For most B2B software companies past a certain size, the majority of this year's new revenue will not come from a new customer. Benchmarkit's 2025 B2B SaaS Performance Metrics Benchmarks, built from private B2B SaaS operating data for fiscal 2024, put expansion ARR at 40% of total new ARR at the median, rising to 58% for companies between $50 million and $100 million in ARR and 67% for companies above $100 million. High Alpha's 2025 SaaS Benchmarks, drawn from more than 800 respondents, describes the same shift from the other side: below roughly $20 million in ARR, growth comes mostly from new logos, and above $50 million, roughly 60% of new ARR comes from customers the company already has.

That single fact has quietly rewritten the job. The account executive who used to be measured on logos is now measured on a base, and the stage that used to sit as an afterthought at the end of the pipeline, expansion, is where most of the number lives. What has not changed is how most teams run it. Expansion is still treated as new business with the hard parts removed: same product, warm relationship, paperwork already signed, shorter cycle. The evidence does not support that reading. Expansion is a complete sale run under conditions that make disciplined selling harder rather than easier, because the customer already keeps a scorecard on you, the internal politics are settled, and your ask competes with money the customer has already spent.

Start with how expansion buyers actually feel about what they buy. In a Gartner survey presented in June 2023, 60% of technology buyers involved in decisions to renew or expand "as-a-service" agreements said they regret nearly every purchase they make, a 6 point increase over 2020. Consider who those people are. Not cold prospects. Not first-time buyers without a reference point. These are people inside an existing vendor relationship, buying more of something they already own, reporting regret at a level that would be treated as a crisis if it turned up in new business.

The programs built to capture that revenue do not fare much better. Schmitz, Lee and Lilien, writing in the Journal of Marketing in 2014, open from the finding that roughly three quarters of all cross-selling initiatives fail, and that they fail for sales force reasons rather than product or market reasons. Their own study, using matched company records plus surveys of salespeople and sales managers inside a biotech firm, found that cross-selling performance depended on whether supervisory behavior and compensation design actually supported the adoption behavior being asked of reps. Cross-selling is not a product decision that the sales team then executes. It is a selling problem that the product merely makes possible.

Then there is the question of whether the expansion revenue is worth having at all. Shah, Kumar, Qu and Chen studied cross-buying across five firms in both consumer and business markets and published the results in the Journal of Marketing in 2012. Between 10% and 35% of customers who cross-bought were unprofitable, and those customers accounted for between 39% and 88% of the firms' total losses from customers. The unprofitable cross-buyers were not randomly distributed. They clustered into recognizable behaviors: customers who demanded disproportionate service, customers who reversed revenue through returns and credits, customers who moved only on deep discounts, and customers with a hard ceiling on what they could spend. When those customers bought more, the losses grew rather than shrank.

None of this should surprise anyone who read Reinartz and Kumar's analysis of 16,000 customers across four company databases, published in Harvard Business Review in 2002, which found the link between customer loyalty and customer profitability far weaker than the industry assumes. Long tenure does not imply high margin. Neither does a larger contract.

Precision matters here, because two different numbers get discussed as if they were one. Gross revenue retention measures what survives of the existing base after churn and downgrades, and it cannot exceed 100%. Net revenue retention adds expansion on top of that base. In Benchmarkit's 2024 data, the median company posted 88% gross revenue retention (down from 90% across three years) and 101% net revenue retention. Read together, those two numbers say something uncomfortable: at the median, expansion is not funding growth. It is refilling a leaking bucket, and only just. The market has noticed. In Gainsight's Customer Success Index 2025, produced with Benchmarkit from more than 400 companies and customer success leaders, 55% named gross revenue retention a top measurement and 76% named customer retention a primary revenue metric, a deliberate move away from letting net retention hide the churn underneath it.

So here is a working definition the Sales Agent can reuse.

An expansion opportunity exists when a customer that has already realized measurable value from what it bought has a named, currently funded problem your next offering solves, and a decision path that is live for that new problem.

Everything else is a quota wish with a customer's name attached.

Four conditions carry that definition into practice. Following the qualification rule established earlier in this series, that every claim carries a source, a date and a falsifier, an expansion is warranted only when all four hold and each one has evidence behind it.

1. Realized value, stated in the customer's numbers. Not usage. Not logins, seats provisioned, or a health score the vendor calculated for itself. The test is whether someone on the customer side can state, without your prompting, what changed because of the original purchase and roughly what it was worth. The falsifier is blunt: if you cannot name the person who would say it and the number they would use, the value has not been realized, it has been asserted. Expansion built on asserted value inherits the regret rate in the Gartner data, because the customer is being asked to double a bet it has not yet been shown it won.

2. A job the customer already names. Cross-sell fails at the rate Schmitz and colleagues report partly because it begins from the vendor's catalog rather than the customer's agenda. The test is whether the problem your expansion addresses appears in the customer's own language, in their planning documents, board updates or stated team priorities, before you introduce the product. The falsifier is discovering that the only place the problem exists is your account plan.

3. A decision path that exists for this purchase. The most common and most expensive expansion error is assuming the original approval transfers. It rarely does. Expansions routinely need a new budget line, a different owner, a fresh security or procurement review, and a buying group that may barely overlap with the one that signed the first contract. The test is that you can name the budget, the approver and the process for this specific purchase, with dates attached. The falsifier is a champion who says "leave it with me" and cannot describe who else has to say yes.

4. Account economics that survive the addition. This is the condition almost nobody writes down, and the cross-buying research says it matters most. Before pursuing an expansion, ask what the account will cost to serve once it closes. If the additional revenue arrives with support load, custom work, discounting that resets the renewal price, or a pattern of credits and reversals, the expansion can make the account worse while making the quarter better. The falsifier is an honest look at service tickets, discount history and credit notes across the last two renewals.

An expansion that clears all four is a real opportunity and deserves a real plan. One that clears three is a project, not a forecast entry. One that clears fewer is a renewal conversation wearing a growth costume.

Re-run discovery, do not recycle it. Familiarity is the enemy in this stage. The seller believes they know the account because they know the original deal, and skips the work of establishing what is true now. Treat an expansion like a new deal in one specific respect: the gap, the cost of inaction, the decision path and the risk of change all have to be re-established for the new problem, with new evidence and new dates.

Stop reading satisfaction as purchase intent. The Wallet Allocation Rule research by Keiningham, Aksoy, Williams and Buoye makes an argument most account teams have not absorbed. What predicts share of wallet is the rank a customer assigns you relative to the other options it uses, not how satisfied it reports being. Satisfaction on its own explains a trivial share of category spending. A customer can be genuinely happy with you and still give the next dollar to the vendor it ranks first for that specific job. The practical consequence: a strong survey score or a warm quarterly review is not evidence for condition one or condition two. Ask what else the customer uses for adjacent jobs, and where you rank for the job you want to expand into.

Remove the friction you control. In the same Gartner research, 95% of technology buyers said they would have preferred a fully digital experience for their expansion purchase. That is close to unanimous, and it reads as an instruction. Pricing that requires a call to decode, order forms that restart legal review, and quotes that take a week are not neutral, they are a tax on the growth motion. More recent Gartner work, showing that most B2B buyers now prefer a rep-free experience while still turning to sellers to validate what they found on their own, tells you where seller time belongs: not gating information, but validating a decision the customer has already started making.

Rebuild the buying group before you need it. The people who bought the original contract may no longer be there. Roles turn over on a shorter cycle than multi-year agreements run, and the champion who understood why you were chosen is often the single point of failure in both renewal and expansion. Breadth of contact is the mitigation, and it has to exist before the expansion conversation opens, not after.

Time expansion to value, not to the quarter. The temptation is to open the expansion when the seller needs it. The evidence says to open it when the first condition is provably true. An expansion pushed before realized value converts a satisfied customer into a skeptical one, and the regret data suggests the effect compounds across the relationship.

Be willing to decline. Given the cross-buying findings, the highest-return decision in some accounts is not to expand: fix the economics, reduce the service load, or let the account stabilize before adding to it. Declining an expansion is a legitimate account planning outcome and should be sayable out loud in a pipeline review.

Track gross and net revenue retention separately and never quote net alone. Break expansion down by source (more seats, more usage, a new product, a price increase), because those four have completely different repeatability and only some of them survive a downturn. Report the share of expansion opportunities that had a documented realized-value baseline before the opportunity was created, which is the leading indicator for everything above. And keep one post-close measure: the rate at which expansions are downgraded, credited or reversed within two renewal cycles, which is the only honest read on whether the growth was real.

Expansion is where the two halves of a revenue organization make contact. It is a sale, so it needs qualification, discovery, a decision path and a business case, exactly like a new logo. It is also the accumulated result of every promise made during the first sale, and it cannot outrun a value story that was never delivered. The companies that get this right are not better at asking for more. They are better at being able to prove, on the customer's terms and in the customer's numbers, that the last thing they sold worked.