Available 2025 data show vertical SaaS has generally been one of the stronger parts of software, not among the worst-performing, and there is no clear evidence that AI‑driven in‑house tools and pricing compression uniquely damaged its valuations or revenues.

Key points:

• Valuations and relative performance. Multiple independent valuation and M&A analyses in 2025 find that vertical SaaS companies still trade at a premium to horizontal SaaS and are viewed as especially attractive:

  • AGC Partners’ early‑2025 report shows vertical SaaS trading at about 7.0× EV/Revenue vs. 4.8× for horizontal SaaS, with similarly strong growth and margins, implying investors pay up for vertical focus rather than shun it. (agcpartners.com)
  • A Baker Tilly / AInvest summary of Q3 2025 data likewise reports vertical SaaS firms at 7.0× EV/revenue vs. 4.8× for horizontal, explicitly saying vertical SaaS companies "continue to outperform horizontal peers." (ainvest.com)
  • A 2025 valuation overview notes that AI‑first and vertical SaaS are the most sought‑after categories, often rewarded with 8–12× revenue multiples, while broader horizontal platforms without clear ROI sit around 3–5×. (ful.io)
  • A broad SaaS metrics benchmark (2,000+ companies) highlights "Vertical SaaS outperforming": vertical players are 1.5–3.3× more likely to be outlier winners and account for 54% of Q3 2025 SaaS M&A, with the conclusion "Vertical‑first wins." (rockingweb.com.au)
  • A July 2025 sector note from JM Financial contrasts vertical vs. horizontal SaaS since the 2021 peak: horizontal SaaS endured sharper multiple compression, while vertical SaaS names like AppFolio, Procore, Veeva, Samsara, and Guidewire generally maintained 5–15× EV/Sales, with "more orderly and muted de‑rating" and structurally stronger pricing power due to sticky, mission‑critical workflows. (fr.scribd.com)

• Stock and growth benchmarks do not show vertical SaaS as a bottom tier.

  • A First Analysis quarterly review does show that in the June 2025 quarter, the vertical SaaS group’s shares rose only about 5% vs. 13% for its broader SaaS universe, lagging categories like cybersecurity and data visibility. That is mild underperformance for that quarter, not "worst‑performing software assets" for the year. (firstanalysis.com)
  • The Software Equity Group’s SEG SaaS Index reports that "Vertically Focused" public SaaS names saw EV/revenue medians fall from 4.5× (3Q24) to 3.2× (3Q25) while the overall index went from 5.9× to 5.3×, indicating some relative multiple compression—but again, several other categories (e.g., analytics, sales & marketing) also sit in the low‑multiple range, so vertical is not uniquely or clearly the "worst" segment. (softwareequity.com)
  • By contrast, other 2025 performance roundups emphasize that industry‑focused, vertical platforms are among the winners: a 1H’25 SaaS stock review notes that industry‑focused platforms like Veeva and Guidewire outperformed horizontal solutions, and calls out "Vertical Specialization Advantage" as a key driver. (linkedin.com)
  • A separate SaaStr analysis, "The Vertical SaaS Gold Rush," shows many vertical or non‑tech B2B names (Samsara, ServiceTitan, Toast, Shopify, etc.) growing 2–3× faster than generic enterprise SaaS; most horizontal incumbents like Salesforce are in single‑digit growth. (saastr.com)
  • Multiple benchmark summaries (e.g., OpenView/WeBuildSaaS, LinkedIn analyses) show median growth of ~45% for vertical SaaS vs. ~28% for horizontal, with top‑quartile verticals at ~100%+ YoY growth—again describing vertical SaaS as "crushing" horizontals, not as a laggard. (webuildsaas.com)

• AI’s effect: more tailwind than systemic damage to vertical SaaS.

  • A 2025 vertical‑SMB SaaS benchmark from Tidemark concludes that "Vertical SaaS has arrived" and that AI is now a revenue driver, with companies that adopt AI in their products and GTM seeing stronger growth, rather than mass churn to in‑house tools. (tidemarkcap.com)
  • Coverage of the "third wave" of vertical SaaS (cloud + fintech + AI) from investors like a16z describes AI as expanding vertical SaaS TAM by automating labor‑intensive workflows and enabling higher net revenue retention—again positioning AI as a core growth catalyst for vertical SaaS firms such as Toast, not as a force that is broadly eroding their revenue base. (yourstory.com)
  • Two detailed essays from Reformation Partners (Jan 2024 and Nov 2025) explicitly rebut Friedberg’s thesis that internal AI‑built tools would make vertical SaaS "doomed." They argue, with examples from construction, hospitality, legal, and other sectors, that most vertical SaaS customers lack the in‑house engineering depth to build and maintain comparable tools; that internal builds carry higher ongoing risk and cost than simply buying the vertical standard; and that AI productivity gains accrue to vertical SaaS vendors themselves, allowing them to ship better products or cut prices while preserving margins. The author concludes, nearly two years into the AI boom, that the original “vertical SaaS is dead” prediction has not been borne out. (medium.com)
  • More general commentary on AI and SaaS economics in 2025 acknowledges some margin and pricing pressure across SaaS as AI features become table stakes, but does not single out vertical SaaS as uniquely impaired; instead, it typically notes that commoditization risk is greatest for generic horizontal tools, while domain‑specific, data‑rich platforms retain differentiation. (implicator.ai)

Putting this together:

• Vertical SaaS did not become "among the worst‑performing software assets" in 2025. Across valuations, M&A share, and growth metrics, it is more often described as outperforming horizontal SaaS and as a favored category for investors.
• While there is some evidence of moderate multiple compression and a few quarters of relative underperformance for certain vertical cohorts, that is part of a broader SaaS repricing and does not match the severity or uniqueness implied by the prediction.
• The specific causal story—that enterprises would rapidly adopt AI‑driven in‑house tools, causing widespread churn and pricing compression that disproportionately harms vertical SaaS—has not materialized at scale. Available data and investor commentary instead show AI being integrated into vertical SaaS products and business models, largely as a competitive advantage rather than an existential threat.

On balance, Friedberg’s prediction that vertical SaaS would be one of the worst‑performing software assets in 2025 due to AI‑driven in‑house tools and pricing compression is not supported by the observed 2025 data, so it is best classified as wrong.