The System Everyone Was Asking For

There was a stretch around 2017 where I could tell you exactly what we were building, but not when, not how many, and not who needed to know.

We had a full management team by then — marketing, ops, purchasing, assembly, quality, R&D, development, HR, accounting. Every chair was filled. On paper it looked like a company that had grown up. In practice, the way work actually moved through the building had not changed since we were a fraction of the size. The person with the most pull in the room decided what we were doing, when, and how. Everybody else found out when it landed on their desk.

You could see it most clearly in the handoffs, because that is where it broke. Marketing could not schedule a photo shoot or start building collateral, because nobody could tell them when a product would actually be ready to shoot. Purchasing could not order parts on time, because there was no signal that said go and no projection telling them how many. Same story in assembly, in quality, in every department downstream of a decision. Not because anyone was bad at their job. Because there was no mechanism that told one department what another one was about to need.

This was right as the GC1K and more Spartan models were coming down the line — real products, real money, real launch dates — and we were aligning on all of it by hallway conversation and whoever happened to be in the room.

I would love to tell you I had some flash of insight here. I did not. I went looking for help. I found a book on stage-gate development, read it cover to cover, and realized the thing we were missing had a name and a shape.

But I did not walk out of that book and start drawing gates on a whiteboard. The first thing I did was interview every department stakeholder, one at a time. Before I built anything, I wanted to know what each of them actually needed — and when, and from whom.

Two things came out of those conversations. The first was the gap I expected: no visibility into production timelines, no settled bill of materials, no handoff that told a department when something was coming, what it was, and how many. The second one surprised me. I had braced for resistance — people protecting turf, "we have always done it this way." That is not what I got. Every department was asking for the same thing: a way to understand the development cycle. They were not fighting a system. They were starving for one.

That reframed the whole project. I was not imposing process on people who did not want it. I was building the thing the floor had been quietly requesting the entire time.

So we built it. Five gates for full development programs — concept through launch, each gate a decision point where a project had to earn its way forward before the next department spun up. Three gates for product improvements and smaller projects, because forcing a minor change through five gates is its own kind of waste. Fielder's choice on which track a program ran. The gates were not there to slow things down. They were there to make the handoff real — so that when purchasing got the go-ahead, it was a real go-ahead, with a settled BOM and a number behind it.

Visibility and buy-in jumped almost immediately. For the first time, a department could look at where a program sat and know what was coming and when. That part felt great.

Here is the part I will not dress up: it added work. Specifically, it added work to my development team. Gates have to be prepared for. Documentation has to exist. Approvals have to be chased and recorded. We traded a chaos that was cheap for the person at the top and expensive for everyone downstream, for a discipline that cost my team more up front and saved the whole company on the back end.

I would make that trade every time — because what we bought was not speed. It was alignment. The entire building finally rowing in the same direction, on the same information, at the same time. You do not get that for free, and you do not get it by accident.

That is the lesson I carried out of it: a real process costs something, and if it does not, it is not real. The discipline is not in the meetings or the templates. It is in the handoff — making sure the next person down the line gets what they need before they need it. And it is in what you choose not to build yet.

If I had AI...

The single biggest lift in that whole project was the interviews — sitting down with every stakeholder, one at a time, to map what they needed. Today I would send a structured intake, let AI run and synthesize the first pass, and walk into the room already knowing the shape of the answer. The gate machinery itself — approvals, change requests, status, the endless "where does this program stand" — is exactly the kind of coordination AI runs now without a human chasing it. And document control and distribution, getting the right BOM at the right version to the right department, is the handoff problem I was solving by hand, solved automatically.

It would have taken real time off the build.

But here is the line I keep coming back to, and it is the whole reason I write this series: AI could have helped me build that system faster than I would believe. It could not have noticed the system needed to exist. Nobody handed me a ticket that said "your handoffs are broken, go fix them." I had to feel it, name it, go find the book, and do the interviews. AI runs the machine beautifully. It still does not decide to build one.

So here is my question for you: what is the system everyone in your shop is quietly asking for — that nobody has been given permission to build?

Matt Foster spent 18 years inside an Arkansas manufacturer — from a napkin sketch to a $400M acquisition — building the systems a growing company needs before anyone thinks to ask for them. He now runs Kre8ive Studio, helping small manufacturers build the process the floor is already begging for.