# First product brief

Pretectal · Updated October 2, 2026
Guide: https://pretectal.com/start-building

Use this brief to turn one real task into a small prototype. Replace the prompts below with your own decisions. Keep the brief beside your project and update it when the scope changes.

## 1. One person and one task

- Product / working name:
- Person or team it serves:
- Task they need to complete:
- How they do it today:
- Where they get stuck:
- Who confirmed this problem matters, and when:
- Example of a correct result:

## 2. The smallest useful version

- First action the user takes:
- Information they provide or select:
- What the product does:
- Result the user receives:
- What happens with empty, incorrect, or unexpected input:
- Three features we are deliberately leaving for later:

Write three observable acceptance checks. Each should name an input or action and the expected result.

1.
2.
3.

## 3. Building approach

- Tool chosen and why it fits this task:
- Where the preview lives:
- Sample data and content we may use:
- Where working versions are saved:
- How to restore the previous working version:
- Existing project instructions the AI must follow:

Will AI help build the product, run inside the finished product, or both?

- AI's intended contribution:
- Decisions the product owner will make:
- If the product uses live AI: how outputs, failures, and costs will be evaluated:
- Connections, accounts, or shared data the prototype does not yet need:

## 4. First build request

Help me build a small prototype for this task.

USER:
CURRENT TASK:
SUCCESS:
INPUTS:
FIRST VERSION:
OUT OF SCOPE:
ACCEPTANCE CHECKS:

First identify missing information and propose the smallest implementation plan. Use sample data. Do not connect live services or publish the prototype. Build one step at a time. After each change, explain how I can test it, what remains untested, and how I can return to the last working version.

## 5. Observe the first use

- Tester and date:
- Task they were given:
- Version and device tested:
- Could they finish without coaching?
- Correct result reached?
- Where they hesitated or needed help:
- Most important fix:
- Result after retesting:
- Feedback to revisit later:

If you measure improvement, record the same task under both processes. Include the date, conditions, sample size, and what you actually measured. Do not infer time saved from sign-ups or a successful demonstration.

- Previous process observation:
- Prototype observation:
- What this comparison can and cannot establish:

## 6. Handoff

- Status: prototype / ready for a limited trial / in regular use
- Version reviewed and date:
- Named owner:
- Intended users and approved use:
- How to update content:
- Where code and data are kept:
- Access rules, if applicable:
- Known limits and unresolved issues:
- How to report a problem:
- How to restore a working version:
- Next feedback date:
- Criterion for continuing, changing, or retiring the tool:

Companion checklist: https://pretectal.com/resources/first-product-checklist.md