Business Model

For a long time the sentence companies used to introduce themselves was "we build this product." That sentence carried what the company made. It carried nothing about whether the company worked. A product being good and a product leaving money behind are two separate facts, and only the first one was in the pitch.

So investment reviews and planning meetings kept stalling in the same place. The presenter would spend twenty minutes on features, someone across the table would ask where the money comes from, and the answer would come back as "once we have enough users, there will be a way." The next question — how much more it costs when one more user signs up — was usually where the talking stopped. The feature list existed. The user count existed. The box connecting those two to money was empty.

An empty box splits the room inside the company too. Engineering argues that growing usage comes first, sales argues that charging comes first, and nothing written down anywhere settles which one is right this quarter. Both sides can point at something real, so the argument repeats at the next meeting with the same two positions and no new evidence.

How much that box matters shows up when the same product turns into different companies. Take one collaborative document editor. A company charging per person per month, a company charging by storage used, and a company charging nothing and selling ads run on completely different logic. The first one spends its energy adding seats inside accounts it already has. The second builds whatever makes people upload more files. The third needs time spent in the app, so it sends more notifications. The screens look alike, but the part of the product each team touches next quarter is not the same, and neither are the numbers they read out in meetings.

Business model is the name given to that empty box. It asks a company to write down, in one piece, who it serves, what those people get, how the company collects payment for it, and what it has to spend to deliver. Writing it as one piece matters because the items hold onto each other. Change the customer and the price you can charge changes. Change the price and the cost you can carry changes. Change the cost and the product you can build changes. Fix one box on its own and the rest fall out of line, and scattered across separate documents that misalignment goes unnoticed.

The terms filed under business model fall into three groups: what the structure is made of (components), which arrangements already exist in the world (types), and what form you write it down in (tools).

Components are the three pieces you get when you take the structure apart — money coming in, money going out, and the two set against each other for a single customer.

1. Revenue model — the paths money arrives through. Common ones are a subscription at a fixed amount each period, usage-based billing that charges for what was consumed, a commission taking a percentage of each transaction, and advertising sold instead of a usage fee. What gets decided here is the method of collection, which is why how much to charge belongs to pricing as separate work. Only once the method is settled as a subscription does the question of 9,000 won versus 15,000 won a month have a place to sit.
2. Cost structure — the shape of what goes out. Split it into fixed costs, which stay put whether customers arrive or leave, and variable costs, which move with customer count. At Pagedeck, salaries for 45 people and the office lease are fixed — doubling signups this month does not move them — while the server bill for storing documents rises as usage rises. The ratio between the two sets the character of the business. A company that is mostly fixed cost loses money until it crosses breakeven, and past that point each additional customer adds margin quickly. That is why the same monthly loss reads as alarming in one company and as expected in another: what matters is how many more paying teams sit between today and the crossing point.
3. Unit economics — profit and loss measured on one customer or one transaction. Company-wide results can be made to look better for a quarter by cutting marketing or delaying hiring, and per-customer math does not bend that way. If bringing in one account costs 60,000 won, the account pays 9,000 won a month, and 2,000 won of that is cost of service, recovering what the account cost takes close to nine months. If the average account stays six months, growing the customer base makes the hole deeper, and what that calls for is a change to the structure rather than a larger advertising budget.

Types are arrangements enough companies have walked through that they carry names. Starting from a shape that already works beats designing one from a blank page.

1. Two-sided market — a structure that only runs once sellers and buyers are both present. Nobody lists a place to stay when no travelers come, and no travelers come when nothing is listed, so the early work is filling one side by force, with money or with effort. Once both sides pass a certain size, growth on one side pulls the other side along, which makes the position hard for a later entrant to take. Which side pays is its own decision. Charge both and neither shows up; leave one side free and the price on the other has to cover that side too.
2. Asset-light model — earning by connecting what already exists rather than owning it. Linking people who have cars with people who need rides instead of buying cars, or renting someone else's servers instead of building a server room. Upfront investment stays small and a drop in demand is easier to survive, but sitting on top of someone else's asset means their change in terms moves your entire cost base.

Tools are the format that lays every item on one page so the gaps show. The business model canvas belongs here. Nine boxes — customers, value, channels, revenue, costs and the rest — sit on one screen for a team to fill in together. Its usefulness is that an item nobody thought to ask about becomes a visibly blank box.

The three groups prop each other up rather than sitting side by side as options. A type supplies the starting shape, components translate that shape into this company's numbers, and a tool collects those numbers on one page where the missing ones stand out. Written in that order, Pagedeck reads like this: the shape is let teams try it free and charge once the team commits, money arrives through per-seat subscriptions and enterprise contracts, and money leaves as a mostly-salary fixed base with a server bill that tracks storage. Write that much down and one number turns out to decide what the company keeps — the share of free teams that convert to paid. Looking at the three groups separately never reaches that conclusion.

"The feature is good, but I can't see the business model. Where does money come in from this?"

What to build is settled, and the path from building it to revenue has not been explained.

"This isn't a price change, it's a business model change. Charging by storage instead of by seat means rewriting how we sell and which numbers we watch."

Changing the method of collection drags in work well outside the product.

"A million free users is just a cost until there's a structure that charges someone."

User count on its own is not a business yet; at this stage it's a server bill.

The meeting where this comes up most at Pagedeck is the quarterly business review Grace runs. Team conversions to paid rose last quarter while the money left over stayed flat, so Grace asks the room to recompute the numbers for a single account. Sofia says closing one enterprise customer takes sales five months on average. Mia adds that those five months also generate product requests that land on her team. Set against the size of those contracts, whether that cost is worth carrying becomes visible right there.

The conclusion does not always come out as raise prices. Last quarter the decision was to stop taking on the segment that needs five months of sales attention, and to widen the path where a team signs up with a card on its own instead. When who pays and how they pay changes, the feature list for the next quarter changes with it. The item at the top of Mia's list came out as an admin screen for inviting people and switching payment methods rather than anything to do with editing documents. A decision about the path money takes reordered the product screens.

The most common mistake is using business model as another way to say how we make money. "Our business model is subscription" sounds natural enough to say, which makes the shorthand easy to fall into. But a subscription is one method of collection. In front of it sits who the customer is and what they get, and behind it sits what serving them costs. Drop both ends and the only problem left to discuss is price. Every time margin runs thin the only lever in reach is charging more, while the fact that the cost of bringing in a customer is climbing faster than the price never comes up. Plenty of other levers exist without touching the collection method at all — narrowing which customers to serve, cutting the hands needed to support each one, or shifting acquisition toward a cheaper path.

The second mistake is treating it as settled once written. A document prepared for investors in the first year does feel like a finished blueprint, so the assumption is fair. In practice the structure has to move when the customer moves. A company that served individuals and starts taking enterprise customers finds contract length, payment terms, and the headcount needed for support all different, and those differences hit both what comes in and what goes out. Hold onto the original structure while the reality underneath shifts and the numbers break first. Revenue climbs, what's left over shrinks, and the quarter ends with nobody able to point at the cause.