What Makes a Framework a Framework

A framework that only works in the place it was born is not a framework. It is a case study with numbered headings.

That is a harsher standard than it sounds, and most methods you meet fail it quietly. Someone builds something that works, writes down what they did, numbers the steps, and calls it a framework. But what they wrote down is a description of their situation, not a set of principles. You can tell the difference with one question: what survives when you take the domain away?

The eight-phase framework this site is built around comes from a factory. It was built inside a custom manufacturing operation, proven there, and everything I have written about it uses factory nouns: bills of materials, workstations, production variance. I have claimed from the start that every business sits on a product and that the architecture is industry-agnostic. But a claim like that deserves scrutiny, not repetition. So this piece does the honest version of the exercise: take each phase, strip the manufacturing away, and look at what is left. If what remains still tells you what to do, it is an invariant. If nothing remains, it was factory detail all along.

Phase by phase

Product formulation

The published principle: “You can’t automate what you haven’t decomposed.”

In the factory this meant decomposing a few hundred catalog items into bills of materials, labor minutes, and overhead, down to the atoms. Nothing about that sentence requires a factory. Imagine a caterer who wants to quote events without guessing. The homework is identical: every dish decomposed into ingredients, prep minutes, and plate cost. The nouns change; the move does not. Name the atoms of what you sell, because everything else will be computed from them. Decompose before you systematize.

Operational mapping

“Map the machine before you optimize it.”

The factory version mapped the workflow across seven workstations and linked every operation back to the product data, workarounds included. A landscaping contractor would do the same thing with a crew’s day: load-out, drive time, the cut, the cleanup, who waits on whom. Until the day is mapped, “we should be faster” is a feeling, not a plan. Map what actually happens before you touch any of it.

Relational database

“One source of truth.”

In the factory this is one relational core connecting materials, products, projects, clients, and quotes, so every report pulls from the same place everything else does. Now picture a repair shop with jobs in one app, parts in a spreadsheet, and customers in somebody’s phone. That is three memories that disagree. Put the job, the part, and the customer in one structure and reporting stops being archaeology. The invariant is the same five words in any industry: one source of truth.

Knowledge base and SOPs

“Capture the tribal knowledge before it walks out the door.”

The factory version wrote down the steps everyone “just knew” and made them retrievable. An independent clinic has the same exposure: the intake exceptions only the longest-serving person at the front desk can handle. Every “you just have to ask her” is an undocumented procedure carrying real operational risk on two legs. Externalize the tribal knowledge, whatever your tribe happens to make.

Project intelligence

“Every completed project is a test of your system’s accuracy.”

This is where the factory data got loud. Comparing estimates to actuals surfaced an operation estimated at more than twenty minutes that actually took about seven, a roughly 3x error that had been priced into every relevant quote. The caterer meets the same truth by comparing quoted kitchen hours to the hours the event actually consumed. The gap between estimate and reality is the most valuable data the business produced that week, in any business. Every completed job is a test of the system, if you grade it.

Configure, price, quote

“A quote is not a creative act. It’s a calculation.”

The factory assembles quotes from real costs: specs plus materials plus labor plus margin. The landscaping bid works the same way once phase one exists: area, crew rate, material, margin, assembled from decomposed components instead of “what did we charge them last time.” If you cannot reduce pricing to a formula, even a complex one, you do not yet understand your costs well enough. Price as calculation, not improvisation.

Sales and client management

“Sales is not separate from operations.”

In the factory, the promises made while selling are tied to production truth: what the shop can actually deliver, by when, at what cost. A repair shop lives this invariant every time it promises a turnaround its bays can actually honor, because delivery data has corrected the promise. The principle underneath is the one the next section is about, because it is the structural heart of the whole framework.

Content and marketing

“If Phases 1–7 are done right, marketing is just publishing what your system already knows.”

The factory version of this phase is the site you are reading. The clinic’s version would be publishing what its own records already show, real wait times, real outcomes, instead of inventing claims. When the system underneath is honest, marketing stops being fiction-writing and becomes publication. Publish what the system knows.

The loop is the tell

If you only remember one part of this piece, make it this one, because it is the strongest evidence that the eight phases are a framework and not a numbered story.

Phases 5 through 7 are not a sequence. They are a bidirectional loop: production truth corrects quote accuracy, quote accuracy disciplines the promises made while selling, and those promises are tested against production truth on the next job. Take away the factory and the loop does not collapse; it just changes clothes. Delivery truth, pricing accuracy, and the promises you make must continuously correct each other in any business that estimates anything. The caterer, the contractor, the repair shop, the clinic: each one either closes that loop or slowly drifts into quoting fiction.

Here is why this matters for the framework question. A checklist transfers trivially and proves nothing; anyone can renumber their steps for a new industry. An architecture transfers only if the relationships between the parts survive the move. The loop is a relationship, not a step. Its transfer is the real test, and it is the part I would watch most closely in any instantiation outside the factory.

The test ahead

Everything above is an argument, and I want to be honest about its weight. Argument is not evidence. Today the framework has one full instantiation: the factory system, built and in daily use, whose shape I have documented in detail elsewhere on this site. The invariants in the right-hand column of that diagram are a prediction about every instantiation that does not exist yet: whatever the domain, it will have to name its atoms, map its work, keep one source of truth, write down its tribal knowledge, grade every job, price by calculation, and close the loop between what it promises and what it delivers.

A framework earns the noun through transfer, and transfer is a test you can fail. That is exactly what makes it worth claiming. If the eight phases ever run somewhere with no workstations and no board feet, the interesting result will not be that the phases were followed. It will be whether the loop held.

If you run a factory, the blueprint will turn these phases into a concrete starting plan for your shop today. If you run anything else, read the right-hand column of that diagram once more, slowly. My bet is that most of it reads like a description of your business too. That bet is the framework.