Hero Image

Case Study

Behind a changing warehouse, outdated numbers told the wrong story.

KAWAII SLIME COMPANY × AUTOBIZZ

Every internal topping bin variant at Kawaii Slime Company carried the same par level of 4,000 units and an alert threshold of 2,000, whether it moved fast or barely moved at all. AutoBizz replaced those fixed figures with weekly measurements built from real consumption, calculated separately for two of their stores and written straight back into Shopify.

“The right stock, at the right time, without the guesswork”

Client

Kawaii Slime Company

Sector

E-commerce

Solution

Consumption-Based Par Levels

Timeline

August 2026

Par levels at KSC now follow what the warehouse actually uses

AutoBizz helped KSC keep the right amount of each topping variant in stock without relying on guesswork. We built an automated system that checks the last 30 days of actual sales and stock movements to work out how much each store needs and when it should be restocked. Every week, the system updates these numbers automatically in Shopify, giving each store a more accurate stock plan based on real demand. No manual calculations, no outdated stock targets, just a smarter way to manage inventory and reduce the risk of running out of popular toppings or keeping more stock than needed.

  • ShopifyQL
  • Shopify
  • Google Apps Script
  • Shopify Metafields
  • Inventory Adjustment History
  • MS Excel Online
  • Slack

The Situation

A fixed number was applying everywhere

Every internal topping bin variant carried a Par Level of 4,000 units and an Alert Threshold of 2,000. These items were never sold in the ordinary sense. They moved only as transfers from the warehouse to two channels, each at its own pace.

Services We Provided

  • Inventory Automation
  • Automated Reporting
  • KPI Analysis

Old Routine

  1. 01Toppings transferred from the warehouse to stores
  2. 02Transfers logged as inventory adjustments, not as sales
  3. 03Shopify's sales-based tools saw no demand at all
  4. 04The same 4,000 par level stayed on every variant
  5. 05The 30-Day Demand Plan read those fixed figures as signal

Nobody had set those numbers wrong on purpose. They were set once, as a reasonable starting point, and then quietly treated as truth because nothing ever measured them again.

‘A par level that never moves is not a plan. It is a guess that everyone has quietly agreed to stop questioning’

Image of Client's Previous Situation

The Challenge

Transfers that never touched checkout stayed invisible to Shopify

Because toppings moved by transfer, not by sale, every tool built on sales data looked straight past them.

Image of the challenge client was facing
  1. Day ZeroOne par level of 4,000 and threshold of 2,000 set for every topping
  2. Weeks laterSlow movers sat overstocked against a par level they would never need
  3. Later stillFast movers outran an alert threshold that could not see them coming
  4. Too lateThe 30-Day Demand Plan inherited the stale figures as real signal
  5. Trust spentPlanning numbers that nobody could explain, and nobody could safely rely on

System Architecture

The System Design Board

System Architecture

SEE THE DELAY BEFORE IT BECOMES A CUSTOMER PROBLEM

The Turning Point

The data was there all along

The answer was sitting inside Shopify. Every product transfers left a record in the inventory adjustment history, even though it never reached a sales report. Reading consumption from those adjustments, instead of from checkout, turned an invisible flow of stock into a weekly measurement.

An overview of client's turning point after the solution
Consumption report: 30 days of store adjustments, per product

The Solution

Par levels built from real consumption, every week

Both figures were rebuilt from Shopify's own inventory adjustment history, with separate formulas and the results written straight back into Shopify every week.

01

The Hidden Ledger

Consumption From Adjustments

A tested ShopifyQL query reads 30 days of inventory adjustment history for internal products, capturing corrections, manual adjustments and other movements that never appear as sales.

02

The Fixed Roster

Scoped Product List

The automation runs on a defined list of Internal Use Only, matching each variant against the query output so nothing outside scope is touched.

03

The Twin Engines

Separate Channel Formulas

Different Stores move at different speeds, so each channel gets its own par level formula instead of one shared number.

04

The Early Warning

Calculated Alert Threshold

The alert threshold is recalculated from the same consumption data, so reorder warnings rise and fall with actual use.

05

The Clean Count

500-Unit Rounding

Topping variants comes in 500 units each bag, so par levels are rounded to match calculation. That way, every target is a whole number of bags the team can actually send.

06

The Direct Line

Shopify Metafield Write-Back

Both values are written straight into Shopify metafields, where the 30-Day Demand Plan picks them up without any copy and paste.

07

The Weekly Pulse

Scheduled Recalculation

Stores are scheduled at different time of a week, with a consumption report shared to the team after each run.

What Changed

What changed under the hood

Before

Par level fixed at 4,000 for every variant

Now

Par level recalculated weekly for each variant

Before

Alert threshold fixed at 2,000 everywhere

Now

Alert threshold recalculated weekly from real consumption

Before

No data source, just a static guess

Now

30 days of inventory adjustment history per product

Before

Same number for each of the stores are set once

Now

Formulas matched to each channel's restock schedule, updated weekly

The business outcomes after using our solution

The Results

What the numbers look like now

2 Hrs

Time saved per week

No longer spent reviewing and updating par levels manually.

39%

Less overstock

Reduction in excess units held on slow-moving toppings.

Weekly

Refresh cycle

All reports are scheduled on weekly basis and fully automatic.

References

Documentation this build relies on

  1. 01

    ShopifyQL inventory adjustment history

    Schema used to read event-level inventory adjustments as consumption.

    View Source
  2. 02

    Shopify Custom Data (Metafields)

    How custom fields like par levels attach to Shopify products.

    View Source
  3. 03

    Shopify Admin GraphQL: metafieldsSet

    Mutation used to write par levels and thresholds back to Shopify.

    View Source
  4. 04

    Google Apps Script Installable Triggers

    Time-driven triggers that schedule the weekly report runs.

    View Source

Get Started

Ready to automate your operations?

If your team is still spending hours updating spreadsheets, chasing suppliers or manually monitoring operational data, AutoBizz can help turn that process into a connected automated system.

FAQ

Frequently asked questions

Still have questions? Book a call and we'll answer anything.

What is a consumption-based par level?

A consumption-based par level is a target stock quantity calculated from how much of a product is actually used over a recent period, rather than a fixed number set once. For Kawaii Slime Company, AutoBizz calculates par levels for internal topping bin variants from 30 days of Shopify inventory adjustment history, runs separate formulas for each store channels, and refreshes the figures every week so they follow real usage.