Approval SLA & Bottleneck Tracker

The problem this kills

You know approvals are slow. You just can't prove where. Was it the legal review, the budget sign-off, or the one director who sits on things for a week? Without numbers, every process-improvement meeting turns into finger-pointing and guesswork, and you end up "fixing" a step that was never the problem.

Most teams don't have approval-cycle analytics because the data lives in exports nobody wants to wrangle: a CSV of items, steps, approvers, and timestamps. Turning that into "the budget gate averages 4.2 business days and breaches SLA 38% of the time" is exactly the kind of repetitive, fiddly analysis that's perfect to hand to software - software you can now build yourself.

This Implementation Plan hands you a tool that does the measuring for you, so you can fix the actual slow gate instead of guessing.

What you'll build

A small, private web app - just for your team - that:

• Imports your approval history from a spreadsheet or CSV (item, step, approver, submitted, decided).
• Computes turnaround per step in business hours, not raw calendar time, so weekends and nights don't make every gate look slow.
• Measures SLA breaches against the targets you set per step or per approval type.
• Ranks the bottlenecks - the slowest steps, the slowest approvers, and the worst-case (not just average) waits.
• Lets an owner exclude outliers (that one item stuck for 90 days while a vendor was on leave) - with a required note explaining why.
• Holds a human approval gate: the bottleneck report is a draft until an owner reviews the findings and the exclusions and signs off. Only then can it be shared or exported.
• Exports a clean CSV in the columns you choose, so the numbers drop straight into your own slides or system.

What's inside the Implementation Plan

A step-by-step runbook you paste into Claude Code (an AI coding assistant). It builds the whole tool with you, one prompt at a time - no prior coding needed.

It opens by interviewing you about your business - your approval steps, your systems, your data's real column names, your SLA targets, your typical and peak volumes, and your messy edge cases. It then reflects a short tailored spec back to you for a thumbs-up before any code is written. That's the difference between a generic template and a tool that fits how your approvals actually work - whether you're a PMO running stage gates or an ops team chasing PO sign-offs.

Inside you'll find:

• The discovery interview (8-12 tailored questions) and how the agent turns your answers into the data model.
• The full build, broken into copy-paste prompts: login, secure database, importer, the business-hours turnaround engine, the bottleneck report, the owner approval gate, and the CSV export.
• A "No API yet?" fallback so you can build the entire tool today using a Google Sheet or CSV as the data source - no integration required.
• A verification checklist so you know it actually works.

The governance it includes (this is the point)

This isn't a throwaway script - it's built like an internal tool you can stand behind:

• Login so only your team can open it.
• Row-level security so each organization only ever sees its own data.
• A complete audit trail - who imported what, who excluded which outlier and why, and who approved the report, all time-stamped.
• A human-in-the-loop approval gate - the report stays a draft until an owner reviews the findings and exclusions and approves. Nothing gets published or fed into a process change automatically.
• Duplicate guards - the same (item + step) row can't be counted twice, even if you re-import the export.

Who it's for

PMO leads, process owners, and ops directors who own a workflow with multiple approval steps and need to prove where the time actually goes - so the next process change targets the real bottleneck. If you can use a spreadsheet, you can build this.

You've got this - paste the first prompt and let the plan interview you.