The Wrong Parts Don't Become Candy

In 1956, a General Foods chemist named William A. Mitchell was deep in a problem that, on paper, sounds almost embarrassingly simple: carbonated soda, without the can. His approach was methodical. Pressurize a sugar mixture with carbon dioxide. Cool it. Let it harden. Drop the tablet in water and — fizzy drink on demand. Clean. Elegant. Commercially obvious.

It didn't work.

What came out of the process wasn't a dissolvable soda tablet. It was something that had never existed before: sugar crystals with CO2 trapped inside them at high pressure. Tiny, crackling, effervescent little grenades that detonated on contact with saliva.

Mitchell didn't throw them out. He tasted them.

Twenty years later, Pop Rocks became one of the fastest-selling new products in candy history. The mistake Mitchell made in that General Foods lab — documented, reproducible, and fully understood because his process was controlled enough to understand it — became a product category that still exists seventy years on. It is one of the greatest accidental discoveries in American manufacturing history.

It is also completely irrelevant to your work. And that's exactly the point.

Let's talk about what a mistake looks like in your world.

Not a candy lab in 1956. Your world — where the thing you're building might navigate a battlefield, regulate a heartbeat, stabilize a satellite, or land a 50,000-pound aircraft without a pilot.

In that world, a wrong component isn't an interesting anomaly. It's a nonconformance report, a root cause analysis, and a conversation with your customer that nobody wants to have. A supplier substitution that slipped through without engineering review isn't a curiosity — it's a potential AS9100 finding, a line-down event, or a field failure with consequences that land on the front page of trade press and in front of program managers who don't forget.

A miscounted kit doesn't slow down a candy launch. It stops a production run. It grounds a test campaign. It pushes a program milestone that took eighteen months to negotiate.

The margin for serendipity in aerospace is zero. In medtech it's negative. In defense it doesn't exist as a concept.

Mitchell had the luxury of tasting his mistake and asking what is this? You have the luxury of getting it right the first time, or explaining to a customer why you didn't.

Here's what the Pop Rocks story is really about — and why it still matters even if you'll never accidentally invent a new product category. Mitchell's mistake was legible.

That's the word that matters. Legible. He knew what he put in. He knew what the environment was. He knew what the process was supposed to produce. So when it produced something else, he didn't just have a strange result — he had a strange result he could actually understand, trace, and reproduce.

That's not luck. That's rigor.

The discovery wasn't serendipitous because Mitchell was lucky. It was serendipitous because his process was documented tightly enough that "unexpected output" was a data point, not a mystery. Most manufacturing environments aren't built that way. Most procurement processes aren't either.

Think about the last time something went sideways in your supply chain. Wrong MPN on a line item. A part that arrived looking right but came from an unauthorized distributor. A kit that went to the line missing three passives that were supposed to be in a separate bag that someone forgot to consolidate. A lead time that changed after PO issuance and nobody caught it until the build was already scheduled.

Now ask yourself: how long did it take to find out what happened? How far back did you have to trace? How many people had to reconstruct events from memory, email threads, and spreadsheets that hadn't been updated since the last engineer who understood them left the company?

If your answer is "too long" — that's not a people problem. That's a systems problem. And it's a systems problem that gets more expensive every single time.

There's a particular kind of engineer — you probably know them, you might be one — who is extraordinary at the actual work. Circuit design, mechanical integration, software architecture, systems engineering. The kind of person who can hold an entire program in their head and spot a second-order risk that everyone else missed.

That person is also, at many hardware companies, the person who ends up managing the parts.

Chasing quotes on DigiKey at 9pm. Reconciling a distributor invoice against a BOM that's three revisions out of date. Arguing with a supplier about a minimum order quantity on a component that's only needed in quantities of four. Manually updating an inventory spreadsheet that four other people also have edit access to and nobody trusts.

This is not what they were hired to do. It is not what they are good at. It is not what moves your program forward. And it is not, by any measure, an acceptable use of the most expensive and specialized people on your team.

The companies that build the hardest products in the world — the ones doing things that genuinely matter, in regulated industries where failure has real consequences — need a procurement process that is as rigorous, as traceable, and as reliable as everything else they do. Not because procurement is glamorous. Because when it fails, everything downstream fails with it.

Cofactr was built to be the procurement execution layer that hardware teams don't have time to build themselves.

We handle the full stack: planning, purchasing, warehousing, inspection, kitting, and shipping. Not as a vendor you manage, but as an execution partner that removes procurement from your critical path entirely. Every part tracked. Every supplier accountable. Every kit built to your exact spec and shipped when your line needs it.

When something unexpected happens — and in hardware, something unexpected always eventually happens — you have a complete, traceable record of every component, every source, every inspection, every deviation. Not because we're optimistic about surprises. Because when they come, legibility is the difference between a two-hour root cause and a two-week investigation.

You build. We procure.

That's not a marketing line. It's a precise description of the division of labor that lets your engineers do the work they actually came to do.

Every kit Cofactr ships includes a pack of Pop Rocks.

We put them in the box deliberately. Not as a joke, and not as a gimmick — but as a small, tangible reminder of the thing this whole story is really about.

Mitchell's accident became a legend because his process was controlled enough to make the accident meaningful. Because he could look at what happened and understand it, trace it, and reproduce it. Because the discipline around the experiment was tight enough that when the experiment failed, it failed informatively.

That's what we're building every time we ship a kit. Not candy. Not chaos. A process disciplined enough that when you open that box, every component is exactly what it's supposed to be, sourced from exactly where it should be, and traceable all the way back if you ever need to know.

The Pop Rocks are a reminder that only some mistakes are okay.

Yours aren't.

Cofactr is a procurement execution platform built for hardware companies in aerospace, defense, medtech, robotics, and autonomous systems. If your procurement process isn't as rigorous as the products you build, let's talk.