Un-Factory Foods™
Stock Our Products Become a Maker
Why It Works Our Foods Articles Resources Stock Our Products Become a Maker
By Matthew Myers Quality

One Recipe, A Hundred Kitchens

How The Un-Factory holds a single quality standard across a distributed maker network.

For a buyer, a distributed food brand raises one fair question before any other: if this product is made in a dozen different kitchens, how do you know each one is the same, and how fast can you act when one isn’t?

This is our answer. It rests on two systems we built. The first controls who is allowed to make our product. The second controls how every unit is made, recorded, and traced. Together they mean an Un-Factory product is not simply made locally — it is made by a qualified maker, to a fixed standard, on a signed and traceable record, every run. Every line we carry — Dare to Pop popcorn, Noble Pretzel, and Slow Rise crisps — runs on both.

Two systems, one standard

Centralized manufacturing controls quality by consolidating production into one building. We give that up on purpose — the product is made locally, close to the shelf it sells on. The trade is real, and we don’t pretend otherwise: made in many places, a product will drift unless something holds it to a single standard.

Two things hold it. Qualification decides who enters the network. Run Ticket governs how they produce once they’re in. A maker cannot reach the second without clearing the first.

Part one: qualification

Not everyone who wants to make our product gets to. Before a maker produces a single retail unit, they pass a qualification process that gates entry to the network.

Qualification is the gate. It means every maker producing under our brand has been vetted as a facility, as an operator, and as a person — before the production system ever hands them a run.

Part two: Run Ticket

Once qualified, a maker produces under Run Ticket — the application that runs the decentralized factory. It moves quality control off a factory floor and into software every maker works from, identically, on every run. One run, start to sealed case, it governs five things.

Formulation is calculated, not improvised

A run begins with an order, not a recipe. Run Ticket takes the order quantity and works backward to every input required — scaled to the exact batch, itemized per mix and per full run. The maker weighs to the number on screen.

Recipes are held centrally and versioned; a maker receives a formula and cannot alter it. Removing manual scaling removes the failures it causes — rounding, mis-converted measurements, batches wrongly doubled. A specification change propagates to every maker’s next run at once, and prior versions stay on record against the batches they made. There is one source of truth for what each product is, and it is not the maker’s memory.

Process is sequential and signed

Each product’s run is a numbered sequence of steps. Each is signed off in-app — initials and system timestamp — before the next unlocks. A step cannot be signed out of order, and a late signature is recorded as late rather than backfilled to look on-time.

This yields a complete, time-stamped production record for every run, generated as a byproduct of the work rather than reconstructed afterward. It is the difference between a supplier who claims a process and one who can show it, run by run.

Each product has a defined critical control point

Every line has one step that determines whether the product meets standard, and the ticket enforces it.

For Slow Rise, it’s a 24-hour ferment — the step that develops the product’s structure and flavor, and the one most vulnerable to being shortcut, since under-fermented dough can look correct on the surface. The ferment is not a single confirmation; it requires four checks logged across the 24 hours, each timestamped and initialed against a defined visual standard, and the step will not close until all four are present. A gap in those checks is recorded and does not clear on its own.

For Dare to Pop, control sits at the cook and coat — temperature, timing, and seasoning coverage. For Noble Pretzel, at the bath and bake. Each product’s ticket places its controls at its own decisive step. The principle is constant: identify where the product is made or ruined, and make that step non-skippable and self-documenting. Checks are timestamped and gaps are retained; nothing is erased.

Every run is verified by photograph

Each run ends with a standardized photographic record. The finished product is photographed on a supplied reference card carrying a dimensional scale and a grey balance patch, so images captured under different lighting are normalized and directly comparable — color, size, and finish measurable against every other run in the network.

What each photo must show is defined per product. A Slow Rise crisp is photographed snapped in half, because the 24-hour ferment is legible in the cross-section and a shortened one is not — a control that can’t be satisfied without performing the step. Popcorn is documented for coat and unpopped count; pretzels for color and crust. Runs are reviewed on a sampled basis, with any flagged run pulled for direct review.

Traceability, labeling, and recall

Every unit is labeled from its run record. The label carries the production date, a batch ID, and the maker’s cottage-law information — printed directly from Run Ticket, not entered by hand, so the compliance line can’t be mistyped and the batch code can’t be transposed.

That batch ID is the spine of the system. From any single bag, it resolves to the full production record: which maker, which kitchen, which day, which ingredient lots, which signed steps, and the finished-product photographs. Ingredient lot numbers are captured at the scaling step, so traceability runs from the retail unit back through the batch to the specific supplier lots it contained.

The consequence is the one that matters to a buyer: in a recall, an affected lot can be identified to the individual batch and maker within the hour, and every retail unit carrying that batch ID is accounted for. We also hold a retained sample from each lot, sealed and labeled, so a reported issue can be assessed against physical product rather than description alone.

Accountability and enforcement

Qualification gets a maker in. Standard keeps them in.

Missed checks, late signatures, and process gaps are recorded against the maker. A maker operating to standard competes on the same shelf as every other maker under our name, and their work is only worth what the least careful maker’s work makes it worth. Enforcement protects that maker — and the buyer — from anyone cutting a corner.

Accumulated failures carry escalating consequences, up to removal of a maker’s right to produce under our brand, following a defined process with severity tiers so a paperwork lapse, a quality deviation, and a food-safety breach are treated as the distinct events they are. A maker who falls below standard is caught by the record and removed by the process. The right to keep producing is contingent on continuing to meet standard.

In short

Centralized manufacturing controls quality by consolidating production. We control it in two moves: we qualify who is allowed to make the product — facility, interview, and background screening — and we govern how they make it through Run Ticket, which generates a signed, timestamped, photographed, lot-traceable record as the work is done.

A buyer stocking an Un-Factory product is stocking a distributed network with vetted makers, factory-grade traceability, and a documented enforcement standard behind every unit on the shelf.

Made Local. Sold Local.

← All articles Become a Maker →