Tie every complaint to the revision and lot that shipped.

Complaints counted against what shipped, by revision and lot, so a problem shows up as a rate with a cause.

Start with a development review How it works

Shrinks after washing · by lot24 complaints · 2,000 shipped

Lot 3 · 21 of 67531Case opened
Lot 2 · 2 of 6753.0Below the line
Lot 1 · 1 of 6501.5Below the line

Per 1,000 shipped. A case opens when a lot crosses the rate your team sets

One symptom, counted per 1,000 shipped, by lot. The lot that stands out opens a case.

Sound familiar?

We know people say it shrinks, and we can’t say which batch or how many.
“Twenty-four complaints” means nothing until you know how many units shipped.
Support sees the complaints. Development never sees the pattern.

Without a denominator and a lot, a fixable batch problem looks like a product problem, and the wrong thing gets redesigned.

What changes.

TodayWith the engine

Counting: Complaints counted as they arriveComplaints per 1,000 shipped, by revision and lot
Symptoms: Tags typed by whoever answeredClassified the same way every time, with the customer’s words kept
The join: The marketing SKU onlyOrder to shipment to revision and lot
Unknowns: Dropped from the countShown as unknown, never guessed
Who sees it: Support, in a weekly exportDevelopment and quality, when a lot crosses the line

How it works. 5 steps, 5 stages.

AI1
Rules and automation3
Your team1
Approval

1Collect
1. Rules and automation
  Pull complaints from every source
  Returns, support tickets and reviews, deduplicated by order.
  - Shopify
  - Gorgias
2Classify
1. AI
  Classify what went wrong
  Symptom and use condition, like “shrinks after washing”, with the customer’s words kept.
3Join
1. Rules and automation
  Join to revision and lot
  Order to shipment to the revision and lot it shipped from. Anything unmatched stays unknown.
  - Shopify
  - NetSuite
4Count
1. Rules and automation
  Count against what shipped
  Complaints per 1,000 shipped, by lot, variant and period.
5Flag
1. Your teamApproval
  Review the cases that cross the line
  Quality opens an investigation. A serious safety signal skips the count and goes straight to your escalation path.

Every shipment adds to the denominator

What you get. Real work, every week.

Shrinks-after-washing mentions

Mentions27
Tied to a lot24

Three mentions had no order. They’re counted as unknown, not guessed

How many complaints could be tied to an order and a lot. The rest are reported as unknown.

Complaints per 1,000 shipped

Counted against	Shipped	Complaints	Per 1,000
All of revision E	2,000	24	12
Lot 3, substitute thread	675	21	31
Lots 1 and 2, approved thread	1,325	3	2.3

The same complaints, cut three ways. The denominator is what makes the lot visible.

How you’ll know it’s working.

Complaints per 1,000 shipped, by lot
The rate that separates a batch problem from a product problem.
Share of complaints tied to a revision and lot
The unmatched share tells you how far to trust the rest.
Classification checked against your quality team
A wrong tag sends a case to the wrong owner.

Baselines agreed before we build.

Works with what you have.

Commerce

Shopify
NetSuite

Support

Gorgias

Fulfillment

ShipBob

We configure first: Your helpdesk and returns tool already tag complaints. We start from those tags and check them, rather than add another system.
We build when: Complaints are tagged but never joined to a lot or revision, so nobody can count them against shipments.

What a first build includes.

Built in your accounts, around your team. Scope, fee and baseline agreed before we start.

1Design
2Build
3Run it, or hand it over

Start with a review Scope this build

We build

A complaint feed from returns, support and reviews
Symptom classification checked by your quality team
A join from order to shipment, revision and lot
A cohort table with rates per 1,000 shipped

We start from

Orders and shipments, with lot numbers from your 3PL or factory
Returns, support and review history
Your quality thresholds

From your team

Owner: Quality lead or product development lead
Involves: Quality
Customer service
Development
Sourcing
Logistics

Questions teams ask.

What if our 3PL doesn’t track lots?

Then the first step is to capture lot numbers at receiving. Until then, complaints stay shown as unknown. We never guess a lot from a date.

Does AI tell us the cause?

No. It classifies and counts. Quality or engineering establishes the cause, and a rate that rises and falls by lot shows where to look.

Who owns it once it’s running?

You do. It’s built in your accounts, documented, with training for the people who run it. Ownership usually sits with your quality lead or product development lead. We can keep running it with you, or step back.

How do we start?

Usually with a development review: two to four weeks, a fixed fee credited to the build, and a plan that’s yours. If the job is already clear, we can scope the build directly.