Review Request Campaign Tracker

The problem this kills

You know reviews drive sales, so you run review-generation campaigns. But the work is a swamp. You export a customer list, eyeball who looks happy, hand-pick who to email, and try to remember who you already asked last month so you don't annoy them again. You paste people into a marketing tool, hope nobody unhappy slips through, and then have no real record of who was invited, who got a reminder, and who actually left a review. By the third campaign you're guessing, double-asking loyal customers, and accidentally emailing someone who gave you a 2-star CSAT last week.

The painful part isn't sending email — it's the judgment and the bookkeeping around it: who is eligible, who did we already ask, who is unhappy or opted out, and did it actually work? Spreadsheets can't enforce a frequency cap or a satisfaction threshold, and they definitely can't stop you from asking the same person for the same Google review twice.

What you'll build

A small, private web app for your marketing team that turns review-generation into a safe, repeatable workflow:

1. Import your customers from a CSV or Google Sheet (or your CRM later) — name, email, recent purchase, satisfaction score, and prior-request history.
2. Auto-select candidates by your rules: only satisfied customers (above your CSAT/NPS threshold), only those past your "cool-down" window, never anyone opted out or already asked for that platform.
3. Show the marketer a clean review screen with the proposed recipient list — who's in, who's excluded, and exactly why each person was excluded.
4. Require a human approval before a single email goes out.
5. Send the request and timed reminders through Resend, matched to the right review platform and template.
6. Track every response — invited → reminded → reviewed — and respect frequency caps and opt-outs forever.
7. Export a campaign CSV and a results CSV in the exact columns your reporting needs.

What's inside the Implementation Plan

The plan is a complete, paste-and-go runbook for an AI coding agent (Claude Code). You don't write code — you paste, answer questions, and approve.

It opens by interviewing you about your business. Before it builds anything, the agent asks how you actually run review campaigns today: which platforms you collect on, how you measure "happy," your cool-down window, your real customer field names, your peak volumes, and your messy edge cases (refunds, repeat buyers, multiple platforms). Then it reads back a short tailored spec, gets your thumbs-up, and shapes the data model, the eligibility rules, and every later step around your answers — so you get a tool fit to your business, not a generic template.

Inside you'll find: the discovery interview, a tailored data model, the candidate-selection engine, the marketer approval gate, the Resend sending + reminder logic, response tracking, and the CSV import/export fallback so you can build the whole thing today even with no CRM integration.

The governance it includes (this is the point)

Reputation work touches real customers, so the plan bakes in guardrails you'd normally have to fight for:

• Login so only your team can open the tool.
• Row-level security so each organization only ever sees its own customers and campaigns.
• A complete audit trail — who selected the list, who approved it, who was emailed, and when.
• A hard human-in-the-loop approval gate — the tool drafts the recipient list; a marketer reviews and approves before any request is sent.
• Duplicate guards keyed on email + platform, so the same customer is never asked twice for the same review, plus frequency caps and permanent opt-out handling.

Who it's for

Marketing and customer-marketing ops people who run review-generation programs and are tired of hand-picking lists in spreadsheets, worrying about over-asking, and having no clean record of what worked. If you can use a spreadsheet and write a clear sentence, you can build this.

You've got this — paste the first prompt and let the agent interview you.