How to SPARK Better AI Conversations

Transform ChatGPT from answer bot to thought partner

If you're like most people, you're probably using ChatGPT and other AI assistants as fancy search engines. Ask a question, get an answer, move on with your day. Maybe you use them to help draft emails or generate images, but your interactions likely remain brief and transactional.

I was guilty of this too.

The problem is that we're still operating with what I call the "Google mental model" - where humans request information and computers provide it. This approach severely limits what's possible when working with today's AI tools.

The Missed Opportunity

When we treat AI as merely a question-answering machine, we're missing out on its potential as a thought partner. We're using a Ferrari as if it were a bicycle.

Based on insights from innovation experts Jeremy Utley and Ethan Mollick, I've completely transformed how I approach these interactions. The results have been eye-opening - more nuanced responses, better solutions to complex problems, and genuinely surprising insights I wouldn't have discovered otherwise.

Introducing the SPARK Framework

To transform your AI interactions from transactional exchanges to meaningful collaborations, I've developed the SPARK framework:

S - Set a Goal

Clearly define what you want to achieve. Rather than asking a simple question, frame your interaction around a specific outcome.

Instead of: "What strategies work for product launches?"
Try: "I need to develop a launch strategy for our new productivity app targeting remote professionals."

P - Provide Context

Share relevant details about your situation. The more context you provide, the more tailored and useful the response will be.

Instead of: "What are good marketing channels for a productivity app?"
Try: "For our productivity app launch, we have a small team of 5, a $30K marketing budget, and 3 months to prepare. Our app integrates with Slack and Microsoft Teams, and we've already secured 200 beta users who've provided positive feedback."

A - Assign a Role

Tell the AI what expertise it should adopt. This helps frame its approach and can dramatically change the quality and focus of responses.

Instead of: Letting the AI respond to your productivity app questions as a general assistant
Try: "You're a product marketing director who has successfully launched productivity SaaS products at both startups and established companies like Atlassian and Asana."

R - Reveal Your Role

Define how you'll actively participate in the conversation. This establishes expectations and creates more productive back-and-forth.

Instead of: Passively receiving advice about your app launch
Try: "Ask me specific questions about our target market, competitive landscape, and available marketing channels so you can tailor your recommendations."

K - Keep It Conversational

Make it a dialogue, not a one-off exchange. The real magic happens when you engage in multiple turns of conversation.

Instead of: Ending the conversation after getting initial productivity app launch strategies
Try: "After you suggest launch strategies, I'll share our budget limitations and marketing team composition to refine the approach."

SPARK in Action: From Personal Decisions to Business Strategy

A Simple Personal Example

Recently, my wife and I were considering getting a dog for our family. Our first instinct was to ask a simple question:

"What kind of dogs are good around kids?"

But instead, we applied the SPARK framework:

Set a Goal: "I need help deciding what type of dog we should consider getting for our family."

Provide Context: "We are a family of six with young kids. We tend to be away from the house a few hours a day. We have a yard and would like a dog that can play with the kids. Our oldest child has mild allergies to some animals."

Assign a Role: "You're an expert on dogs and the various needs and temperaments of different breeds. Your job is to help match families with the best dogs based on the needs of both the family and the dog."

Reveal Your Role: "Ask me questions about our family lifestyle, preferences, and constraints to help determine which breeds would be the best match for us."

Keep It Conversational: We then had a productive dialogue where the AI asked about:

• Our experience with dogs

• How much time we could dedicate to training

• Our activity levels

• Space constraints in our home

• Grooming preferences

• Size considerations

With each answer, the AI refined its understanding and eventually provided us with a thoughtful list of breeds we hadn't even considered, along with detailed explanations of why each would be a good match for our specific situation.

Applying SPARK in Business

The same framework can deliver even more value in professional contexts. Imagine a team needing to revamp their customer onboarding process. Instead of asking:

"What are best practices for SaaS customer onboarding?"

Here's how they could apply SPARK:

Set a Goal: "I need to redesign our customer onboarding process to improve our 30-day retention rate, which currently sits at 65%."

Provide Context: "We have a B2B productivity app with a 14-day free trial. Users need to connect their work calendar and task management tools to see value. Currently, only 40% complete all integration steps. Our support team is small with just two customer success managers handling all accounts."

Assign a Role: "You're a customer success director who has optimized onboarding flows at companies like Calendly and Monday.com, achieving 90%+ activation rates and significant improvements in retention."

Reveal Your Role: "Ask me questions about our current onboarding flow, user feedback, and technical constraints so you can provide targeted recommendations."

Keep It Conversational: This would lead to a productive dialogue where the AI might ask about:

• Current drop-off points in the onboarding funnel

• Existing customer feedback on the process

• Technical limitations for integration simplification

• Resources available for implementation

• Current metrics and measurement practices

With each answer, the AI is refining its understanding and working with you to co-create an onboarding redesign strategy, including specific intervention points, communication templates, and a phased implementation plan.

This framework scales naturally from everyday decisions to complex business challenges. In both cases, the quality of insights are dramatically better than what you'd receive from simple queries.

Beyond Simple Questions

I've applied this framework to even more complex scenarios:

• Working through strategic decisions by assigning the AI the role of various stakeholders (customer, competitor, etc.) to pressure-test my thinking

• Improving my writing by having the AI play the role of different audience members and providing feedback from multiple perspectives

• Exploring complex problems by having the AI represent experts from different backgrounds (neuroscience, behavioral science, sociology, and philosophy) and facilitate a structured debate

In each case, the quality of insight was dramatically better than what I would have received from a simple query.

The Shift in Mindset

Using the SPARK framework requires a fundamental shift in how we think about AI tools:

1. From answer provider to thought partner: The AI becomes a collaborator in your thinking process, not just a source of information.

2. From transaction to relationship: Each interaction builds on previous exchanges, creating a richer context for better results.

3. From passive consumption to active co-creation: You remain engaged and responsible for guiding the conversation toward valuable outcomes.

Getting Started with SPARK

You don't need to use all elements of the framework for every interaction. Start small:

• For your next factual question, add context about why you need the information

• When seeking advice, specify what role or expertise you want the AI to adopt

• After receiving an initial response, follow up with a clarifying question instead of ending the conversation

Over time, incorporating these elements will become second nature, and you'll find yourself having increasingly valuable exchanges.

The Cost-Benefit Calculation

Yes, using the SPARK framework takes more effort than firing off quick queries. It requires thoughtfulness about your goal, careful articulation of context, and ongoing engagement.

But consider the trade-off: a few extra minutes of setup for significantly more valuable and tailored insights. In many cases, this investment pays dividends in better decisions, more creative solutions, and deeper understanding.

Beyond Individual Use

This approach isn't just valuable for personal use. Organizations can benefit enormously by training teams to interact with AI tools using structured frameworks like SPARK:

• Product teams can explore design alternatives more thoroughly

• Marketing can generate more targeted campaign concepts

• Customer service can develop more nuanced response templates

• Strategy teams can pressure-test assumptions from multiple angles

The companies that master these collaborative approaches will gain significant advantages in creativity, problem-solving, and decision-making.

The Future of Human-AI Collaboration

As AI systems continue to advance, the ability to engage with them effectively will become an increasingly valuable skill. The most successful individuals won't be those who can craft the perfect one-shot prompt, but those who can maintain productive, iterative dialogues that leverage both human and artificial intelligence.

The SPARK framework is just the beginning. As we develop better methods for collaboration with AI systems, we'll unlock new possibilities for enhancing human creativity, decision-making, and problem-solving.

Don't just prompt. SPARK a real exchange.