The ask lands in every board meeting the same way. We need more ARR, and we need it showing up this year, not in the plan for next year.
The reflex answer is the pipeline. More spend, more reps, more top of funnel. It's a real answer with a real problem: a new-logo dollar is the slowest and most expensive dollar your company knows how to earn. The cycle is long, the spend is upfront, and the win rate is a coin you flip against strangers.
There are only four things that move ARR, and three of them live in the customers you already won.
The four levers
New logos. Sell to somebody who isn't a customer. This is the one every plan is built on, and it's the one with the worst economics of the four. It's also the only one that grows the base, so it never goes to zero. It just shouldn't be carrying the whole number.
Expansion. Existing customers buying more: more seats, more volume, the next tier, an adjacent product, a service they didn't take at first. This is the lever with the best economics and the least machinery behind it in most companies.
Price. What you charge for the same thing. Real, frequently underused, and constrained by how much value customers can actually point to. A price increase into a base that isn't getting value is a churn event with a delay on it.
Retention. Not additive, but it decides whether the other three compound or leak. Every dollar you don't lose is a dollar you don't have to win again, and the win-again version costs new-logo prices.
Notice what that list actually says. Three of the four operate on people who already pay you. Most ARR plans put nearly all the weight on the one that doesn't.
The math on second-order revenue
Every dollar of expansion revenue skips the most expensive parts of the new-logo dollar.
No ad spend. No cold outreach. No qualification, because the customer is already qualified. No trust-building, because the trust is the thing you already built. Typical cost lands at a third of new-logo revenue or less, and the cycle is a conversation between people who already work together rather than a procurement gauntlet.
Speed is the underrated part. An expansion motion started this quarter produces ARR this quarter, because some of your customers are approaching readiness right now. That count exists today. Most companies just can't see it, which is the actual reason the base never shows up in the ARR plan. Not because the money isn't there. Because nobody is instrumented to see that it is.
Why it isn't in your plan
Ask why the ARR plan is all new logos and you'll hear versions of the same answer. Expansion isn't predictable. We can't forecast it. It just sort of happens.
All true, in the same way an unplanted field is unpredictable.
Belief without machinery produces random revenue. Random revenue can't be planned. So the plan defaults to the expensive lever that at least has a spreadsheet behind it, and the cheaper lever stays in the category of pleasant surprise.
The machinery that makes it plannable
None of this is exotic. It's four things, and most companies have none of them written down.
An inventory of what customers can buy next. Not your price list. A mapping of what each account currently has against what exists, so the gap is a number rather than a feeling. Most teams cannot answer "how much room is in this account" without opening a spreadsheet and guessing.
Milestones that make readiness observable. The question is not whether a customer seems happy. It's whether they've reached the point where the next thing genuinely helps them. That needs to be something you can see in behavior rather than something a CSM feels on a call, or it can't be counted, forecast, or acted on by anyone but the person who felt it.
Something held back on purpose. If everything is bundled into what they already bought, there is no next thing to present when they arrive. Deliberate unbundling isn't withholding value, it's making sure there's a real, useful step available at the moment a customer becomes ready to take it.
An owner with a number. Expansion that belongs to everybody belongs to nobody. It doesn't need to be a new team. It needs one person whose forecast includes it and who gets asked about it in the same meeting where pipeline gets asked about.
Build those four and expansion stops being upside. It becomes a line item, which is what the board was actually asking for.
What to do in the next ninety days
In order, because the order matters.
Count what's there. Take your top fifty accounts and mark what each one has against what they could have. You're not building a forecast yet, you're finding out whether the room is large enough to bother with. Almost always it is, and almost always it's larger than the ARR plan assumes.
Pick one observable milestone. One signal that says a customer is ready for a specific next thing. One, not a scoring model. You're testing whether readiness can be seen at all before you invest in seeing it well.
Work the accounts that already cleared it. Some customers passed that milestone months ago and nobody told them what was next. That's not a pipeline you build, it's a backlog you clear, and it's the fastest ARR available to you.
Then instrument it. Only after you've done it by hand and know it works. Automating a motion you haven't run produces a system that reliably executes a guess.
The fastest ARR you can add is sitting in accounts you already won. It's cheaper than the pipeline, it closes faster than the pipeline, and it's the only ARR lever whose ceiling you set yourself.
One caution worth carrying into all of it. Increasing ARR and improving ARR are different jobs, and it's entirely possible to do the first while quietly doing damage to the second.
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.