iHeartRaves / INTO THE AM · CST × Product

Pam Website QC

Pam checks every product page against your source of truth, and flags anything missing, wrong, or out of order, before it goes live.

Deterministic. Fast. No guesswork. · Share-ready overview for the team

TL;DR

Pam compares a finished product page to what you said should be on it and returns a clean pass / fail / fix list in seconds.
The one thing the team has to build is a source of truth: a simple sheet, one row per product, columns for the fields that must be there.
Pam pulls the page data from Shopify (read-only), so the check is structured and reliable, not screen-scraping guesswork.
It is deterministic work: the comparison is rule-based, so it is cheap, consistent, and runs on every launch.
To start we need a filled template for 5 to 10 pilot products. Then Pam QCs every new product before it goes live.

The problem it kills

Manual page QC is slow, and things slip live: wrong price, missing size chart, no fabric content, images in the wrong order, a color or size that never got added. Each one is a return, a support ticket, or a lost sale. Pam catches it in seconds against what you already decided the page should have.

How it works

STEP 1

Give Pam the source of truth

What must be on the page, per product.

STEP 2

Point Pam at the page

The live URL or the Shopify product.

STEP 3

Pam checks field by field

Source vs page, line by line.

STEP 4

Pam returns a QC report

All clear, or a specific fix list.

What you provide (this is the part the team is stuck on)

The unlock is one structured source of truth. Simplest version: a Google Sheet, one row per product, columns equal the required fields. Pam reads it. You only fill the fields you care about. Blank means not checked.

Recommended columns:

Product name / handle / live URL
Price (and compare-at price)
Variants: every color and size that must exist
Description: required sections or key phrases
Images: required count and required shots (front, back, detail, on-model)
Size chart: required? which chart
Fabric / material content
Care instructions
Fit notes (runs small / large, model height and size worn)
Tags / collections it must belong to
Meta title and meta description (SEO)
Badges (new, bestseller, etc.)

What Pam checks

Presence: is each required field on the page at all?
Match: does the value match the source (price, variants, fabric)?
Order: are images and sections in the expected order?
Completeness nudges: "no size chart found", "only 2 images, you usually have 5", "missing fabric content".

Sample QC report

PASS · present and correct FAIL · missing or wrong WARN · mismatch to confirm SUGGEST · nice to add

Basic Tee: Black

PASSPrice $24 matches source.

PASS6 of 6 variants present (S through 3XL).

FAILSize chart missing.

FAILImage 3 (back shot) not found. Only front and detail uploaded.

WARNFabric on page reads "95% cotton / 5% spandex" but source says "100% cotton".

WARNMeta description is 184 characters, over the 160 limit.

SUGGESTAdd an on-model image. This style usually has one.

How it is built (note for whoever builds it)

Pam pulls product data straight from Shopify via a read-only API (structured and reliable). Optional: also scan the live page for display-only items like the size-chart widget and badges.
The comparison is deterministic and rule-based. Pam only uses judgment on fuzzy text matches. That keeps it cheap, fast, and consistent.
Pilot on 5 to 10 products, tune the field list, then scale to every launch.

What we need to start

The source-of-truth template filled for a handful of pilot products. We provide the sheet.
Shopify read-only access for the brand. Already in place.
2 to 3 sample product pages to pilot.

Then Pam QCs every new product before launch, and can re-check live pages on demand.

Phases

PHASE 1

Pilot

Manual trigger, a few products, prove it.

PHASE 2

Every launch

On-demand or scheduled before go-live.

PHASE 3

Full catalog

Catalog audit plus ongoing monitoring of live pages.