SaaS
What One SaaS Sale Reveals About AI Defensibility
ZyraTalk drew interest from strategic buyers but no private-equity offer. Einar Vollset explains why revenue durability became the harder question.
Listen to the audio analysis
Create your ownWritten for builders and creators.
High growth and high retention used to be the golden ticket for a SaaS exit, but in the age of AI, they might not be enough to save your valuation.
Editor's note
Why this matters now
ZyraTalk sold an AI receptionist for HVAC firms and other service businesses. When Discretion Capital took it to market, it had strong growth, retention and integrations, according to adviser Einar Vollset. More than 20 management meetings with private equity firms produced no offer. Strategic buyers competed for it instead.
For a SaaS founder thinking about an eventual sale, the distinction is revealing. In this July 2026 conversation with Rob Walling, Vollset describes buyers asking what would protect revenue if a competitor could reproduce the software. His 5 proposed sources of defensibility explain what that question reaches beyond the product's features.
The source
What it says
Distilled from the original. The notes above and below are the editor's own.
Strong Metrics Did Not Answer Every Buyer's Question
Vollset recalls 22 or 23 private-equity management meetings during the ZyraTalk process and no letters of intent from that group. He says strategic buyers took a different view, and the company was acquired by EverCommerce.
The outcome matters to the argument. ZyraTalk was not a company nobody wanted. Different buyers could value the same business differently. A strategic acquirer may have a reason to combine the product with its existing customers or operations that a financial buyer does not share.
Vollset's account emphasizes uncertainty about future revenue. Growth and retention describe how the business has performed. They do not, by themselves, explain what would stop a new competitor winning customers with similar software.
This is an adviser's account of a sale, not a market-wide study or a rule that all private-equity firms use the same threshold. His discussion focuses on smaller SaaS businesses, including the roughly $1M to $20M ARR range. It supplies no universal valuation premium for any particular moat.
1. Hardware Can Make Replacement a Physical Job
Vollset points to software coupled with hardware, such as grocery-store scales or EV chargers. A competitor recreating a screen has not necessarily reproduced the physical installation or the way the software works with it.
The strength of this protection depends on the integration. Proprietary equipment may make substitution difficult. An interchangeable device with an open interface offers less protection. Hardware also brings its own delivery and support work, so the example is an explanation of a possible barrier rather than a reason for every SaaS company to add a device.
2. A Marketplace Needs Participants as Well as Code
A two-sided marketplace depends on having enough relevant buyers and sellers. Rebuilding the website does not bring either group with it.
Vollset sees that network as a possible source of durability, while warning how hard it is to create in the first place. A founder with access to one side has a different starting position from someone who must recruit both. Neither position guarantees that participants will stay or that a rival cannot enter.
The distinction for a buyer is whether the business has an active network that makes the service useful. A marketplace-shaped product without those participants has not established the same advantage.
3. Shared Work Makes a System Harder to Replace
The system-of-record argument concerns the information and habits that accumulate around a product. Messages, approvals, business history and team routines can make a switch more involved than moving a feature list.
Vollset uses familiar business tools such as Slack and QuickBooks to illustrate that dependence. A replacement has to support the work people already coordinate through the system, including the context they need to keep doing it.
For a founder, this raises a more useful question than whether a competitor can draw the same interface: what would customers have to move, reconnect or relearn to keep their operation working?
4. A Continuing Data Source Is Different From a Snapshot
Vollset distinguishes a static dataset from information that keeps arriving and changing. A rival may obtain a snapshot without being able to reproduce the continuing collection process.
His examples include services built around specialized business or financial data. The possible advantage comes from sustained access, coverage and freshness. It does not follow that every dataset is defensible or that making customer data harder to export creates a durable business.
The buyer's question is what allows this company to keep supplying useful information that a competitor cannot readily obtain on similar terms.
5. Accountability Can Be Part of the Product
A business using software for consequential operations needs a way to resolve failures. Vollset argues that an established provider's reputation and support can therefore matter even when a new tool reproduces much of the functionality.
That is a concrete part of the offer: somebody takes responsibility for fixing the problem and keeping the service dependable. It requires delivery over time. A familiar name or a support email address alone does not prove the service will meet that standard.
The interview does not quantify what customers will pay for this assurance. It explains why a feature comparison may omit something a business buyer values.
Editor's note
What to do with this
Vollset's framework is useful for describing what a customer would have to replace alongside the software: equipment, a network, shared work, continuing information or a trusted operating relationship. Those are different sources of durability, with different costs to build and maintain.
ZyraTalk still found a buyer. The concerns that kept private equity firms from making an offer did not prevent strategic buyers from competing for the business. For the product side of the same question, read Walling on AI features that finish useful user work. Useful functionality can earn adoption. The sale discussion asks what helps that revenue endure.
The original
What SaaS Buyers Actually Want in 2026
Rob Walling · 19 July 2026
Read next
SaaS
Adding AI to SaaS: Start With the User’s Unfinished Work
Rob Walling’s eBay example shows why generating text can leave the hard work untouched. His 6 patterns help locate a more useful AI feature.
4 min read9 min listen
SaaS
Why a SaaS Price Cut Can Bring More Churn
Rob Walling examines a founder’s price cut and shorter customer lifetimes, then separates testing a higher price from selling to a different buyer.
4 min read7 min listen