As product managers, we often find ourselves backed into a corner when tasked with go-to-market strategies, pricing models, or pitching business cases. We need to show how ‘X’ number of customers at ‘Y’ price will lead to a projected ‘Z’ revenue. Enter Generative AI to the rescue! With AI tools like Copilot and ChatGPT, we can navigate these challenges more effectively and make data-driven decisions with ease. In this post, we’ll focus on the ‘X’ factor by imagining our CEO, inspired by their pet turtle’s recent misadventure of mistaking wrapping paper for a flower, tasks you with an AI initiative for a “Petcare on Demand” service. The turtle needed after-hours care while your CEO was on the road, sparking the idea. Your mission, should you choose to accept it, is to use AI to estimate the market potential and develop a go-to-market strategy that will make even the most skeptical stakeholder say, “Wow, I never thought of using Gen AI it that way!”
Understanding TAM, SAM, and SOM
First, let’s familiarize ourselves with TAM, SAM, and SOM. Think of these as parts of
the funnel to whittle down the massive market potential into actionable insights:
- Total Addressable Market (TAM): This is your market’s Mt. Everest – the total revenue if you had 100% market share. Imagine every pet owner on the planet using your app.
- Formula: TAM = Total Potential Customers × Average Annual Revenue per Customer
- Serviceable Available Market (SAM): A more realistic slice of TAM. It’s the portion your product can actually serve, considering factors like geography and target demographics.
- Formula: SAM = Target Market Potential Customers × Average Annual Revenue per Customer
- Serviceable Obtainable Market (SOM): The smallest yet sweetest slice. This is what you can realistically capture after accounting for competition and your market strategy.
- Formula: SOM = SAM × Projected Market Share
Prompt Engineering: The Secret Sauce
To get the most out of generative AI, we need to master the art of prompt engineering. It’s like wielding a superpower, knowing just the right words to use to get the perfect response. Here are a few techniques to keep in your toolkit:
- Zero-Shot Prompts: Ask the AI directly with no context. Great for straightforward tasks.
- Few-Shot Prompts: Provide a few examples rather than lengthy instructions. Perfect for more nuanced tasks, especially where complex structures are involved.
- Chain-of-Thought Prompts: Engage in a strategic, step-by-step dialogue with the AI to tackle complex problems. Imagine this as working with your product trio.
The MoE Toolbox: Mixing and Matching
Just like a handyman has multiple tools in their toolbox, we have multiple AI tools for product managers at our disposal. Here’s how to mix and match them for the best results:
- Use search engines for search work. Even though tools like CoPilot are augmented with ChatGPT, they still excel at finding existing information.
- Use AI Chatbots for Generative AI work. While tools may be augmented with Copilot, the strength of tools such as ChatGPT or Claud lies in creating new content and ideas.
- Use an MoE-like approach. The best result is obtained by incorporating and combining a mixture of experts, or in this case, responses from multiple models.
TL; DR: Use the right tool for the job, and don’t be afraid to get creative and combine the use of the two! For example, cutting-and-pasting responses from one to another.
Example: Estimating Market Potential for “Petcare on Demand”
Setting the Scene
Getting back to our original dilemma, your boss, inspired by their pet turtle’s recent adventure of eating wrapping paper thinking it was a flower, tasks you with an AI initiative for a “Petcare on Demand” service. The turtle needed after-hours care while your boss was on the road, sparking the idea. Your mission? Use AI to estimate the market potential and develop a go-to-market strategy.
Step 1: Define the Problem
Why: Clearly defining the problem ensures the AI gives you relevant data and insights.
Prompt to Copilot: “Define products or services that connect pet owners with local pet care providers for services like dog walking, pet sitting, and grooming. Describe key features and target audience.”
Outcome: You get a problem description in the context of what the CEO wants, ensuring subsequent data aligns well with your target product outcome.
Step 2: Identify Key Data Sources
Why: You need to know where to look to build your TAM/SAM/SOM based on reasonably reliable and suitable market data.
Prompt to Copilot: “List 5 authoritative sources for pet industry market data in the US.”
Outcome: A list of reputable sources to fuel your analysis.
Example Output: 1. US Census Bureau’s American Community Survey: Pet ownership by household and location. 2. APPA National Pet Owners Survey: Pet ownership trends and spending habits. 3. IBISWorld’s Pet Services Reports: Market size, growth, and key players. 4. Euromonitor’s Pet Care in the USA: Market trends and consumer behavior. 5. Packaged Facts’ US Pet Market Outlook: Market growth forecasts.
Step 3: Gather TAM Data
Why: Making a guestimate of the total market size based on available and adjacent market data helps you see the full revenue potential.
Prompt to Copilot: “Using APPA and IBISWorld data, estimate the total number of pet-owning households in the US and their average annual spending on pet services.”
Outcome: An estimation of the TAM, giving you the big-picture view.
Example Output: – 70% of US households own a pet (~90.5 million homes). – Total revenue for pet services is $9 billion. – Average annual spending on pet services: $100 per household.
Step 4: Collect SAM Data
Why: To avoid boiling the ocean, we narrow down to SAM to focus your efforts on where you can realistically make an impact.
Prompt to Copilot: “Identify the top 5 US metro areas with the highest pet ownership and spending habits.”
Outcome: Insights into specific markets where your app can thrive.
Example Output: 1. Los Angeles: 72% pet ownership, $120 annual spending. 2. New York City: 70% pet ownership, $110 annual spending. 3. Chicago: 68% pet ownership, $105 annual spending. 4. Dallas-Fort Worth: 67% pet ownership, $100 annual spending. 5. Atlanta: 66% pet ownership, $95 annual spending.
Step 5: Analyze Data with ChatGPT
Why: To understand trends, growth opportunities, and challenges, and make informed decisions.
Prompt to ChatGPT: “Analyze the market potential for ‘Petcare on Demand’ in the US using the provided TAM and SAM data.”
Outcome: A comprehensive analysis of market potential, highlighting key trends and opportunities.
Step 6: Estimate SOM
Why: Estimating SOM tells you what you can realistically capture, giving you a clear target to aim for.
Prompt to ChatGPT: “Estimate the Serviceable Obtainable Market (SOM) for ‘Petcare on Demand’ in the top 3 metro areas, considering market saturation and competition.”
Outcome: A realistic projection of the obtainable market share in targeted metro areas.
Step 7: Develop Strategies
Why: To create a targeted go-to-market plan that maximizes your chances of success, using well-known frameworks to ensure thorough planning.
Prompt to ChatGPT: “Based on this session, start scaffolding a draft go-to-market plan by first mapping a Ron Zemke and Chip Bell customer journey map, then filling out Alexander Osterwalder’s business model canvas.”
Outcome: A strategic plan for market entry, covering targeting, partnerships, promotion, and differentiation using recognized frameworks.
By following these steps and using specific data sources and tailored prompts, you can leverage both Copilot and ChatGPT to better guestimate the TAM, SAM, and SOM for your “Petcare on Demand” app effectively. This method adds applicable or adjacent data to help your scaffold a strategy as smooth as a well-groomed poodle.
Conclusion
Now that you have a Generative AI approach to shaping a draft market analysis, let’s talk about how to use AI moving forward. Should we lean on generative AI as a source of truth? No way! Should we leverage generative AI to jump-start great conversations with real, live human beings? Absolutely!
And should we take ai courses for product managers from Productside via their “AI Innovation for Product Managers” class to learn how to build the right thing using AI? You betcha! AI is a fantastic tool, but the real magic happens when we combine its insights with our creativity and strategic thinking. So, dive in, get the training, and start using AI to make smarter, more informed product decisions.