From a kitchen recipe to a documented, scalable process
How to turn the recipe in your head into a written spec a stranger's crew could run, spot what breaks when you scale up, tell whether your product forces a process-authority filing, sketch a shelf-life test — and name who owns the food-safety plan and the records once it's a process.
Chef Diego runs a real food plant. If this page didn't get you there, tell us — a person reads every message.
By the end of this lesson you can take the recipe that lives in your head — or on
a stained index card — and turn it into a written specification a crew you'll
never meet could run, spot the problems that show up only when you make ten times
as much at once, tell whether your product is the kind the FDA requires an expert
to sign off on, and sketch a test that proves how long it keeps. This is the
thing you actually hand a co-packer, but it matters just as much when you make the
product yourself. And it carries the same thread as the rest of this course: once
the recipe becomes a documented process, someone owns the food-safety plan, the
process, and the records — and this lesson names who.
Formulation versus a manufacturing recipe
Two documents hide inside what you probably just call "the recipe," and telling
them apart is most of the work.
The first is the
.
It's the what, and it's the part you protect: your formulation is the thing
that's genuinely yours.
The second is the
.
It's the how. A card that reads "a pinch of this, cook until it looks right"
holds a formulation but no manufacturing spec — which is exactly why no one but
you can make it.
Turning the first into the second is the real work of product development. The
spec is what makes the product come out the same batch after batch, and it's what
lets you hand the making to someone else — a new hire, a second shift, or a
co-packer's crew. Building it out properly — the bill of materials, the yield, the
units that actually add up — is its own craft, and
recipes, BOMs, and yield
walks through it in production detail.
What breaks when you scale up
Scaling a recipe is not multiplication. You can't take a pot that serves twelve,
multiply every number by fifty, and get the same product out of a kettle. Three
things change when the batch gets bigger, and each one can quietly ruin a run.
The first is
.
At the stove you never counted the loss; at scale it decides your cost per unit,
so you measure it instead of assuming it.
The second is mixing and heat. A big vessel heats slowly, cools slowly, and
stirs unevenly. A sauce that thickened in a home pot in five minutes behaves
differently in a two-hundred-gallon kettle, and a starch that set perfectly on
the stove can break at volume.
The third is the
.
Heat moves through a large batch differently than a small one, so the time and
temperature that worked at home are a starting point, not the answer.
The fix for all three is the same: run a pilot batch — a small production run on
the real equipment, at the real scale. Measure the actual yield, taste it against
your standard, and record the numbers that matter: a cook temperature, a final
pH, a fill weight. Then correct the specification to match what really happened. A
recipe that hits at the stove has told you almost nothing about how it behaves on
the line.
When your product needs a process authority
Most products you can develop, document, and scale on your own judgment. Two
kinds you can't — federal rules require an outside expert and a filing with the
FDA before you sell them. This is the part first-timers most often miss, and the
reason for it is concrete: the hazard these rules exist to prevent is botulism,
from Clostridium botulinum growing in a low-acid or acidified food that wasn't
processed correctly. That's why the FDA doesn't leave the process to guesswork.
Two product classes fall under these rules:
An
.
Many pickles, some hot sauces, and marinated or brined vegetables land here.
They're regulated under
21 CFR Part 114
(verified July 2026).
A
.
Canned vegetables, soups, and many shelf-stable sauces sit here, under
21 CFR Part 113
(verified July 2026).
For either class you need a
.
The process authority sets your
.
Under
21 CFR 114.83,
an acidified scheduled process has to be established by a qualified person with
expert knowledge; the low-acid rules require a process at least equivalent to one
a competent processing authority establishes (verified July 2026).
Then there's the filing. Under
21 CFR Part 108,
a commercial processor of acidified or low-acid canned foods has to register the
establishment with the FDA on Form FDA 2541 — within 10 days of first processing —
and file each scheduled process before packing a new product. The FDA's
registration and process-filing page for these foods
is where that's done and where the current forms live (verified July 2026).
Here's the honest scope. Plenty of products never touch any of this — a baked
good, a dry spice blend, a fresh product kept and sold refrigerated. But whether
your product is acidified or low-acid canned turns on its finished pH, its water
activity, its packaging, and its process together — not on a single reading from
your own pH meter, and never on what a blog says about a product like yours. The
determination is exactly what a process authority is for.
Get the determination before you scale, not after
Don't classify your own product from a home pH reading or a similar jar on a
shelf. The finished equilibrium pH, water activity, packaging, and process
interact, and the rules turn on all of them at once. If there's any chance your
product is acidified or low-acid canned, have a process authority confirm it in
writing — and file with the FDA where it's required — before you commit to a
production run. Confirm the current requirements at the
FDA's acidified and low-acid canned foods page
rather than trusting any summary, including this one.
Shelf-life study basics
Every product has a
.
The date you print should come from evidence, not a guess — and the evidence comes
from a
.
There's no single federal recipe for a shelf-life study, but the shape is
consistent. Store samples the way the product will actually live — ambient,
refrigerated, or frozen. Pull them on a schedule. At each pull, check safety
first, then the quality attributes that tend to fail first for your product:
texture, color, flavor, separation, or rancidity. The date you set is the point
before things cross the line, with a margin to spare.
One piece of regulatory honesty. For most foods, a date on the label isn't
required by federal law at all — the single federal exception is infant formula.
The USDA's Food Safety and Inspection Service is plain about it: except for infant
formula, dates are not an indicator of a product's safety and are not required by
federal law; a "Best if Used By" date speaks to quality, not safety (verified July
2026 at
FSIS, Food Product Dating).
Your retail buyers will expect a date anyway, and a date you can defend needs a
study behind it — which is why you run one even when no rule forces the label.
Safety and quality are two different clocks
For an acidified or low-acid canned food, the scheduled process is what makes
the product safe — that clock belongs to the process authority. The shelf-life
study is about how long quality holds: flavor, texture, appearance. Don't let a
long shelf life reassure you about safety, or a validated process reassure you
about quality. They answer different questions.
How dates, date codes, and first-expiry-first-out actually run on your floor is
its own topic —
shelf life and expiry
covers the operations side.
Who owns the plan once it's a process
Now the thread that runs through this whole course, one last time. A documented
process doesn't, by itself, say who's responsible for it. Someone still owns the
,
any scheduled-process filing, and the traceability records. Name that owner out
loud.
If you make the product yourself, it's all yours: the specification, the
shelf-life evidence, the establishment registration and scheduled-process filing
if your product needs one, the food-safety plan, and the lot records.
If a co-packer makes it, ownership splits — and the split has to be written down.
The co-packer runs its plant's food-safety plan and, for an acidified or low-acid
canned product, generally holds the FDA registration and the scheduled process for
its own facility and line, because that filing is tied to specific equipment. You
still own your product's formulation and specification, and your side of the
traceability records — under the FDA's Food Traceability Rule (FSMA 204), each
company keeps the records for the events it performs, and one party's records never
cover for the other's. Which party files the scheduled process, and who keeps
which record, belongs in the co-packing agreement in plain words — the same
discipline
finding and vetting a co-packer
laid out. The records themselves, and the rule's compliance date, live in
what FSMA 204 asks you to record.
The document makes the product repeatable; the contract makes the ownership clear.
From here
You've turned a recipe into a specification, found the places scale-up will test
it, learned whether your product needs a process authority, and sketched the study
that sets your date. One question is left before the product can go on a shelf: how
do you label it? A label is a legal document — the statement of identity, the net
quantity, the ingredient list, the allergen declaration, and the date you just
learned to justify. That's where this track turns next, in the course on your
label and your liability.