How a Business Actually Works

Module 01

How a Business Actually Works

Before frameworks, before finance: a clear mental model of how any company creates value, makes money, and loses it.

Everything starts here. If you can explain, for any company, how it makes money and where that money could leak away, you are already more useful than most people who give business advice. The vocabulary comes later — the model comes first.

The value loop

Strip away the jargon and every business runs the same loop: it acquires inputs (people, materials, software, capital), does something to them that creates value a customer wants, sells that output for a price, and keeps the difference between price and cost as profit. Then it reinvests some profit to do it again, bigger. A "good business" is one where that loop is profitable, repeatable, and hard for others to copy.

Your job as a consultant is almost always to find where the loop is weak: the company can't acquire customers cheaply enough, or it can but they don't come back, or they come back but each sale loses money, or it makes money per sale but can't grow without costs exploding. Naming which link is broken is most of the value you add.

Business model: how the money actually flows

A business model is the specific answer to four questions: who is the customer, what do you sell them, how do you deliver and charge for it, and why is that profitable. Common shapes you'll meet constantly:

  • One-time sale — make a thing, sell it once (a bakery, a furniture maker). Simple, but you start every month at zero.
  • Subscription / recurring — charge repeatedly for ongoing access (software, gyms). Predictable revenue; the whole game is keeping customers from leaving.
  • Marketplace / platform — connect buyers and sellers, take a cut (ride-hailing, app stores). Hard to start, very hard to displace once it works.
  • Services — sell expertise and time (agencies, consultancies — including yours). Easy to start, hard to scale because you're selling hours.
  • Advertising / data — give something away free, monetise attention or data. Needs huge scale to work.

The model shapes everything else. "Cut costs" means something completely different to a subscription business (protect the experience that keeps people paying) than to a one-time manufacturer (squeeze the supply chain). Always identify the model first.

B2B vs B2C — two different worlds

Selling to businesses (B2B) and selling to consumers (B2C) behave differently in ways that change every recommendation. B2B usually means fewer customers, larger deals, longer sales cycles, multiple decision-makers, and buying driven by return on investment. B2C means many customers, smaller transactions, fast emotional decisions, and buying driven by desire and convenience. Advice that fits one can be actively wrong for the other — a B2B company's growth problem is often "the sales process," while a B2C company's is often "the funnel and the brand."

The functions inside the machine

Inside any company, work clusters into a handful of functions. You don't need to master each, but you must know what each worries about, because that's where problems hide:

  • Product / Operations — building and delivering the thing. Worries about quality, cost, speed.
  • Sales & Marketing — getting and keeping customers. Worries about demand and acquisition cost.
  • Finance — the money, the books, the forecasts. Worries about cash and profitability.
  • People / HR — hiring, retaining, organising humans. Worries about talent and culture.
  • Leadership / Strategy — deciding where to point everything. Worries about the future and the competition.

The value chain & where advantage lives

A sharper lens (from Michael Porter): picture the company as a chain of activities that each add value, from inputs to the customer's hands — inbound supply, operations, outbound delivery, marketing, service — supported by infrastructure, people, tech, and procurement. At each link, ask: is this where we're special, or is this just cost? Competitive advantage lives in the links you do better or cheaper than rivals; everything else is a candidate for simplifying or outsourcing.

The durable forms of advantage — a company's moat — are worth memorising, because "do they have a moat?" is a question you'll ask constantly: cost advantages (you can produce cheaper), network effects (the product gets better as more people use it), switching costs (leaving is painful), intangible assets (brand, patents), and efficient scale (the market only supports one or two players). No moat means profits get competed away no matter how good the team.

Growth versus profit — the tension at the centre

New consultants assume "grow" and "make money" point the same way. Often they fight. Growth usually costs cash upfront (hiring, marketing, inventory) before it pays back, so a fast-growing company can be profitable on paper yet starved of cash — or deliberately unprofitable while it lands a market. The right answer depends entirely on the model and the moat. Your job is to make the trade-off explicit, not to assume one side.

Key takeaway

For any company you advise, be able to say in plain language: this is the model, this is whether it sells to businesses or consumers, this is how a unit of value flows from input to paying customer, this is whether it has a moat, and this specific link is where the problem is. If you can't yet, you don't understand the business well enough to advise it.

For Orelis & the app

Your intake flow should force this model out of the user early: what do you sell, to whom (B2B or B2C), how do you charge, where's the moat, and what's the one number that's hurting? An AI consultant that skips straight to advice without pinning down the model is guessing. The model is the context that makes every downstream answer specific instead of generic.

Test yourself

Q1Pick a business you know well. In two sentences: what's its model, and which link of the value loop is most fragile?
Show a worked answer
There's no single right answer — the skill is the structure. Example (cafe): one-time/repeat sale of food and drink to local foot traffic; the fragile link is repeat visits, because a cafe lives or dies on whether the same people come back, not on whether the coffee is technically excellent. Notice you named the model first, then a specific weak link.
Q2A subscription software company and a furniture maker both ask you to 'reduce costs.' Why is that the same words but a different job?
Show a worked answer
For the furniture maker, costs are mostly materials and production — cutting them can directly lift margin with limited downside. For the SaaS company, the cost that matters is often the experience and support that stop customers cancelling; cut the wrong thing and churn rises, destroying the recurring revenue that is the entire model. Same instruction, opposite risk profile — which is why you identify the model first.
Q3A startup is growing 100% a year but burning cash and asks if that's good or bad. What do you need to know before answering?
Show a worked answer
Whether the growth is buying a moat or just buying losses. Ask: what are the unit economics (does each customer eventually pay back the cost to acquire them — see Module 2), is there a network effect or switching cost forming, and how much cash runway is left? Fast growth on sound unit economics with a forming moat can be excellent; the same growth on broken economics is digging a deeper hole faster. Growth is neither good nor bad until you see what it's buying.