How Much Growth Is Actually Coming From the Existing Base
More than 60% of G2’s own gross bookings come from existing customers, a figure G2’s own head of sales disclosed in a podcast excerpt cited by Gainsight. Private B2B SaaS net revenue retention independently clusters between 101% and 104% across three named benchmark providers, KeyBanc Capital Markets and Sapphire Ventures, SaaS Capital, and a corroborating analysis, meaning the typical company’s existing base is already, on its own, generating more revenue than it started the year with.
Both figures point the same direction from different angles: a mature SaaS company’s growth increasingly runs through the accounts it already has, not only the ones still working through a first sales cycle.
What an In-App Prompt Can Do Well
A well-placed in-app prompt is genuinely good at low-friction, self-serve decisions, a single extra seat, a small usage-limit bump, anything a user can evaluate and approve alone without needing anyone else’s sign-off. For that category of decision, a prompt is faster than a call for everyone involved and should stay exactly where it is.
The mistake is not building in-app prompts. It is assuming the same mechanism scales up to every expansion decision, regardless of size or complexity.
What an In-App Prompt Cannot Do
This is reasoning, not a cited statistic. A prompt cannot answer a question about budget-cycle timing, whether now is actually a workable moment for this specific account to add spend, or read hesitation the way a real conversation can. It cannot make the internal case to a buyer’s own boss the way a rep-assisted conversation, armed with real usage data and a specific outcome the account has already gotten, can.
A genuine tier jump or a multi-year commitment usually involves more than one person on the buyer’s side, and a static in-app prompt has no way to identify or reach the second person who also needs to be convinced.
The Dollar Gap Between an Average Account and a Well-Retained One
Companies with net revenue retention of 120% or higher command a median annual contract value of $61,802, more than double the $26,269 median for companies below that line, according to SaaS Capital’s 2026 survey of more than 1,000 private SaaS companies. That much revenue difference does not deserve a worse channel than the one that closed the original deal, and for a decision of that size, an in-app prompt, however well designed, is a worse channel than a real conversation.
The accounts worth the most are exactly the ones where the case for a human touch is strongest, not weakest, since they are also the accounts with the most complexity, the most stakeholders, and the most at stake in getting the decision right.
Where the Line Actually Belongs
Practitioner guidance, not a cited statistic: a small, single-seat or minor usage-tier add is a reasonable candidate to stay entirely in-app. A genuine plan-tier jump, a multi-year commitment, or any expansion that involves a second stakeholder the original user cannot approve alone, is a reasonable line for where a human conversation should take over instead.
Drawing that line by decision size and stakeholder count, rather than by a blanket rule that treats every expansion the same way, keeps the self-serve motion fast for what it is good at while reserving a real conversation for what it is actually needed for.
Why This Is Not an Argument Against Self-Serve Expansion
None of the above argues for removing in-app prompts or slowing down the self-serve motion that works well for smaller decisions. It argues against treating a static prompt as sufficient for every expansion decision regardless of size, which is a narrower, more specific claim than a blanket case against self-serve.
Human + AI SDRs can run the human side of that split over SMS, reaching out on the larger, higher-stakes expansion moments while leaving the smaller, self-serve decisions exactly where they already work well.
Sources
The external data in this article draws on the sources below. Figures described in the text as estimates or industry triangulations are directional and are not attributed to a single dataset.
- Gainsight, Why Your Customers Are Already Shopping Around (And How to Catch It Early)
- SaaS Capital, What Is the Average Deal Size for Private SaaS Companies
- KeyBanc Capital Markets and Sapphire Ventures, 16th Annual Private Company SaaS Survey (via PR Newswire)