The last piece ended on a hinge. Something that isn't repeatable can't be forecast, and something that can't be forecast can't be a channel.
That leaves a question it didn't answer. What makes expansion repeatable?
The instinct is to reach for a motion. Better coverage, better timing, somebody whose actual job it is. All of that helps and none of it is the thing, because a motion can only work on what already exists to be sold.
A forecast needs two things
Take any forecast you already trust. New business has a list of accounts and a stage each one is sitting in. That's it. That's the whole apparatus. You know what could close and you know roughly when, so you can put a number next to it and then defend the number to somebody who doubts it.
Expansion usually has neither. Ask what a given account buys next and you get a shrug or a wish. Ask when and you get "when they're ready," which isn't a date. It's a hope with a calendar invite.
So the number gets left out, not because anyone doubts expansion exists, but because there's nothing there to count.
Unbundling produces both
Strategic unbundling is the decision about what comes out of the initial sale. The first deal carries what the customer needs and can actually use in the window between signing and first value, and everything else gets held back and attached to the milestone that earns it.
Do that and you've built the two things the forecast was missing.
You have a list. For every account the next item is a known item, because you decided which items were next before anybody signed anything.
You also have a trigger. A milestone isn't a mood. It's an observable thing that either happened or it didn't, and when it happens the item attached to it comes due.
List plus trigger. Same apparatus new business runs on, pointed at the base.
What it changes about the number
An expansion forecast built this way isn't a guess about how customers behave. It's an inventory crossed with milestones, and you already know both sides of that.
Which is the difference between "expansion should be around 15% this year" and "these 40 accounts have this item next, and 12 of them cross the milestone that opens it before Q3."
The first is a hope with a % sign attached. The second survives a board meeting.
The part that stings
If expansion isn't forecastable in your company, the honest read isn't that expansion is unpredictable by nature.
It's that nobody decided what came next, so there's nothing there to predict.
Nobody forecasts a surprise. Pricing is what stops it being one.
Lincoln Murphy formally named and popularized Customer Success starting in 2010 and has spent 15 years connecting it to expansion revenue and commercial outcomes. Read The Premise.