Land and expand: Selling to the person who never used it

A team of six adopts your product, loves it, and pays $90 a month on a company card. Eighteen months later somebody wants to roll it out to four hundred people.

Nothing about the first sale prepares you for the second. The person who adopted the product is not the person who signs that contract, has not used it, and does not care about the things your champion cares about. They are thinking about security review, data residency, what happens when the champion leaves, whether this overlaps with something the company already pays for, and who is accountable when it breaks.

Land and expand fails most often because teams treat the second sale as a larger version of the first. It is a different sale, to a different person, with different objections, and the product’s job changes completely at that boundary.

What is land and expand?

Land and expand means winning a small initial commitment (one team, one use case, one project) and growing revenue inside that account over time, not negotiating the whole thing up front. It is the natural commercial shape of product-led growth, because a product that can be adopted without permission is a product that necessarily lands small.

Expansion arrives in four ways, and they are not equally available to every product:

There is a fifth route that only some products have, and it is easy to miss because it starts outside work: people who use something at home bring it to the office. Products with a personal or family tier get advocates who arrive already fluent, and the business conversation starts when a compliance requirement makes the personal version untenable.

Most companies can do one or two of these well. Trying them all at once produces a pricing page nobody can read.

Your champion has to survive a meeting you are not in

Mapping the second buyer is the practical core of the whole pattern.

At some point somebody inside the customer will make the case for spending real money on you, in a room you are not invited to, against alternatives you will not hear about. Everything that determines whether that goes well is something you gave them beforehand, or failed to.

So the question is not what you would say. It is what they can hold up. Four things earn their place:

Evidence of use they did not have to assemble. Who is using it, how often, what they have produced with it. If your champion has to build that from screenshots, most champions do not, and the case gets made on impressions and loses to a cheaper suite deal.

A per-team or per-department cost breakdown. The person approving wants to know what this costs and what it replaces, which is as much a packaging question as a sales one. A single line item for four hundred seats invites a negotiation; a breakdown invites a decision.

The security and compliance answers, in a form they can forward. Not a promise that you will respond to a questionnaire. The completed questionnaire, the certifications, the data-handling summary, findable without asking you. Expansion deals die here, and they die quietly.

A short answer to “why not the thing we already pay for.” Your champion will be asked this, and if you have not given them an answer they will improvise one, badly.

The real competitor is consolidation

Expansion is rarely lost to a better product. It is lost to a suite the company is already buying.

The argument that kills bottom-up growth at scale is not “this is worse” but “we are standardizing, and this is one of eleven tools doing overlapping things.” That conversation happens on a spreadsheet, driven by someone who has never used any of the eleven, at a moment you may not know about.

Two things make you survivable in it. Being genuinely used, in a way that shows up in the usage data. A consolidation exercise sees measured usage and nothing else, whatever the tool’s quality. And being spread across more than one team, since a tool that three departments depend on is a harder line to delete than a tool one department likes. That is the same insight as distributed integration adoption: lock-in that no single decision-maker can undo.

Landing small and staying small is not land-and-expand. It is a discount. If accounts arrive and never grow, the small first deal was not a strategy, and the thing to measure is expansion velocity, not how many logos you added.

Narrow, deep adoption in one team is more fragile than shallow adoption across four, even though it looks healthier on every engagement metric you have.

Accounts with no room to grow

Willingness to sell is where product-led companies get stuck on principle. Past a certain deal size somebody has to answer questions in a meeting, and refusing to staff that is not purity, it is leaving the largest deals to competitors who will. The product-led part is that usage tells you which accounts deserve a human, which is precisely what a product-qualified lead is for.

Two organizational decisions decide whether the human half helps or fights the product half. Put both motions on the same targets, because a sales team measured on new logos and a growth team measured on self-serve revenue will compete for the same accounts and route around each other. And keep pricing, packaging and billing under one owner, since expansion stalls in the handoffs between teams that each control part of the offer.

One caveat belongs here because it is the most common misreading of this pattern. A land-and-expand motion that works does not mean a go-to-market that is efficient. Product-led companies that move upmarket often report strong enterprise net revenue retention alongside sales and marketing spend that consumes a large share of revenue, because expanding enterprise accounts is expensive work even when the product does much of the persuading. Check the filings of any public company in your category before assuming the two move together. Product-led growth is supposed to improve that ratio, not disguise it, so watch what you spend to expand alongside what expansion earns.

Net revenue retention hides more than it shows

One percentage is the headline number for this pattern, and it is also where the self-deception lives, because it conceals everything that would tell you what to do next.

1. Decompose NRR into its four sources: seats, usage, tier and price. Expansion driven by price increases is not the same business as expansion driven by adoption, and only one of them compounds. Split it before you argue about it.
2. How many distinct teams are active inside your largest accounts? This is the consolidation-survival number. Compare it against seat count as a renewal predictor in your own data, because one of the two is usually much more informative and it is rarely the one on the dashboard.
3. What share of expansion happened with no human involved? That fraction is the honest measure of how product-led your expansion actually is, as opposed to how product-led your acquisition is.
4. How long from landing to first expansion, and is it shortening? A stable long gap usually means expansion is being discovered by accident rather than designed.

Published NRR benchmarks exist and are widely misused. Benchmarkit put median net revenue retention for private B2B software at about 106% in its 2025 data and nearer 101% in 2026.¹ So the commonly repeated “healthy is above 120%” bar sits well above the median: it describes an unusually good result while being presented as a minimum, and a company at the median reads its own dashboard as a failure.

Arm the champion

If your usage is growing and your revenue is not, build the forwardable artifact first: that is very likely the constraint. Publish your security answers where a stranger can find them without contacting you, which unblocks deals nobody has told you are blocked.

Then read your own accounts properly. Decompose NRR into its four sources, and count teams rather than seats inside your largest ones, treating single-team concentration as the risk it is instead of the success it resembles. Pull your last twenty expansions and find what they had in common thirty days beforehand: that shared behavior is your expansion signal, it is rarely what anyone guessed, and it becomes the usage threshold at which a human gets involved, so the product decides the timing and not the calendar.

The last two require talking to people. Call whichever account grew most last quarter and ask how they went from one team to several, because the answer describes the expansion motion you actually have rather than the one on your slides. Then ask your last three lost expansions what they consolidated onto. It beats any win-loss survey, and the answer is usually not a direct competitor.

So the revenue half of the flywheel comes down to six decisions: whether a stranger can start at all, what they get for nothing, whether taking it away works better than withholding it, when to ask, what unit to charge for, and who signs when the account outgrows one team.

All six run on customers willing to bring other people in, and that willingness has been treated here as something that happens. It does not have to be left to chance. It can be built, paid for, and made worth someone’s while, which is the last stage of the wheel and the one most companies never design on purpose: two-sided referrals make the introduction someone else’s idea.

Updated August 2026