Gmail Wow Moments

See Wow Moments Framework for core principles and selection guidance.

Discovery Insights (the goldmine)

Best when: Discovery searches found personal signals

The discovery data contains tiered categories of personal signals. Lead with high-tier discoveries (1-2) when present.

• "I noticed you travel quite a bit" (from travel category)
• "You're active on [platform]" (from linkedin, github, instagram, etc.)
• "Looks like you have pets" (from pets category in sent mail)
• "You're using Linear for project management" (from project_mgmt, tier 2)
• "You've got a Vercel setup — side projects?" (from hosting, tier 4)

These are high-confidence because they come from actual emails — verification codes, receipts, confirmations.

Quantitative Insights

Best when: patterns data shows clear relationship metrics

Use specific numbers when they're surprising:

Conversation depth:

• "Your deepest conversation was with [X] — [N] back-and-forth messages"
• "You have [N] people you actually go back-and-forth with (multi-turn conversations)"
• "Most of your contacts are single exchanges — [N] one-and-done conversations"

Time investment:

• "You've been emailing [X] for [N] days — your longest-running conversation"
• "[X] is your newest contact — started [timeAgo]"
• "You have [N] sustained relationships (30+ days of back-and-forth)"

Confidence signals:

• "[N] of your top contacts are high-confidence (sustained, multi-thread)"
• "Some recent contacts ([N]) need more time to assess"

Look at patterns.deepestConversation, patterns.longestRelationship, patterns.multiTurnContacts, and patterns.sustainedRelationships.

Relationship Dynamics

Best when: Clear imbalance or strong bidirectional signal in contacts data

• "[X] emails you way more than you email them — you're in demand"
• "You and [X] have a real back-and-forth going — [N] messages over [spanDays] days"
• "[X] is your most active contact — and it's mutual"
• "Most of your top contacts are bidirectional — real conversations, not newsletters"

Look at contacts[].bidirectional, receivedFrom, sentTo, and new fields: spanDays, avgMessagesPerThread, confidence.

Inbox Patterns

Best when: Clear patterns in volumeByDate or labels

• "[Day] is your heavy inbox day" (from volumeByDate aggregated by day of week)
• "[Label] is where the action happens" (from labels with high counts)
• "Your inbox is surprisingly clean — mostly real people, not noise"

Email Personality (optional)

Best when: A clear archetype emerges from the data

• The Networker — Many bidirectional contacts, broad relationships
• The Focused — Few contacts but deep engagement (high avgMessagesPerThread)
• The Broadcaster — Sends more than receives, outbound communicator
• The Hub — Receives way more than sends, people come to them
• The Deep Thinker — Few contacts but very long conversation threads

Gmail-Specific Priority

1. Discovery first — High-tier discoveries (1-2) show you learned something personal
2. Quantitative insights — Use patterns data for specific, surprising numbers
3. Relationship dynamics — Strong bidirectional patterns or clear imbalances
4. Confidence context — Mention when contacts are high-confidence vs. emerging