Focus Groups + Market Testing

Level 4 · Course 26

Your agent will build a convincing case for its own work. That is useful when drafting and dangerous when deciding. A focus group breaks the loop. You give controlled context to fresh evaluators and ask them what they see before your main agent explains why it is right.

This course installs a lightweight testing engine. Your agent defines what is being tested, chooses evaluator profiles, controls context, collects verdicts, extracts patterns, and decides what changes. The goal is not to outsource judgment. The goal is to get outside reads before you spend money, ship copy, launch a feature, or trust an idea.

A good focus group is small, fast, and specific. Three evaluators can catch what one loaded agent misses. After this course, your factory has a way to test products, pricing, copy, workflows, and assumptions against agents that do not already live inside your story.

The Profile Format

Every evaluator in a focus group starts as a profile. The profile controls what context they carry, what they're sensitive to, and what they should ignore. Same format every time — the differences between profiles are the data.

Who they are
One sentence. Role, situation, and relationship to the product being tested.
Pain level
Low / Medium / High — plus one sentence on why. What problem are they trying to solve, and how urgently?
Money sensitivity
$10 matters / $100 is fine / $500 is nothing. Sets how they read pricing without being told what to think about it.
What to show them
URL, screenshot, or copy — exactly what they receive. Nothing else. No context, no framing, no hints.
What to ignore
Instructions to suppress: don't mention price, don't mention the brand, don't explain the backstory. Keeps the test clean.

Worked Example — Restaurant Menu Page

A restaurant owner wants feedback on a new menu page before it goes live. They run three evaluator profiles, each receiving the page cold with no context about who built it.

The Evaluator Profiles

Profile A — Frequent diner who tries new spots weekly

Pain: low. No particular urgency — they're browsing, not desperate. Money: medium ($100 is fine for a dinner out). Needs: menu clarity and pricing they can scan in under 30 seconds. What to show: the menu page URL. What to ignore: don't mention the restaurant name or who owns it.

Profile B — Parent of three looking for family-friendly options

Pain: medium. Picky eaters at the table, possible allergy concerns. Money: high ($500 is nothing — eating out is frequent). Needs: kids menu visibility, allergy information, indication of noise level or atmosphere. What to show: the menu page URL. What to ignore: don't mention the owner or any backstory.

Profile C — First-date planner looking for the right ambience

Pain: low. Wants to get it right, not fix a crisis. Money: low ($10 matters — they're watching the spend). Needs: photos of the interior, whether reservations are required, a clear sense of the vibe from copy alone. What to show: the menu page URL. What to ignore: don't mention price increases, ownership, or anything about the business side.

The Question Template

Every evaluator gets the same five standard questions. After each standard question, one customized follow-up can be added based on what that profile cares about most.

  1. Describe what this page is offering in one sentence. No interpretation — just what you see.
  2. Would you make a reservation based on this page alone? Why or why not?
  3. What's missing that you'd want before you felt confident? List anything.
  4. How does the pricing feel — too high, unclear, or appropriate? First impression only.
  5. Honest first impression in two sentences. No softening. Say what you'd say to a friend.

The Synthesis Format

After all three evaluators respond, your agent produces one synthesis. The format is fixed — patterns first, outliers second, fix list third.

Agent synthesis: "All three evaluators said the menu was hard to navigate on mobile — items loaded but category headers didn't stand out. Two of three flagged the same gap: no indication of whether reservations are required or walk-ins are welcome. Profile B couldn't find allergy information anywhere on the page and would not have made a booking. Pattern: navigation and booking clarity are the blockers. Outlier: Profile C mentioned the font choice felt 'too corporate for a neighborhood spot' — worth noting but not a priority fix. Fix list: mobile nav, reservation status visible above the fold, allergy section added or linked."

You decide: what's the next thing you ship that someone else should see first? Three evaluators, five questions, one afternoon.

Your Agent PDF

Your agent executes the PDF. You read the page. No copying. No manual setup.

Download PDF — Course 26

Questions? [email protected] — My Courses