Hire an AI new customer onboarding reviewer

Everything it can hand back

Your task sets the actual outputs. The specialist picks the methods that cover them.

• A chart of each sign-up week's share at each step
• The step that loses the most, and whether it repeated
• A verdict on last week's fix, with the numbers
• The themes behind the stall, in customers' words
• One fix, a first-use flow or a 14-day sequence, with the one rate it should move
• A list of messages and changes to approve

How it does the job well

• Check the current policy and the customer's order or account facts before replying.
• Answer the actual question in the first two lines, then explain.
• Match the customer's tone. Acknowledge the problem once, without grovelling.
• Hand refunds outside policy, legal threats and safety issues to a person.

Mistakes it avoids

• Fixing a step that lost the most for one week.
• Changing two things in the same week, then not knowing which one worked.
• Averaging all customers together and losing the before-and-after.
• Writing a longer email sequence when the stall is in the product.
• Promising refunds, dates or fixes the policy does not allow.
• Copy-paste replies that ignore what the customer said.

What a good result looks like

An example, not your task: You sell a product people must set up or use for the first time before they see the point.

• This week in one line. Sign-ups, the share reaching the first win, and the step that loses the most.
• The chart. Each sign-up week's share at each step, with the newest week at the bottom.
• Last fix. The step, the rate it was meant to move, and a verdict of helped, no change or worse.
• What customers said. Themes behind the losing step, with quotes from new customers.
• The one fix. The cause, the skill run, what changes, the one rate to move and the check date.
• Changes to approve. Each email, text or flow change in one line, with where it goes.

How the work is checked

• Every rate is read from the group rows, not averaged across all customers.
• The last fix was judged before a new one was proposed.
• Only one fix was proposed, aimed at one step and one rate.
• Every email or text uses only claims from the brand brain or product facts.
• States which checks passed, failed or could not run, and names missing inputs.

These checks have not been run against a model yet. Your AI uses them to check its own work.

A second opinion

Before it hands back, your AI reviews the draft the way Eric Ries would. This is a way of reviewing the work. It does not mean they wrote or endorse it.

Capabilities

• Build the plan once. Turns the product and its customers into onboarding steps, milestones and owners, saved after you approve them.
• Read the groups. Adds this week's row, finds the step that lost the most and judges any live fix against the groups before it.
• Hear the customers. Pulls the themes behind the losing step from new-customer messages, with quotes.
• Decide. Waits for a step to lose the most two weeks running. Keeps a fix that helped. Calls a fix that did nothing by name.
• Fix the one step. Finds the cause, then redesigns the first-use flow or rewrites the 14-day sequence, depending on where the stall is.
• Log and hand back. Writes the fix with its rate and check date, then stops for your yes.

How the role works

1. Checks your task, context and available tools.
2. Picks the methods needed and prepares a work plan.
3. Carries out supported tasks, with each method's limits and review gates.
4. Checks the real deliverables and corrects supported faults.
5. Hands back files, evidence and unresolved gaps.

Your specialist's working setup

Download the base specialist without a task form or account. README.md and SETUP.md explain how to use the folder. SOUL.md defines its scope. INTAKE.md asks missing questions inside your AI. PLAYBOOK.md, SKILLS.md, BEST_PRACTICES.md, CHECKS.md and HANDOFF.md guide the work and its checked deliverables.

Complete associated skills, supporting files and licences are included. Choose only the methods needed for the actual task. Keep context, working files and accepted feedback in your AI project.

Use this specialist through MCP or the API

Call prepare_specialist with role_id customer-onboarding-keeper only. No task is required to load it. Your AI asks what job you need done, what you want back and which relevant context it can use.

For task-led hiring, call find_specialist, explain the suggested specialists and let the user accept or change the team.

Complete ZIP: Download base specialist. API: GET /api/v1/specialists/customer-onboarding-keeper/pack. The host AI supplies its own model, tools and execution; a pack does not connect accounts or run a task.

Inputs and helpful context

The specialist starts from the task brief. Only active methods require their inputs.

For the primary method, when used

• Each sign-up week's count at every onboarding step, from your product, store or email tool
• What the product does and what the first win is (advised)
• Support messages, survey answers or cancel notes from the first 30 days (optional)
• Your brand brain, for any email or text (optional)
• Last week's plan and chart, once they exist

Helpful context (optional)

• Day-7 replies or survey answers from new customers
• The date each onboarding change went live, so the verdicts are honest

Tools and access

• Claude Code: It can keep the plan, chart and fixes files between weeks. You start the run each Monday. It does not run on a timer.
• Your email or product tool: Export counts per sign-up week at each step. Most tools can group by sign-up date.
• Any good chat AI: Reading a table and writing the verdict is text work.
• A spreadsheet or analytics tool: Builds the group table from an export.
• Klaviyo: Paid subscription. It can send on the delivered event (step 10), branch subscribers from one-time buyers and run email and SMS from one flow. The delivered-event trigger may need a paid plan.
• Any good chat AI: Any current chat AI can write the copy once you give it the facts.
• A subscription app such as Skio or Recharge: Paid subscription. It tells you who is a subscriber and when the next order ships (steps 4 and 8).
• Any good chat AI: A chat AI can read a CSV and group comments into themes.
• Google Sheets or Excel: Use them to clean and export your survey results first.

Functions and deliverables

• New customer onboarding reviewer: Track where new customers get stuck. Review the last fix and propose one change for the weakest step.
  • A chart of each sign-up week's share at each step
  • The step that loses the most, and whether it repeated
  • A verdict on last week's fix, with the numbers
  • The themes behind the stall, in customers' words
  • One fix, a first-use flow or a 14-day sequence, with the one rate it should move
  • A list of messages and changes to approve
• Customer onboarding planner: Describe your product and customers. Get phased onboarding steps, email templates, milestones and owners.
  • A phased onboarding plan
  • Task checklists with owners
  • Email and message templates
  • Milestones and escalation rules
  Source: Burhan Sebin.
• Cohort funnel analyst: Read your results by join week or month. See which changes moved which rates, and whether the report can be trusted.
  • A table of each group's path
  • Which changes moved which rates, and which are unproven
  • Signs that you are stuck
  • A pass or fail on whether the report can be trusted
• User retention reviewer: Describe where users drop off. Get a diagnosis, ranked fixes and a score out of 10.
  • A score out of 10 with the reasons
  • The main cause of drop-off
  • Fixes for onboarding and prompts
  • Small first steps to try
  • A list of changes in priority order
  Source: Wondel.ai.
• First-use onboarding designer: Design a first-use onboarding flow around one early win. Give it your product and users. You get fewer setup steps and a habit cue.
  • Onboarding flow with a reason for each step
  • First-win definition
  • Habit cue and return trigger
  • Different paths by user readiness
  • List of setup steps to cut
  Source: sickn33.
• New customer onboarding sequence writer: Write a new customer onboarding email sequence and SMS plan for days 0 to 14, plus a 100-day touch map, from your product facts.
  • A day-by-day send plan for days 0 to 14 (email and SMS) and a touch map out to day 100
  • Full copy for about six emails and four texts
  • A pre-arrival sequence and a welcome insert
  • A review and referral timing rule for each product, and a survey to find the habit linked to staying
• Feedback theme finder: Share customer ratings and comments to get score summaries, counted themes, quotes, and recommended actions.
  • A scorecard with NPS and CSAT and how many responses sit behind each
  • Themes with counts and exact quotes
  • Detractor and promoter comments looked at separately
  • Trends against the earlier period, if you gave one
  • Actions tied to specific themes
  Source: Burhan Sebin.
• Brand profile builder: Share your website. Get a reusable profile covering your products, customers, voice and approved claims, with sources.
  • A folder of eight plain files, from a start file for your AI to a claims list and a change log
  • A tag on every fact, and every claim marked approved, pending or banned
  • A two-sentence diagnosis of your biggest gap, and a list of what's missing and who can answer it
  • A short slice for your next step, if you name one
• Eric Ries: Turn a product plan into testable guesses using Eric Ries’s public approach. Start with the riskiest assumption and measure what customers actually do.
  • Your problem worked on the way Eric Ries would: their questions, their method, their push-back
  • The answer or the work, in Eric Ries's voice, using their ideas by name
  • Anything that goes beyond their public record, marked
  • One suggested next step

Tasks you can use it for

• You sell a product people must set up or use for the first time before they see the point.
• Cancels cluster in the first month and you fix the welcome email every time.
• You made an onboarding change and cannot tell whether it worked.

Limits to check before starting

• You sell one-off purchases, or your product has no first win: look at your sales and reviews by hand, or ask a person to call a few buyers.
• Your customers are already past 30 days and you want to know who may cancel: use the Customer renewal risk reviewer.
• You have no step counts by sign-up week: start with a simple export from your product, store or email tool, then come back.

Best way to use this role

Give it one clear outcome, the actual files and a definition of done. Keep the work in one AI project so it can reuse checked context and earlier files.

• Start a chat with the agent. In Claude Code, add it as a skill or save it in .claude/agents, then ask it to run this week.
• Export each sign-up week's counts at each onboarding step. Put the file in one folder, or attach it in the chat.
• On the first run, describe the product and the first win. It builds the plan and saves it after you approve it.
• Read the fix. Make the change in your tools, then tell the agent the date it went live.
• Run it again next Monday. It judges the fix against the groups that came after it.

Read it as data

Read the complete role setup as JSON. Connect your AI to the AI Blocks MCP library and read aiblocks://roles/customer-onboarding-keeper for its role instructions and pinned sources.

This is a prepared AI role. No task has run. Your AI carries out authorized work using its available tools and reports the actual deliverables and checks.