Tools We Use

This guide covers the categories of tooling a product team needs and the conventions that matter inside each. The specific tools (Linear vs Jira vs Shortcut, Notion vs Confluence vs Coda, Claude vs ChatGPT vs Cursor) are implementation details. Pick what your company already runs on. The conventions below apply regardless.

The categories

Category	What it does	Example tools
Issue tracker	Issues, roadmap, project tracking, customer feedback	Linear, Jira, Shortcut, GitHub Issues, Asana
Long-form docs	RFCs, specs, research, process documentation	Notion, Confluence, Coda, Google Docs, Obsidian
Source control	Code, PRs, playbook repository	GitHub, GitLab, Bitbucket
Design	Wireframes, mockups, prototypes	Figma, Sketch, Penpot
Customer comms / support	Tickets, calls, transcripts	Zendesk, Intercom, Plain, Front
Sync comms	Channels, DMs, async updates	Slack, Microsoft Teams, Discord
AI assistant	Feedback synthesis, draft assist, research summarisation	Claude, ChatGPT, Cursor, Amp, Gemini
Recording	Async demos, walkthroughs, customer call recordings	Loom, Tella, Vidyard, Granola

You only need one tool per category. More than that creates copy-paste tax and split-brain documentation.

Issue tracker (e.g. Linear / Jira)

What lives here:

Issues (bugs, feature requests, customer feedback)
Roadmap (what you’re building, what’s next)
Project tracking (status, ownership, progress)

Conventions:

Ticket Naming

Problem-focused, not solution-focused
- ✅ “Users can’t recover from a failed action”
- ❌ “Add a recovery button”
Customer language, not internal jargon
- ✅ “Notifications stop arriving after a user changes email”
- ❌ “Email change event listener edge case”

Project Linking

One project per RFC
RFC link in project description
Keep project status synced with RFC status (Draft → In Progress → Shipped)

Labels

Use product area labels matching your domains
Don’t over-label; 2–3 labels max per issue

Long-form docs (e.g. Notion / Confluence)

What lives here:

RFCs (product requirements documents)
Specs (product, design, technical)
Research docs (customer calls, competitive analysis, problem briefs)
Process documentation (this playbook)

Conventions:

File Naming

RFCs: [Project Name] RFC
Research: [Project Name]: [Research Type] (e.g., “Retry Logic: Customer Research”)

Linking

Always link RFC ↔ tracker project (bidirectional)
Link RFC → research docs in Evidence section
Link specs → RFC in Context section

Status

Keep RFC status in sync with tracker project status
Archive old specs after launch (don’t delete, keep for reference)

Source control (e.g. GitHub / GitLab)

What lives here:

Code
Pull requests
Playbook repository (markdown source, optionally synced to your docs tool)

Conventions:

Playbook Changes

PRs with playbook: prefix in title (e.g., “playbook: Add discovery guide”)
Tag the product team for review on process changes
If you have a docs sync, the sync runs on merge

AI assistant (e.g. Claude / ChatGPT / Cursor)

Whatever your team has approved enterprise access to: Claude (Claude Code, claude.ai), ChatGPT, Cursor, Amp, Gemini. The capabilities matter more than the brand.

What to use it for:

Synthesising customer feedback from support tickets, calls, and channel threads
Drafting RFC Evidence sections (turns raw feedback into structured problem statements)
Summarising competitive research
Generating job stories from interview transcripts

When to use:

You have 10+ pieces of feedback on the same topic → let the assistant find patterns
You need to draft Evidence but have raw notes → feed it the notes, edit the output
You’re comparing 3+ competitors → let the assistant summarise differences, then verify

When NOT to use:

Strategic decisions (the assistant doesn’t know your strategy)
RICE scoring (needs human judgement)
Customer conversations (don’t outsource the relationship)

💡 Always use your company’s approved enterprise account, not a personal one. Personal accounts may not have the right data-handling guarantees for customer information.

Access (template)

New to product? Get access to one tool per category above. Specific tools and invite paths will be company-specific; keep an internal copy of this section with the current names and onboarding owners filled in.

Issue tracker: [your tool], [how to request access]
Long-form docs: [your tool], [how to request access]
Source control: [your tool], [how to request access]
AI assistant: [your approved tool(s)], [account/access process]
Comms: [your tool], [channels to join on day 1]