Ops & Supply

How to manage co-packer production runs as CPG Founders

Ops & SupplyFor CPG Founders3 apps12 steps~24 min to set up

You ship ingredients to your co-packer, wait two weeks, and receive a pallet plus a PDF with numbers that may or may not match what you sent in. You're reconciling yield on a spreadsheet you built yourself, chasing your production coordinator over email to find out when your next run is scheduled, and discovering spec sheet version mismatches only when the wrong label shows up on finished goods. If demand spikes and you need to move a run date, you're texting someone who may or may not respond before the slot is gone. There's no shared system — just your inbox, their inbox, and a Google Sheet neither of you fully trusts.

Build this on Starch → See the apps used

Ops & SupplyFor CPG Founders3 apps12 steps~24 min to set up

Outcome

What you'll set up

A shared production calendar that shows your next co-packer run date, status, and ingredients committed — without sending a single email to ask

Yield variance tracking that flags when your co-packer is reporting 8% shrinkage on a SKU that historically runs at 3%, before you've already paid for the overage

Automated purchase order drafts triggered when your demand forecast says a run needs to start within your lead time window

The Starch recipe

Apps, data, and prompts

The combination of Starch apps, the data sources they pull from, and the prompts you use to drive them.

Apps used

Co-Packer Manager Inventory Planner Lot Tracker

Data sources & config

Co-Packer Manager, Inventory Planner, and Lot Tracker are all currently in development — request beta access to be notified at launch. In the meantime, you can wire Starch to your existing data today: connect QuickBooks from Starch's scheduled-sync integration so bills and vendor payments from your co-packer sync automatically; connect Google Sheets from Starch's integration catalog (queried live when your app runs) to pull production logs your co-packer already maintains; and use Starch's browser automation to pull shipment status from your 3PL's portal — no API needed. Shopify connects live from Starch's integration catalog to feed real-time sales velocity into your production planning.

Prompts to copy

Build me a co-packer production tracker that shows all upcoming runs by SKU, with scheduled date, ingredient quantities committed, expected yield, and actual yield once the run closes. Flag any run where actual yield is more than 5% below expected.

Create an inventory dashboard that pulls current stock from my co-packer, my 3PL in Dallas, and my Amazon FBA reserve, shows days-of-supply per SKU based on a 60-day velocity average, and highlights anything under 30 days of supply.

Set up lot-level traceability so every production run gets a lot number, each lot links to the ingredient COAs and supplier lot numbers used, and I can pull a full one-up-one-down chain for any lot in under two minutes.

Run these in Starch → or paste them into your favorite agent

Walkthrough

Step-by-step

1 Connect QuickBooks — Starch syncs your bills, vendor payments, and purchase orders on a schedule so every co-packer invoice is visible in one place alongside your production data.

2 Pull in your co-packer's production log spreadsheet by connecting Google Sheets from Starch's integration catalog; the agent queries it live each time your production dashboard refreshes.

3 Tell Starch: 'Build me a production calendar that shows every scheduled run for the next 90 days, organized by SKU, with run date, ingredient quantities, expected case count, and current status.' Starch builds the app from that description.

4 Enter your standard yield targets per SKU (e.g., Granola Bar 8oz: 94% yield, Protein Powder 2lb: 97% yield) so the system has a baseline to compare against when actuals come in.

5 After each run closes, enter or import actual yield data; Starch flags any variance beyond your threshold and logs it against that co-packer and that SKU for trend analysis.

6 Wire Shopify from Starch's integration catalog so sales velocity is queried live and feeds into your days-of-supply calculation per SKU across all inventory locations.

7 Use Starch's browser automation to pull current FBA reserve and inbound quantities from Seller Central — no Amazon API setup required — and add that to your inventory picture.

8 Tell Starch: 'When any SKU drops below 45 days of supply and my co-packer lead time is 21 days, draft a purchase order for the next production run and send it to me for review before it goes out.' Starch builds that automation.

9 Assign a lot number to every production run and attach the relevant COAs and supplier lot numbers so you have one-up-one-down traceability for every batch — critical for FSMA 204 compliance and retail audits.

10 Run a mock recall drill quarterly: tell Starch 'Show me every order that contains product from lot 2025-1142 and the customer or retailer it shipped to.' The answer should take seconds, not a day of spreadsheet archaeology.

11 Set a weekly automation: 'Every Monday at 7am, send me a Slack summary of any production runs in the next 30 days, SKUs below 45 days of supply, and any yield variances flagged in the past week.' Starch handles the scheduling and the message.

12 As Co-Packer Manager, Inventory Planner, and Lot Tracker move from beta to live, migrate your custom-built apps into those pre-built templates or fork them to match your specific SKU structure and co-packer relationship.

See this running on Starch

Connect your tools, describe what you want, and the agent builds it. Closed beta is free.

Try it on Starch →

Worked example

Q1 2026 Granola Bar Production Run Reconciliation

Sample numbers from a real run

Ingredients committed to co-packer (oats, honey, almonds)	18,400
Co-packer labor and overhead invoice (QuickBooks bill)	9,200
Expected finished goods at 94% yield — 4,700 cases @ $5.85 COGS	27,495
Actual finished goods received — 4,330 cases (92.1% yield)	25,331
Yield variance flagged — 370 cases short vs. expected	-2,164

You shipped $18,400 in ingredients for your 8oz Granola Bar to your Ohio co-packer on February 3rd with a 94% yield target and an expected finished case count of 4,700. The QuickBooks bill for $9,200 in co-packer fees synced into Starch automatically the day it was issued. When your co-packer uploaded their run closure report to the shared Google Sheet on February 17th, Starch queried it live and immediately calculated actual yield: 4,330 cases, 92.1%, 370 cases short. That triggered a variance flag — your threshold was 93% — and Starch sent you a Slack message that morning with the delta: $2,164 in expected inventory you didn't receive. You opened the production run record in Starch, attached the co-packer's explanation (almond moisture content was high, causing more breakage in the bar former), and logged it against the run. You also had the lot number — 2026-0203-GB8 — linked to the three almond supplier lot numbers that went into the batch, so if there's a quality issue downstream, you can trace it back in under two minutes. Without Starch, you would have found this variance three weeks later when you reconciled the QuickBooks bill against a manual case count — after the product was already allocated to a Whole Foods purchase order.

Measurement

How you'll know it's working

Yield variance per SKU per run (actual vs. standard yield %)

Days of supply per SKU across co-packer, 3PL, and FBA locations

Co-packer on-time production rate (scheduled run date vs. actual completion date)

Time to complete a mock recall trace (target: under 5 minutes for FSMA 204 readiness)

Purchase order lead time adherence (days between PO issue and ingredients-at-co-packer)

Comparison

What this replaces

The other ways teams handle this today, and how the Starch version compares.

Spreadsheets (Google Sheets + Airtable hybrid)

Most CPG founders at this stage live here — it works until you have 8+ SKUs, multiple co-packers, or your first retail audit, at which point the manual reconciliation time and error rate become the actual bottleneck.

Fishbowl or inFlow Inventory

Solid inventory management tools, but they don't talk to your co-packer's production data, don't handle lot traceability the way FSMA 204 requires, and require significant setup time your team doesn't have.

TraceGains or SafetyChain

Purpose-built for food safety and supplier compliance — genuinely strong on COA management and audit trails — but priced for mid-market manufacturers with a quality team, not a founder doing this alongside everything else.

Cin7 or Katana

Good manufacturing-focused inventory tools with production order tracking, but built for in-house manufacturing workflows rather than the co-packer relationship model where you don't control the production floor.

On Starch RECOMMENDED

One platform — copacker manager, inventory planner, lot tracker all running on connected data. Setup in plain English; numbers stay current via scheduled syncs and live agent queries.

Try it on Starch →

FAQ

Frequently asked questions

Co-Packer Manager, Inventory Planner, and Lot Tracker are all listed as 'in development.' What can I actually use today?

You can build a working version of this workflow right now using Starch's custom app authoring. Connect QuickBooks (Starch syncs it on a schedule), Google Sheets from Starch's integration catalog (queried live), Shopify from Starch's integration catalog, and browser automation for any portal your co-packer or 3PL runs. Then describe what you want — a production tracker, an inventory dashboard, a lot traceability log — and Starch builds the app. The dedicated Co-Packer Manager, Inventory Planner, and Lot Tracker apps are pre-built templates that will make setup faster when they launch; the underlying capability is available today.

My co-packer doesn't have an API. Can Starch still pull their data?

Yes. If your co-packer has a web portal you log into, Starch can automate it through your browser — no API needed. If they send you production reports as spreadsheet attachments, you can connect Google Sheets from Starch's integration catalog and query that data live. The only scenario where you'd need to do manual data entry is if your co-packer communicates exclusively by phone or PDF with no digital record — and even then, Starch can be the system of record where you log those numbers.

Does Starch store my lot traceability data, or is it just querying it live?

Data from Starch's scheduled-sync integrations — like QuickBooks, Stripe, and Plaid — is stored in Starch's database and refreshed on a schedule. Data from Starch's integration catalog (like Google Sheets or Shopify) is queried live when your app runs rather than stored long-term in Starch. For lot traceability specifically, you'd want your core lot records — lot numbers, COA links, chain-of-custody events — living in a Starch app that stores them directly, not purely in a live-queried external sheet. That's exactly what the Lot Tracker app is designed to do.

Can Starch generate a purchase order and actually send it to my co-packer?

Starch can draft a purchase order automatically based on your reorder triggers and send it to you for review. For sending to your co-packer, it can push the PO via Gmail (Starch connects directly to Gmail) or attach it to an email that goes out automatically once you approve it. If your co-packer uses a portal for PO submission, Starch can automate the submission through your browser — no API required.

Is Starch SOC 2 certified? My retailer is asking about data security.

Starch is not currently SOC 2 Type II certified. If your retailer or co-packer requires SOC 2 documentation as a condition of sharing data, that's worth knowing upfront. For most co-packer workflows — production scheduling, yield tracking, PO management — the data involved is your own operational data rather than consumer PII, so the compliance conversation is usually more straightforward than it sounds.

We work with two co-packers — one for dry goods and one for liquid products. Can Starch track both in the same place?

Yes. You can build a single production tracker that segments runs by co-packer and by product line, with separate yield targets and lead times for each relationship. Tell Starch: 'Build me a production calendar that tracks runs across Dry Goods Co-Packer and Liquid Products Co-Packer separately, with SKU-level yield targets for each, and flags any variance above 3% within 24 hours of a run closing.' Starch builds that app from the description.