Product-Persona-Prompt Mapping
Map products and personas to the specific questions buyers ask AI
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 5
- Confidence
- 96%
Product-Persona-Prompt Mapping begins with the company's offerings and the customer profiles most relevant to each offering. The marketer then identifies the problems, comparisons, and recommendations those people could ask an answer engine about at different stages of the buyer journey. These combinations generate a structured prompt set that simulates actual customer demand. Unlike conventional SEO research, which may concentrate on a limited collection of short keywords, this method accounts for highly contextual questions shaped by role, location, need, product, and purchase stage. The resulting prompts become the measurement layer for AEO: each one can be submitted repeatedly to answer engines to reveal whether the brand appears, which competitors appear, and where optimization effort should be directed.
Origin
Extracted from Marketing Against The Grain, where Asia Frost and Beerie Amiel demonstrate the prompt-generation framework embedded in HubSpot AEO.
Core principles
- 01AI-search questions vary by product, persona, problem, and journey stage.
- 02Specific prompts reveal opportunities that broad SEO keywords miss.
- 03A useful prompt set should simulate questions real buyers are likely to ask.
- 04Prompt discovery must precede visibility measurement.
How to run it
- 1
Inventory the products
List the products, services, or use cases for which the brand wants to earn recommendations. Keep distinct offerings separate so their customer questions do not become blurred.
Pro tip Start with commercially important products rather than attempting to map the entire catalog at once.
Watch out A generic brand-level prompt set can conceal product-specific weaknesses.
- 2
Define the personas
Identify the ideal customer profiles or personas relevant to each product. Capture differences in role, context, location, needs, and purchasing authority.
Pro tip Use actual customer research and sales conversations when defining persona distinctions.
Watch out Do not assume every customer asks the same questions.
- 3
Generate contextual questions
For every product-persona pairing, generate the likely problems, comparisons, and recommendation questions that person would ask an answer engine.
Pro tip Phrase prompts naturally and include the situational detail a buyer would provide.
Watch out Avoid reducing conversational prompts to short SEO-style keyword fragments.
- 4
Map the journey stage
Classify each prompt by its role in awareness, consideration, or decision-making. This distinguishes problem exploration from direct product-selection questions.
Pro tip Include both unbranded problem prompts and explicit recommendation prompts.
Watch out Tracking only late-stage prompts misses opportunities to shape earlier consideration.
- 5
Create the tracking set
Select the prompts with the strongest relevance to customer demand and business outcomes. Use this set as the stable basis for recurring answer-engine measurement.
Pro tip Expand the set as new customer questions emerge.
Watch out An indiscriminately large set can consume effort without improving strategic coverage.
In the wild
A Dell marketer pairs enterprise laptops with IT leaders and enterprise executives. The awareness-stage set includes questions such as why laptop fleets struggle with premium performance, while the decision-stage set asks for the best premium laptop. Tracking both reveals whether Dell participates throughout the journey rather than only when its name is already known.
→ The marketer gains a commercially relevant prompt set spanning problem recognition and product selection.
Common mistakes
Treating prompts like keywords
Short, generic phrases strip away the product, persona, and problem context that makes answer-engine questions strategically useful.
Using one universal persona
Combining materially different customers obscures how their questions and expected recommendations vary.
Tracking questions without intent
A large prompt list is not useful unless its entries map to real customer problems or stages of the buying journey.
Is it for you?
Best for
Marketing teams that need to determine which prompts matter before measuring or improving their answer-engine visibility.
Not ideal for
Teams without defined products, audiences, or enough customer knowledge to distinguish meaningful prompts from speculative ones.
From the transcript
“we actually have to pick our products and ICPs or ideal customers' profiles or the different personas that we're going to be targeting.”
“what we're gonna do with that is we're actually gonna go and help you figure out what are the best prompts to track for each…”
“we really have to accommodate for that and think about it in a much more specific uh persona and like product specific, problem-specific way than…”
From the episode
How Brands Win in ChatGPT Search