No Right to Exist

The concession

Robert F. Smith, chief executive of Vista Equity Partners, names a third category of enterprise software after describing two categories his firm is actively converting: "if you are basically repurposing data and information that has been available in the public marketplace and selling that, or reselling it, content, you'll end up in [category three], and there's a lot of those...it will eat those companies. No, no question about it."¹

The definition is narrow, which is what makes it useful. The condemned business has three properties at once: its input is public data anyone could always have accessed; its value-add is aggregation, formatting, or interpretation, the work of finding, cleaning, arranging, and explaining; and it owns no workflow, selling the compiled artifact rather than doing anything with it. That combination was a viable business for decades because assembly itself was expensive, someone had to gather the filings, normalize the categories, and write the summary, and the customer was paying for saved labor rather than for secret information. A model that reads everything and assembles on demand removes exactly that cost, which is why this category collapses while workflow-owning categories do not. The complement is the real point: what survives is sovereignty and dominion over workflows and data, and category three simply fails that test on both axes at once, the data is not proprietary and there is no workflow to act inside.

Why the concession matters

Every other statement of a coming SaaS apocalypse in this material comes from someone positioned to benefit from it: Ben Horowitz funding the replacements, Ryan Petersen using the threat as procurement leverage, Brad Gerstner having avoided the sector as an investor. Smith concedes the same dynamic, unprompted, about a category his own asset class contains, with no hedge at all: "no question about it."¹ A concession against interest is the strongest form of evidence available inside a promotional interview, and it is the reason the rest of his three-category framework is worth taking seriously rather than dismissing as talking a book.

The framing also converts an unfalsifiable industry debate into a sorting exercise. "Will AI kill SaaS" has no answer. "Is this company's data public, and does it own a workflow" has one, and can be asked of a specific line item on a specific balance sheet.

What is left unresolved

Smith names no companies directly, leaving the category as a description rather than a list, though the shape matches recognizable businesses: market-research and industry-report publishers, regulatory and compliance-content vendors, aggregated news products, contact-database vendors, and the thinner tier of vertical software whose product is essentially a formatted view of public filings. The category boundary often runs inside a single company rather than between companies, since a vendor can hold genuine dominion over one workflow while running adjacent content products that are pure category three. "Publicly available" also does unexamined work in the definition: data can be technically public and still practically unobtainable at scale, paywalled, rate-limited, or historically accumulated in a way a new entrant could not reconstruct, which is exactly the kind of business that survives despite public inputs. And distribution habits, procurement cycles, compliance requirements, and citation conventions, let businesses in this category persist far longer than the disruption logic predicts; incumbency has outlasted structural disruption before. No company from Smith's own roughly ninety-company portfolio is placed in the category either, which is the obvious question left unasked in the interview that produced the claim.¹