Chapter 5: Modules Are a Commercial Strategy
Week twelve of the sales cycle. The quote review meeting. Seventeen line items on the screen: part numbers, quantities, unit prices.
The customer's operations director scans the list and asks one question: "Can you guarantee uptime?"
The room goes quiet. Because the quote can't answer. Nothing on that screen speaks the customer's language. They asked about an outcome, and we handed them inventory.
Here is the law underneath that silence, and it took me years in the field to say it this simply: you get what you model. If you model parts, you'll talk parts. Your quotes, your sales conversations, your pricing arguments, and eventually your strategy all inherit the structure of your product model. Which means the structure of your product model is not a technical artifact. It is your commercial strategy, whether you designed it or not.
Two truths, one beam
The most useful structural idea I know is embarrassingly simple: your product needs two bills of materials, and they must not be the same one.
Customers buy outcomes. Finance buys parts. So you maintain a Sales BOM, the promise a buyer understands: "500 kW oil-free compressed air, certified Class 0, with guaranteed service response." And you maintain a Cost BOM, the plan your factory trusts: castings, controllers, fasteners, hours. Between them sits a stable mapping layer, and that mapping layer is your structural beam. Engineering can swap a supplier, redesign a casting, change the internals entirely, and the promise the customer bought stays intact. The interface holds while the implementation moves.
Companies that quote from the Cost BOM get the seventeen-line-item meeting forever. Companies that sell from the Sales BOM get to have a conversation about uptime.
What a module actually is
Strip away the methodology decks and a module is this: the smallest stable unit you can build, stock, and source, wrapped in an interface you promise not to break. The interfaces are the strategy. Inside a module, engineers are free. Between modules, nothing changes without a decision, a version, and an impact test, because every downstream quote, price, and factory routing depends on those boundaries holding.
Defining the modules forces the conversations most companies have been avoiding for a decade. Which variety do customers actually pay for? A client of mine offered two hundred motor variants; analysis of sold configurations showed customers cared about five performance envelopes. Two hundred variants was not customer intimacy. It was two hundred prices to maintain, two hundred things that could go wrong in a quote, and a fog that made every AI initiative hopeless. Stop offering 200 motor variants. Offer 5 performance envelopes.
And name every decision. "Environment: Standard, Marine, or ATEX" is a module choice a system can validate and a new salesperson can understand. A free-text field that says "see notes for special coating" is a future warranty claim.
This is why modularization run as an engineering cleanup project fails. Engineering can define the units, but only the business can decide what to offer, what to refuse, and where variety earns its keep. Those are pricing decisions, market decisions, strategy decisions, executed in engineering language. The companies that get this right run modularization from the executive table, with engineering as the master builder rather than the sponsor.
Anti-pattern: The Parts Parade. Quotes that read like a warehouse inventory: forty rows of article numbers marching past a buyer who wanted to know if the machine will do the job. The Parts Parade doesn't just confuse customers. It invites procurement to attack the quote line by line, and price erosion follows.
The test of whether your structure works is not architectural elegance. It's this: can a salesperson in their first month configure a valid machine and explain it to a customer without calling engineering? When the answer is yes, your expertise has become infrastructure.
There's one more structure to fix before the AI chapters, and it's the one where the money leaks directly: price.
Modularization isn't an engineering cleanup. It's your commercial strategy, written in engineering language.