── SKILL
onboarding-analyst✓ APPROVED

Know if a new customer will stick, by day 90.

Tracks Day 30 / 60 / 90 activation, scores the sales-to-success handoff, and catches accounts a fit score wrote off while they were adopting.

Inside: The 30/60/90 milestone framework, the handoff scorer, and the time-to-value read.

01 / HOW TO USE

Install it in one line, or paste it in.

1
In Claude Code (one command)
Copy the install line, paste it into your terminal, and restart Claude Code. The skill installs itself to ~/.claude/skills/ and runs automatically when it is relevant.
2
In Claude, ChatGPT, or a Project (no terminal)
Open the file, then upload it to your chat or paste its contents in. A skill is just a markdown file of instructions, so any capable AI can follow it.
New to skills? A skill is a plain-text file that teaches your AI a workflow. Point any capable assistant at it and it follows the steps, on your data.
02 / MAKE IT YOURS

Connect your context. Set it to your motion.

CONNECT YOUR CONTEXT · AND WHY IT HELPS
a CRM

it reads close date, owner, and handoff fields automatically.

a product-analytics tool

it measures real activation and time-to-value instead of asking you.

a meeting or email tool

it confirms the kickoff happened and the CSM is actually engaged.

a support tool

it catches early-ticket patterns that predict a rocky start.

SET IT TO YOUR MOTION

This was built for a CS team onboarding new customers off a CRM and a product-analytics tool. Set these to your stack:

Run any milestone model you like. The skill checks "did this customer hit the milestones you defined," so point it at your capabilities, not anyone else's.

Set thisWhat it isDefault / Example
PRODUCT analyticswhere activation livesa product-analytics tool
CLOSE datewhen the clock startsaccount created or contract-signed date
CAPABILITIESthe features that mean "activated"your product's key actions
HANDOFF fieldskickoff, notes, owner transferyour kickoff and account-notes fields
TTV targetyour time-to-value benchmarkDay-30 activation (re-tune)
03 / THE FULL SKILL

Everything the skill does, in full.

── WHAT THIS DOES

Watches every new customer through their first 90 days and tells you which ones are actually sticking. It checks Day-30, Day-60, and Day-90 activation against named capabilities, scores how clean the sales-to-success handoff was, tracks how fast each account hit first value, and flags the accounts a fit score wrote off while they were quietly adopting. It is the missing read between the deal closing and the renewal.

── THE METHOD
  1. 1
    Day-30 / 60 / 90 milestone framework

    Each checkpoint is named-capability, not a vague "good activation." A real read looks like "first capability adopted Day 12, second started Day 18, the Day-30 target capability not yet activated." You see exactly what landed and what is missing.

  2. 2
    Handoff quality scorer (multi-source)

    Did the kickoff happen inside the first two weeks? Was the deal context actually transferred, not left as stale sales notes? Has the CSM met the customer? The score reads more than the CRM, it confirms the meeting and the notes transfer across your tools.

  3. 3
    Onboarding-window champion engagement

    Engagement during the honeymoon period is uniquely diagnostic, different from a post-90-day champion check. A champion who goes quiet in the first 30 days is an early red flag worth more than the same silence at Day 200.

  4. 4
    Early false-negative catch

    A customer a fit score floored while they have already adopted several capabilities in their first 30 days is a false negative. That account belongs in the expansion pipeline now, not written off.

  5. 5
    Time-to-value tracker

    Measure the time from contract signed to first power-user threshold, and tag it against your benchmark. "Hit Day-30 activation in 18 days, top of the recent cohort" tells you more than a raw date.

── QUALITY GATES
  • Milestones are named-capability, never aggregate. You see which capability landed and which is missing, with the day.
  • The handoff scorer is multi-source, not CRM-only. It confirms the kickoff meeting and the notes transfer.
  • Time-to-value is benchmark-tagged, so a number always carries "fast or slow versus your cohort."
── OUTPUT (EXAMPLE)
EXAMPLE OUTPUT
FIRST 90 DAYS · 14 customers in the onboarding window

DAY 30 CHECK (8 past Day 30)
  OK    Acme Corp    activated Day 18 (top of cohort)
  OK    Vertex       activated Day 24 (on pace)
  WARN  Blend Labs   Day 30, zero core actions (behind)
  WARN  Northwind    Day 30, champion dark 14 days

DAY 60 CHECK (4 past Day 60)
  OK    Summit       multi-capability adopted
  WARN  Meridian     Day 60, baseline only, nothing adopted

DAY 90 CHECK (2 past Day 90)
  OK    Anvil        power usage on two capabilities
  WARN  Beacon       Day 90, fit score now healthy but no expansion in pipeline

Handoff quality:
  11 of 14 had a kickoff inside 14 days
  WARN  Blend Labs   no kickoff yet (Day 30)
  WARN  Meridian     kickoff happened but context was empty at handoff
  WARN  Northwind    account notes never updated from sales-stage notes

Early false-negatives (1):
  Beacon   fit score floored, but adopted 4 capabilities in 30 days
           should be in the expansion pipeline now

At-risk in first 90 days (3):
  1. Blend Labs   Day 30, zero adoption, no CSM meeting
  2. Northwind    champion not engaged, thin handoff
  3. Meridian     Day 60, still no capability adopted

Next moves:
  1. Emergency CSM intervention on Blend Labs this week
  2. Northwind, re-engage the champion or pivot to a backup
  3. Add Beacon to the expansion pipeline and assign coverage
  4. Handoff-completeness rate needs to hit 100 percent
── WHERE THE NUMBERS COME FROM

The Day-30 / 60 / 90 checkpoints and the two-week kickoff window are defaults, not laws. They suited a product with a fast time-to-value. If your product takes longer to land, stretch the windows. The milestone logic does not change, the thresholds are yours.

── EXAMPLE CHAT
EXAMPLE CHAT
You
Beacon is 90 days in and our fit score has them flagged low. Worth a save call or should we let it go quiet?
CLAUDE
Pulled the first-90-day read on Beacon. Don't let this one go quiet. DAY 90 CHECK Beacon power usage on two capabilities, fit score still LOW Early false-negative: Beacon fit score floored at signup, adopted 4 capabilities in 30 days should have moved to expansion pipeline back on Day 30 The fit score never caught up to what Beacon was actually doing. That's a 60-day-old false negative sitting in the wrong bucket.
You
How do we stop missing this on the next cohort?
CLAUDE
Two fixes. First, re-run the false-negative check at Day 30, not just Day 90, so you catch it while it's still fresh. Second, add Beacon to the expansion pipeline today with the adoption evidence attached so the AE isn't starting from zero. One sharpener: connect your product-analytics tool so activation dates come from real usage instead of the self-reported milestones you pasted in.
04 / GO FURTHER

Where an operator takes this next.

The read above is a one-time pull. Here's what an operator wires up once the manual version proves the false-negative catch is real.

The diagnostic is the easy part. Wiring it to run without you is what actually saves the account.

01
Run it every Monday, not when you remember

A scheduled Claude task pulls the new-customer list from Salesforce and flags Day-30/60/90 misses before the CSM has to ask.

02
Catch the false negative the day it happens

Connect a product-analytics tool so the moment a low-fit account crosses the adoption threshold, it lands in a Slack DM to the account owner instead of waiting for a quarterly review.

03
Close the loop into expansion

Route confirmed false negatives straight into the expansion pipeline in Salesforce with the adoption evidence attached, so the AE opens the conversation with proof, not a guess.

05 / PART OF A BIGGER PLAY

One skill is the on-ramp.

── GET NEW SKILLS IN YOUR INBOX

One operator-built skill a week. Free.