., American Bankers Association).
Product and Support: Offer quick-start guides, provide templates for campaign optimization, and maintain responsive customer support via chat or phone.
The prompt was designed for simple copy-paste into the Personas template from Quartz. If you’ve ever had a stakeholder not read critical information, it’s refreshing to distill what they need to a single page. The more context you add, the better the AI can help you!
Tips for Building Better Prompts
Iterate and Refine: Start with a draft and tweak based on what AI gives back.
Use Templates: Create templates for recurring tasks like persona creation. It saves time. Quartz offers several ready-to-use templates for your playbook.
Test with Your Team: Ask for feedback to see if your prompt needs more context.
Evaluate Outputs: Treat AI results as a starting point and refine them with your own insights.
Why Context-Rich Prompts Matter
Mastering prompt quality isn’t just a technical skill—it’s a game-changer. By crafting context-rich prompts, AI becomes more than just a tool—it becomes a partner that helps you deliver more user-centric and impactful products.
While context-rich prompts are great for detailed tasks, shorter prompts can be sufficient for quick activities like summarizing product features or creating basic lists. Understanding when to use each type is essential for efficiency.
While AI tools have been trained on lots of material, they lack your experience and expertise with your company, market, and products. Like a book-smart intern, you still need to check its work.
AI works best when you give it the right input. Take the time to craft thoughtful prompts, and you’ll see how AI can become a true asset to your product management process.
In today’s product management landscape, AI is a powerful ally. But to get the most out of it, you must provide detailed, context-rich prompts. By focusing on context, you guide AI to deliver more relevant, actionable, and strategic insights that will help your team move faster and smarter.
This part teaches the LLM enough about Quartz so that it can provide helpful output in the form needed for the template. Without it, the model isn’t as well aligned with what we want it to do.
Want to upskill your team with enhanced skills? Contact us about The AI-Empowered Product Manager and other learning programs.