Recall & Field Action Execution Tracker: Run It, Then Prove It Reached Everyone

The problem this kills

A recall or field corrective action is the one moment where your quality system is on trial. The clock is running, a regulator may be watching, and you have to do two hard things at once: actually reach every affected customer, and later prove — with evidence — that you did.

Most teams run it out of a spreadsheet and an inbox. The affected list lives in one tab, the "who did we call" notes in another, returns get logged on a third, and the notification wording sits in someone's sent folder. When the auditor asks "what was your effectiveness percentage, and show me the notices and responses," you spend a frantic week reconstructing a story from fragments. Worse, in the rush a customer gets contacted twice, or — far more dangerous — gets missed entirely.

The math regulators care about is simple but unforgiving: of everyone who should have been reached, what percentage actually responded, returned, or corrected? You need that number to be high, you need to escalate the stragglers who are dragging it down, and you need every notice and every response saved as evidence. You don't need to be a developer to run a recall this cleanly.

What you'll build

A private web tool, just for your team, that runs a recall end to end:

• Import the affected list — every affected customer or unit, from your traceability tool, an ERP export, or a CSV / Google Sheet.
• Lead approves the notice and scope — the recall lead reviews the exact notification wording and confirms which units are in scope before a single email goes out.
• Send notifications and track them — Resend emails the approved notice, and the tool tracks each customer's status: contacted, responded, returned, corrected.
• Compute effectiveness % — the regulator metric, live, as responses come in, with a target threshold you set.
• Escalate non-responders — automatic reminders and a clean list of who still hasn't responded so you can phone or certified-mail them.
• Approve closure and export the record — the lead can only close the recall once the effectiveness threshold and required steps are met, and the tool exports the full recall record — notices, responses, dates, and the final percentage — as evidence.

What's inside the Implementation Plan

The plan is a step-by-step runbook you paste into an AI coding agent (like Claude Code). It builds the whole tool with you, one prompt at a time — no coding experience needed.

It starts by interviewing you about your business. Before any code is written, the plan has the agent ask about your actual recall process: where your affected list comes from and what its columns are called, how you identify a unit (lot, serial, batch, order number), your typical and worst-case recall size, what counts as an effective response in your world (returned? corrected in the field? acknowledged?), your effectiveness threshold, your escalation steps, and the messy exceptions — distributors who hold stock for end users, customers with no email, units already scrapped. Then it reads a short tailored spec back to you and waits for your thumbs-up. The tool you end up with fits your recall, not a generic template.

From there the plan walks through: the data model, login, the import, the recall-lead approval gate for notice and scope, the notification send, response and return tracking, the effectiveness calculation, non-responder escalation, the closure gate, and the final export — each step ending in a ready-to-paste prompt.

The governance it includes (this is the point)

This isn't a glorified spreadsheet. The plan bakes in the controls a regulator and your own quality system expect:

• Login so only your recall team can touch it.
• Row-level security so you only ever see your own organization's recalls and customers.
• A complete audit trail — who imported, who approved the notice, who approved closure, and exactly when.
• A hard human-in-the-loop approval gate — the recall lead must approve the notice content and the recall scope before notifications send, and can only approve closure once the effectiveness threshold and required steps are met. The tool drafts and tracks; a person decides.
• Duplicate guards — the dedupe key is recall-id + customer/unit, so the same customer can't be contacted or counted twice.
• Evidence kept — every notice sent and every response received is stored, so the recall record you export can stand up to an audit.

Who it's for

Quality, regulatory, and operations teams managing a product recall or field safety / corrective action — in medical devices, food, automotive, consumer products, or any regulated industry where you must reach affected customers and prove your reach.

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