Startup Stage

Startup advice contradicts itself constantly. One side tells founders to meet every customer in person and hold each account by hand. The other side says anything that runs on human effort can't grow, so build something that sells itself from the start. Both come from people who have actually built companies, and you can't follow both.

The contradiction comes from where each person was standing. At six people and twelve customers, the fastest path really is the founder picking up the phone and keeping every account close. Run the same play at a hundred people and three thousand customers and you can hire salespeople all year while costs climb and revenue doesn't follow. The first piece of advice is right in the first situation, the second is right in the second.

Pagedeck hit this in year two. One investor said to hire five salespeople and push revenue; another said it was too early to hire anyone. They had read the same materials. What split them was a judgment about whether the company had already confirmed that people would pay, and with no name for that judgment the meeting turned into an argument about who sounded more convincing.

Naming the stretches puts conditions on the advice. "Hire salespeople" becomes "hire salespeople if you have confirmed there are buyers." The line between stretches is drawn by the problem still unsolved, not by a company's age or headcount. A company that doesn't know whether anyone will pay and a company that knows but can't tell whether its selling method survives more headcount are standing in different places. Age is a poor divider: some companies are five years in and still looking for a first paying customer, and some pick up four hundred customers in eighteen months. A name also settles which numbers belong in the room, because a company still hunting for buyers gets nothing usable out of a monthly revenue chart. That same chart becomes the main thing worth arguing about once the buyers are confirmed and the open question is how many more of them exist.

The terms under startup stage fall into two groups: names for the stretches a company passes through in order, and names for the ways the road ends. The first four run in sequence. The last four are closing scenes that can arrive from any stretch.

The stretches divide by which problem is being solved, and there are four.

1. Ideation stage — nothing has been decided about what to build. There is no product and no customer, and the team is somewhere between one and five people. The work here is checking rather than building: talk to twenty candidate customers, hear how they handle the problem today, and sort out whether it is something they already spend money or hours on. The condition for moving on is a single one — real people would pay, and you can say who they are. If seventeen of those twenty call the problem annoying but spend nothing on it, it isn't worth selling into yet.
2. Early stage — the product gets built and pushed against a market. The assignment is product-market fit, the state where what you built lines up with demand in a specific market. The number that matters here isn't revenue, it's the return rate. If 100 teams try the product this month and 15 of them create documents again next month, the fit isn't there; if 60 do, it's close. You leave this stretch when shutting the product down abruptly would make some customers angry.
3. Growth stage — demand is confirmed and the job becomes scale. The question shifts from what to build to how to sell more of it. What you are looking for is not a method that worked once but one that returns the same result when repeated. If a million won of search advertising brought in four paying customers, check whether ten million brings in something close to forty. Repeatable is the operative word here, because a channel that produced four customers once and none over the two months after is a story rather than a method. You leave when the money spent acquiring a customer is reliably recovered from what that customer leaves behind.
4. Scale-up — headcount and revenue grow fast. The problems of the three earlier stretches sat in the market; here they move into the organization. Go from forty-five people to a hundred and fifty and the founder can no longer sit in every decision, and work handled by conversation yesterday starts demanding documents and process. What knocks companies over in this stretch is usually hiring speed and management, not weak demand. Once Pagedeck gets here, product calls that Mia holds alone today split across three teams, and without writing down who decides what, the same feature gets built twice.

The endings describe how the road closes, and there are four.

1. Exit — the umbrella name for turning the shares founders and investors hold into cash. Money an investor put in doesn't come back because the company is doing well; it comes back when someone buys those shares. That scene is scheduled from day one in any company that has taken investment. Which of the three shapes below it takes is rarely the founder's decision alone, since it depends on who is willing to buy at the time.
2. Initial public offering (IPO) — listing shares on a stock exchange so anyone can buy and sell them. After listing, results and risk factors have to be disclosed on a fixed schedule and the company gets a public report card every quarter. Preparation alone usually runs two to three years.
3. Mergers and acquisitions (M&A) — another company buys all or most of the shares and the two become one. It is faster than listing and the terms are negotiated with a counterparty, which is why this ending is far more common in practice. What the buyer wants, whether that is the customer list, the technology, or the team, changes the price, and the same judgment separates this from the next entry.
4. Acquihire — an acquisition aimed at the team rather than the product or the customers. The product often shuts down right after the deal while the people move over to the buyer. For founders it sits between closing the company and selling it.

The money to cross from one stretch into the next usually comes from fundraising, raising capital by selling equity or taking on debt, with each round carrying a name like seed or Series A. The two naming systems get spoken side by side and are easy to confuse, but they run on different axes. A round name says how many times money has come in. A stage name says which problem the company has solved.

The two groups carry the same story from opposite ends. The stretches decide what to solve next; the endings decide when and in what shape the road closes. Leaving the ending as a distant topic has a price. A company aiming at a listing has to set its rules for when revenue counts as revenue against outside standards early, and a company aiming at a sale needs the rights to its core technology held by the company rather than by individuals. Fixing either one later means rewriting past contracts one at a time.

"We're still early stage. Building out a sales team now skips a step."

Confirming that buyers exist comes before scaling up the way you sell.

"Closing a Series A didn't move us into growth stage. The return rate is where it was a year ago."

A round going up and the company's open problem changing are two different things.

"The goal this quarter isn't revenue, it's clearing the stretch. Twenty customers who'd miss us if we disappeared comes first."

What the company needs now is meeting the condition for the next stretch, not making a number bigger.

Pagedeck sits at forty-five people, just past a Series A. Putting board materials together, Grace asked Mia which stretch the company was actually in. Revenue had tripled in a year, but half of it came from eight accounts the chief executive had opened herself. Mia recalculated the figure with those eight removed, and the two of them settled on a company still standing at the doorway of growth stage.

That settled, the quarterly plan changed. A proposal to hire five salespeople came down to two, and the team decided to first find out whether anyone other than the chief executive could open accounts the same way. If the two close three accounts each within three months, the rest get hired. Fixing the stretch turned out to be a way of deciding what not to do this quarter.

Without that question the conclusion would have gone the other way. A table showing revenue tripled makes adding people to go faster look obvious. Asking which stretch first is what made anyone take the number apart, and only then did it surface that half of it came from one person.

The most common mistake is reading the stretches as a ladder you climb once and never come back down. The names are ordered and funding rounds only move one direction, so it looks that way. But what divides the stretches is the problem being worked on now, not elapsed time, and when the problem comes back the stretch comes back with it. If customers in the industry that produced half of revenue cut budgets all at once, a company in growth stage is back to looking for buyers. Starting a new product works the same way: the company overall can be in scale-up while that product begins at ideation. Read the names as a ladder and you treat open questions as settled, skip the checking, and hire against it, and a year later the only thing that grew is fixed cost. Admitting the move backward is faster. Pagedeck decided it was past early stage and then, six months on, watched one major industry cut budgets across the board and found itself back on the earlier question.

Treating round names as stage names is the other common error, since companies that raised a seed and companies that raised a Series A get called early stage and growth stage out of habit. A round describes how many times money came in and on what terms. Investors put it in ahead of the next stretch because they expect the company to get there, which is not the same as having arrived on the day the money landed. Plenty of companies raise a Series A while still looking for buyers. Overlay the two and you start staffing against the round: a Series A means there should be a sales team, so the seats get created first, and people hired before a selling method exists each run their own play and scatter six months later with nothing to show. It bends the other way too. A company whose selling method already works but couldn't raise the next round because of market conditions carries an early round name while doing later-stage work.

Both errors have the same shape. They let a label decide what the company does next, when the label is meant to be a reading of what the company has already done.