Deal Stage

Progress used to be reported as a percentage. "That one's about 70 percent there." The trouble is that only the person saying it knows what 70 means. For one rep it meant the proposal had gone out; for another it meant the customer sounded enthusiastic on a call.

So the same list produced different numbers depending on who read it. When three reps each write 70 percent on three deals and a founder adds them up into a revenue figure for the month, the money that actually arrives can be half that or triple it. Summing subjective confidence yields subjective confidence.

A deal stage replaces that estimate with a fixed slot. Instead of 70 percent, the record says "proposal," and the conditions for entering that slot are written down in advance. The goal is to narrow the room for judgment. It doesn't remove instinct, but it does make everyone use the same slot to mean the same thing.

Labeling positions this way isn't a new idea — it's the same structure a factory uses to mark which step a part has reached. What's different here is that a transaction depends on how people feel, so it can also move backwards.

A stage marks one of the positions a transaction passes through in order. Names vary by company, but the shape is usually five.

1. Discovery call — the first conversation, spent understanding the customer's situation and problem. It's a session for asking, not for presenting.
2. Demo — showing the product directly. Pointing at how the specific pain from the first call disappears on screen moves deals forward more reliably than walking the full feature list.
3. Proposal — sending a document that sets out scope, price, and term. It usually reaches people who were never in the meetings, so it has to hold up without anyone talking over it.
4. Negotiation — working out price and conditions. What moves here isn't only the number: payment timing, contract length, and termination terms are all in play.
5. Closing — confirming the signature process and the approval chain.

What matters isn't the count or the names but the condition for crossing between slots. Without those conditions written as sentences, people inside one company advance deals on different grounds. Write "move to proposal when the customer has named a buying timeframe and the approval chain is confirmed" rather than "move to proposal the day after the demo." The first version keys on what the customer did; the second keys on what the rep did. Only the first is usable for a forecast.

The stage is stored as one field on a deal row inside a CRM, the system that tracks customer contact history. It's the value that splits a pipeline view into columns.

Moving backwards is normal. A deal that reached negotiation and then lost its champion goes back to demo, because the explanation starts over. Leaving it where it was because demoting it feels like a loss means the total counts that deal at a higher probability than it deserves. There are only two ways out at the end: closed won or closed lost.

The value never stands alone. It's always an input to something being calculated or grouped.

The first is the sales forecast. The common method multiplies each stage by its historical close rate and adds the results — ten deals of $15,000 at a proposal stage that has converted at 40 percent produce an expectation of $60,000. Rates are kept per stage because a deal at the first call and a deal in negotiation cannot carry the same probability.

The second is finding where deals get stuck. Count how many entered each slot and how many moved on, and one column usually shows an outsized drop. If deals reach the demo reliably but half of them vanish at proposal, the problem is more likely the price or the document than the product.

The third is telling a rep what to do next. Once the stage is set, the next action is largely determined: a deal at proposal needs the approval chain confirmed, a deal at closing needs a signing date booked. This structure is also what makes it possible to tell a new hire where to start on their first week, instead of leaving them to guess which of thirty open deals deserves a call today.

"This one is still at demo. I haven't heard anything about budget, so I can't move it to proposal."

Stages are being set by the crossing condition rather than by effort. It's also a refusal to inflate the record.

"We're losing a higher share at the proposal stage than last quarter."

An observation that one column specifically got worse. A signal to look at pricing or the proposal itself.

"I moved that one back from negotiation to demo. Their contact changed."

A report that the deal went backwards, and a correction that keeps the record matched to reality.

At Pagedeck, Sofia reviews the list grouped by stage every week. Enterprise deals tend to sit in negotiation for a while, because security review and legal review both land there. Anything that has stayed in negotiation for three weeks or more gets flagged, with a line describing what's blocking it.

Grace reads the same list for the distribution rather than the total. Two pipelines can add up to the same amount while one has everything clustered at the first call, which means almost nothing arrives this quarter. The reverse — half the value sitting at closing — means next month is busy and the month after is empty. Either way, the decision to hire now or wait comes out of that shape.

The most common mistake is reading the stage as a measure of the rep's progress. Advancement looks like the reward for effort, so "I moved three deals forward this week" sounds like an accomplishment in a Monday meeting. But the value describes where the customer has arrived, not what the rep did. Run it the other way and people optimize for advancing deals, and six months later the closing column holds ten deals of which two actually close. The hiring plan and the revenue plan built on that list break together.

The second mistake is assuming more stages mean tighter management. Finer slices feel more precise, but past about ten columns people hesitate over where each deal belongs and end up picking whichever is closest. Before adding a column, check whether the existing ones have their crossing conditions written in a sentence. Adding slots without conditions only makes an inaccurate record look more detailed.