A few years ago, if you wanted to grow a software company, the answer was simple. Hire more reps. Point them at more logos. The market was cheap, money was free, and the spreadsheet only really cared about one thing: how many new names landed this quarter.
That world is gone. It has been gone for a while, actually, but a lot of sales organisations are still behaving as if the removal van is parked outside.
The number that decides whether a SaaS company lives or dies now is net revenue retention. NRR. It is the percentage of revenue you keep and grow from the customers you already have, before you add a single new logo. Above 100% means your existing base is growing on its own. Below 100% means you are running up a down escalator, and every new logo your shiny AEs land is just replacing revenue that quietly leaked out the back.
Here is the part nobody wants to say out loud. Most sales teams have no idea how to move that number, because the skill that moves it was never coached.
The number changed and the job didn't
Let me give you the figures, because vague claims are useless.
Public SaaS companies are running NRR around 110% on a good quarter. The broad market has compressed closer to 101%. Top-quartile companies sit at 115% and above. That spread, 101% versus 115%, is the difference between a company that is dying slowly and a company that is compounding.
And where does that growth come from? Existing customers. Expansion now generates over 40% of new ARR industry-wide, up from roughly 30% in 2019. For companies above $50M ARR, it is over half. Think about that for a second. The majority of new revenue at a mature SaaS company comes from selling more to people who already bought.
Then there is the cost side. Expansion revenue costs roughly $0.27 in customer acquisition cost per dollar earned. New-logo revenue costs around $1.13. One study of 174 SaaS companies put expansion at roughly five times cheaper than going and finding a stranger.
So the maths is not subtle. The cheapest, fastest, most defensible growth a software company has is sitting inside its existing accounts. And yet.
The accounts are there. The skill isn't.
Here is what actually happens in most companies. The new-logo AE closes the deal, does a triumphant handoff, and moves on to the next hunt. The account lands with an Account Manager or a CSM. And then the relationship enters a strange limbo where everyone assumes growth will just sort of happen.
It doesn't. Expansion is not the absence of churn. It is an active commercial discipline, and it is a different discipline from new-logo selling.
A new-logo AE is trained to create urgency, to manufacture a compelling event, to run a fast cycle against a competitor. None of that helps you inside an existing account. The customer already bought. There is no competitor in the room. There is no burning deadline. The skill you need instead is the patient, signal-reading, multi-threading, value-proving work of growing a relationship over nine to twelve months. That is a farmer's skill, not a hunter's, and almost nobody gets formally coached on it.
I have watched this happen up close more times than I can count. A genuinely talented seller, brilliant at landing logos, gets handed a book of existing accounts and quietly underperforms for a year. Not because they got worse. Because the job changed and nobody told them how.
What "coachable" actually means here
When I say expansion is a coaching problem, I am not being cute about it. I mean it has the specific shape of a coaching problem. There is a body of knowledge, the practitioners who know it outperform the ones who don't by wide margins, and the gap between the two is closeable with deliberate practice.
Let me prove it with the actual content. Here is what a competent expansion-focused AM knows that an untrained one does not.
They can read expansion signals. Not "the account seems happy." Happy is a precondition, not a signal. They know that a hit-a-limit event, an account bumping against its seat cap or API ceiling, converts to expansion at multiples of any other trigger. They know a funding round opens a window that lasts weeks 2 to 12, not months. They know a champion getting promoted is the single best signal there is, because it combines proven advocacy with expanded budget authority. The untrained AM sees usage tick up and calls the account "healthy." The trained one knows healthy and expandable are different words.
They run the QBR as a forward conversation, not a status update. The untrained AM presents slides about what the product did last quarter. The trained one uses the customer's own metrics to confirm value, then pivots, deliberately, to "given where you are trying to get to, what should we be doing together that we aren't?" That one question opens expansion, renewal, and churn-risk conversations all at once.
They multi-thread. LinkedIn research says 78% of sales reps are single-threaded. Gartner says the average B2B purchase now involves 11 stakeholders, and over 80% of sellers have lost or stalled a deal because one stakeholder left. The trained AM builds three real relationships across three functions before they need them. The untrained one has one warm friend and calls it a relationship, right up until that friend changes jobs and the account evaporates.
They get the timing right. Expand before the renewal, never at it. The trained AM knows that stacking expansion onto a renewal hands the whole conversation to procurement, whose literal job is to cut. The untrained one "mentions" the upsell during the renewal call and watches both the expansion and the margin die together.
They know when not to expand. This is the counterintuitive one. The trained AM can look at an account with two health-score red flags and correctly decide the right move is retention, not a bigger ask. The untrained one pushes anyway, because pipeline is thin, and burns the relationship doing it.
Every one of those is teachable. Every one of those is a behaviour you can name, model, practise, and correct. That is the definition of coachable.
Why training alone won't fix it
Now, the obvious objection. "Fine, James, so we run an expansion training. Problem solved."
No. And if you have read anything else I have written, you knew that was coming.
Training, the workshop kind, has a known and dismal failure rate. People lose more than half of what they learn within 24 hours. A two-day expansion course produces a room full of people who can define a hit-a-limit signal on Tuesday and have forgotten it by the Thursday they actually need it. The course teaches the knowledge. It does nothing about whether the knowledge gets executed in the live account, on the real QBR, with the real customer on the line.
This is the gap I spend most of my professional life thinking about. The distance between what a rep learns and what a rep actually does in the field. Training puts knowledge in the room. It does not change behaviour in the workflow. And expansion is a workflow problem to its bones, because the signal you need to read is in the usage data this Tuesday, and the QBR you need to run differently is on the calendar for next Wednesday.
You cannot fix that with an annual offsite. You fix it where the work happens, at the moment the work happens, with coaching that shows up in the rep's actual week.
The expansion playbook is a coaching system
So here is the argument, landed.
Net revenue retention is the number that matters. Expansion is the lever that moves it. Expansion is a distinct, learnable skill that most AEs and CSMs were never coached on, because the industry spent a decade obsessed with new logos and never built the muscle for growing the ones it had.
The good news, and there is genuinely good news, is that expansion is one of the most coachable skills in all of sales. It is pattern-heavy. It rewards discipline over charisma. The failure modes are well-known and repeatable, which means they are well-known and correctable. An AM who reviews their last ten expansion losses honestly, names the pattern, and gets coached against it will outperform one who doesn't. That is not theory. That is just what happens.
The companies that win the next few years will not be the ones that hired the most hunters. They will be the ones that figured out expansion is a coaching problem, and then actually coached it, in the workflow, in the week, against the real account. That is exactly what AI sales coaching is built to do.
The logos you already have are the cheapest growth you will ever find. Stop treating them like a finished sale.