Report back

Report Back

Turn a day's outreach activity into a clear digest the user can drop into Slack, email a manager, or paste into their weekly review doc.

When this skill triggers

• User asks for an end-of-day summary
• User asks to "wrap up the day"
• User pastes activity data and asks for a report
• User is prepping a Slack update, weekly review, or CRM notes
• User asks about outbound performance today or this week

Inputs the skill expects

The user provides some subset of these numbers (missing ones get flagged, not fabricated):

• Sends today (LinkedIn DMs, connection notes, emails)
• Replies received
• Reply breakdown (positive / objection / DNC / no-fit)
• Meetings booked
• Meetings held
• Drafts pending review
• Follow-ups sent
• Any specific wins or losses worth noting

Output format

Return a clean structured digest. Two formats depending on what the user needs:

Format A — Slack-ready digest (default)

📊 OUTREACH DIGEST · [Day, Date]

SENDS
→ LinkedIn: [X]
→ Email: [X]
→ Connection notes: [X]
Total: [X]

REPLIES ([X] total)
→ Positive: [X]
→ Objections: [X]
→ Not-a-fit / DNC: [X]

MEETINGS
→ Booked today: [X]
→ Held today: [X]

IN THE QUEUE
→ Drafts pending review: [X]
→ Follow-ups scheduled tomorrow: [X]

WORTH FLAGGING
[1-2 lines on the most important thing that happened — a big reply, an objection worth learning from, a signal that fired, etc. This is where operator taste comes in.]

TOMORROW'S FIRST MOVE
[The single most important thing to do first thing tomorrow]

Format B — Weekly review (when user asks for a week summary)

📊 WEEKLY OUTREACH REVIEW · Week of [Date]

VOLUME
→ Sends this week: [X] (vs. last week: [X])
→ Reply rate: [X]% (vs. last week: [X]%)
→ Meeting book rate: [X]%

WINS
→ [Bullet 1]
→ [Bullet 2]
→ [Bullet 3]

LOSSES / LEAKS
→ [Bullet 1 — a specific pattern that isn't working]
→ [Bullet 2]

WHAT WORKED (that we should double down on)
→ [Pattern or signal type or message angle that outperformed]

WHAT DIDN'T (that we should retire or fix)
→ [Pattern or signal type that underperformed]

NEXT WEEK'S PRIORITY
[One-sentence commitment on the biggest lever to pull]

Rules

Never do these:

• Fabricate numbers. If the user didn't provide a metric, mark it as "not tracked" or "not provided," never guess.
• Add motivational filler ("Great job today!" or "You're crushing it")
• Include commentary on numbers unless the user asked for analysis
• Turn a report into a coaching session unless the user asked for coaching
• Add graphs or charts unless the user asked (this is a text digest, not a dashboard)

Always do:

• Preserve the exact numbers the user provided
• Flag anomalies in the "worth flagging" section (a 20% reply rate is worth calling out; so is a 0.5% one)
• Recommend one specific action for tomorrow, not a list
• Keep the tone flat and operational — this is an ops digest, not a pep talk

Examples

Example 1 — daily digest

Input:

"Wrap up today: 8 LinkedIn DMs sent, 2 emails, 3 connection notes. Got 4 replies total — 1 positive (they want a call), 2 timing objections, 1 DNC. Booked 1 meeting (positive reply). Have 3 drafts waiting for me to review. 5 follow-ups scheduled tomorrow. The DNC was a Head of Sales at a Series C company who told me my messaging was tone-deaf — worth learning from."

Output:

📊 OUTREACH DIGEST · Tuesday, Jul 8

SENDS
→ LinkedIn: 8
→ Email: 2
→ Connection notes: 3
Total: 13

REPLIES (4 total — 30.8% reply rate)
→ Positive: 1
→ Objections: 2 (both timing)
→ Not-a-fit / DNC: 1

MEETINGS
→ Booked today: 1
→ Held today: not provided

IN THE QUEUE
→ Drafts pending review: 3
→ Follow-ups scheduled tomorrow: 5

WORTH FLAGGING
The DNC feedback from the Series C Head of Sales — "tone-deaf messaging" — deserves a real look. Pull that thread, look at what triggered the reaction. Fixing this in the next 24 hours prevents a pattern from calcifying.

TOMORROW'S FIRST MOVE
Review the 3 pending drafts before touching anything new — clearing the review queue prevents the send-first-review-later drift.

Example 2 — weekly review

Input:

"Weekly review please. This week: 47 sends (last week 32), 6 replies (12.7%, last week was 15.6%), 2 meetings booked. Wins: got a Series B founder to reply positively, closed a warm re-intro from a champion job change. Losses: 3 timing objections in a row from Series C companies — feels like something. What worked: the champion-job-change angle. What didn't: the funding-round hook — sent 8 with 0 replies."

Output:

📊 WEEKLY OUTREACH REVIEW · Week of Jul 1

VOLUME
→ Sends this week: 47 (vs. last week: 32) — up 47%
→ Reply rate: 12.7% (vs. last week: 15.6%) — down 2.9 pts despite volume increase
→ Meeting book rate: 4.3% (2 of 47)

WINS
→ Series B founder positive reply → real pipeline
→ Champion re-intro closed → clean warm play
→ Weekly volume up 47% without a total collapse in reply rate

LOSSES / LEAKS
→ 3 timing objections from Series C companies in a row — pattern, not noise
→ Reply rate slipped 2.9 points as volume went up (classic diminishing return)

WHAT WORKED (that we should double down on)
→ Champion job change angle. Highest conversion by far — one warm re-intro is worth 20 cold sends.

WHAT DIDN'T (that we should retire or fix)
→ Funding round hook: 8 sends, 0 replies. Either the message is wrong or the timing (post-announcement week is peak pitching, everyone's saturated). Fix: try the funding hook 30 days out instead of week 1, or drop it entirely.

NEXT WEEK'S PRIORITY
Double the champion-job-change queue — set aside 2 hours to build a fresh list of ex-customers who've moved to ICP-fit accounts. That's your highest-ROI channel and it's not saturated.

Example 3 — incomplete data

Input:

"Wrap up today. I sent 10 messages. Got 2 replies."

Output:

📊 OUTREACH DIGEST · [Today]

SENDS
Total: 10 (breakdown by channel not provided)

REPLIES (2 total — 20% reply rate)
Breakdown not provided (positive / objection / DNC)

MEETINGS
Not provided

IN THE QUEUE
Not provided

WORTH FLAGGING
20% reply rate is high — worth understanding if the 2 replies were substantive or one-line brush-offs before celebrating. If both are real, this was a strong send. If both are "not interested," volume is up but signal isn't.

TOMORROW'S FIRST MOVE
Track the channel + reply type breakdown for tomorrow's sends. The 20% number is meaningless without knowing whether it's positive interest or polite rejection.

Common failure modes

Failure: report includes generic praise or motivational commentary. Fix: Keep it operational. The user isn't asking to be complimented — they're asking for a digest.

Failure: fabricated numbers appear because the report "wants to be complete." Fix: If a metric wasn't provided, mark it as "not provided." Never fill gaps with plausible-sounding guesses.

Failure: report doesn't flag anything as unusual even when the numbers clearly indicate a pattern. Fix: Operator taste matters here. If reply rate drops 30% week-over-week, that's worth flagging. If it stays flat, don't invent something to say.

Failure: recommends 5 things for tomorrow instead of 1. Fix: Priorities compress. One clear tomorrow-morning action is more valuable than a list of five.

Related skills

• If the report surfaces an objection pattern worth studying, use objection-response to workshop new response patterns
• If a signal type consistently underperforms, use signal-triage to review the tiering rubric — the signal might be miscategorized