The Vertical Software Selloff: How LLMs Are Redefining Industry-Specific Software

In recent weeks, nearly $1 trillion has been wiped from software and services stocks. FactSet dropped from a $20 billion peak to under $8 billion. S&P Global lost 30% in weeks. Thomson Reuters shed almost half its market cap in a year. The S&P 500 Software & Services Index, comprising 140 companies, fell 20% year to date.

This dramatic selloff has left many wondering what's fundamentally changed about vertical software—the specialized industry-specific applications like Bloomberg for finance, LexisNexis for legal, and Epic for healthcare that have historically commanded premium pricing and customer loyalty.

Nicolas Bustamante, founder of Doctrine (now Europe's largest legal information platform) and Fintool (an AI-powered equity research platform), has been on both sides of this disruption. With a decade of experience building vertical SaaS and now developing AI-powered alternatives, Bustamante offers a unique perspective on why the market is revaluing these companies and what comes next.

The Ten Moats of Vertical Software

Vertical software companies have historically been protected by what Bustamante identifies as ten distinct moats. LLMs are systematically dismantling some while leaving others intact, creating a fundamental shift in what makes these companies valuable.

1. Learned Interfaces → Destroyed

A Bloomberg Terminal user has spent years learning keyboard shortcuts, function codes, and navigation patterns. These aren't intuitive—they're a language. Once fluent, switching platforms means becoming illiterate again. "We're a FactSet shop," "We're a Lexis firm," "We're a Bloomberg house"—these statements reflect software muscle memory, not just data quality.

At Doctrine, Bustamante experienced firsthand how interface mastery created switching costs. "We had a team of designers and a small army of customer success managers whose entire job was onboarding lawyers onto our interface," he explains. "Every UI change was a project: user research, design sprints, careful rollouts, handholding."

At Fintool, this entire cost center disappears. "Our users type what they want in plain English and get an answer. There is no interface to learn because it's all chat," Bustamante notes. "That entire cost center, the designers, the CSMs, the UI change management, it just doesn't exist."

2. Custom Workflows and Business Logic → Vaporized

Vertical software encodes how an industry actually works. A legal research platform doesn't just store case law—it encodes citational networks, Shepardize signals, and litigation workflows. Building this business logic took years of development from engineers who understood both the domain and technology.

"At Doctrine, building a legal research workflow took a team of engineers and legal experts over several years," Bustamante recalls. "The business logic was spread across thousands of lines of Python, custom ranking algorithms, and hand-tuned relevance models."

At Fintool, the same DCF valuation methodology is encoded in a markdown file that took a week to write. "A portfolio manager who's done 500 DCF valuations can encode their entire methodology without writing a single line of code," Bustamante explains. "Years of engineering versus one week of writing. That's the shift."

3. Public Data Access → Commoditized

A massive portion of vertical software's value proposition was making hard-to-access data easy to query. Companies like FactSet built thousands of parsers for SEC filings, each requiring maintenance as formats changed.

"At Doctrine, we built NLP pipelines for different case laws: named entity recognition to extract judges, courts, legal concepts. Dedicated ML models to classify decisions by legal domain. Custom parsers for every court, each with its own formatting quirks," Bustamante describes.

At Fintool, none of this infrastructure was necessary. "Frontier models already know how to navigate a 10-K. They understand the difference between GAAP and non-GAAP revenue. They can parse a nested table without being taught the schema," he explains. "The parsing infrastructure that took Doctrine years to build is now a commodity capability that comes free with the model."

4. Talent Scarcity → Inverted

Building vertical software required people who understood both the domain and technology—a rare combination that created natural barriers to entry.

"At Doctrine, hiring was brutal. We didn't just need good engineers. We needed engineers who could understand legal reasoning. These people barely existed," Bustamante recalls. "Every week, we held internal lectures where lawyers taught engineers how the legal system actually worked. It took months before a new engineer was productive."

At Fintool, domain experts write their methodology directly into markdown skill files without needing to learn programming. "The engineering is handled by the model. The domain expertise, which was always the abundant resource, can now become software directly without the engineering bottleneck," Bustamante observes.

5. Bundling → Weakened

Vertical software companies expanded by bundling adjacent capabilities. Bloomberg added messaging, news, analytics, and trading. Each new module increased switching costs.

"At Doctrine, bundling was the growth strategy. We started with case law search, then added legislation, then legal news, then alerts, then document analysis. Each module had its own UI, its own onboarding, its own customer workflows," Bustamante explains.

LLM agents break this bundling moat because the agent itself becomes the bundle. "At Fintool, alerts are a prompt. Watchlists are a prompt. Portfolio screening is a prompt. There's no separate module for each. There's no UI to maintain," Bustamante describes. "Why pay Bloomberg's premium for the entire suite when an agent can cherry-pick the best provider for each capability?"

6. Private and Proprietary Data → Stronger

Some vertical software companies own or license data that doesn't exist anywhere else. Bloomberg collects real-time pricing data from trading desks worldwide. S&P Global owns credit ratings and proprietary analytics.

"If your data genuinely cannot be replicated, LLMs make it MORE valuable, not less," Bustamante argues. "Bloomberg's real-time pricing data from trading desks? Can't be scraped. Can't be synthesized. Can't be licensed from a third party."

The test is simple: Can this data be obtained, licensed, or synthesized by someone else? If no, the moat holds. If yes, the company is at risk of commoditization.

7. Regulatory and Compliance Lock-in → Structural

In healthcare, Epic's dominance isn't just about product quality—it's about HIPAA compliance, FDA certification, and the 18-month implementation cycles that hospitals endure.

"HIPAA doesn't care about LLMs. FDA certification doesn't get easier because GPT-5 exists. SOX compliance requirements don't change because Anthropic released a new plugin," Bustamante notes. "A hospital can't replace Epic with an LLM agent because the LLM agent isn't HIPAA certified, doesn't have the required audit trails, and hasn't been validated by the FDA for clinical decision support."

8. Network Effects → Sticky

Some vertical software becomes more valuable as more industry participants use it. Bloomberg's messaging function is the de facto communication layer for Wall Street.

"LLMs don't break network effects. If anything, they might make communication networks more valuable. The information flowing through these networks becomes training data, context, signal," Bustamante explains.

9. Transaction Embedding → Durable

Some vertical software sits directly in the money flow. Payment processing for restaurants. Loan origination for banks. Claims processing for insurance companies.

"When you're embedded in the transaction, switching means interrupting revenue. Nobody does that voluntarily," Bustamante states. "Stripe isn't threatened by LLMs. Neither is FIS or Fiserv. The transaction processing layer is infrastructure, not interface."

10. System of Record Status → Threatened Long-Term

When your software is the canonical source of truth for critical business data, switching isn't just inconvenient—it's existentially risky.

"AI agents don't just query existing systems. They read your SharePoint, your Outlook, your Slack. They collect data on the user. They write detailed memory files that persist across sessions," Bustamante warns. "Over time, the agent accumulates a richer, more complete picture of a user's work than any single system of record. The agent's memory becomes the new source of truth."

The Net Effect: Barrier to Entry Collapses

Add it all up, and five moats are destroyed or weakened while five remain intact. But the five that break are the ones that kept competitors out.

"Before LLMs, building a credible competitor to Bloomberg or LexisNexis required hundreds of engineers who understand the domain, years of development time, massive data licensing deals, sales teams that can sell to conservative enterprises, and regulatory certifications," Bustamante explains. "The result: most verticals had 2-3 serious competitors."

After LLMs, a small team with frontier model APIs, domain expertise, and good data pipelines can build a product that handles 80% of what a vertical software does within months. Fintool was built by a team of six and now serves hedge funds that previously relied exclusively on Bloomberg and FactSet.

"The critical insight is that competition doesn't increase linearly—it explodes combinatorially," Bustamante warns. "You don't go from 3 incumbents to 4. You go from 3 to 300. And that's what craters pricing power."

This is a Multi-Year Transition, Not an Overnight Collapse

The market may be overestimating the speed of this disruption.

"Enterprise revenue doesn't disappear overnight. FactSet's clients are on multi-year contracts. Bloomberg Terminal contracts are typically 2-year minimums," Bustamante points out. "A $50 billion hedge fund isn't going to rip out S&P Global CapIQ tomorrow because Claude can query SEC filings. They'll evaluate alternatives over 12-18 months."

But here's the thing the market already understands: you don't need revenue to decline for the stock to crash. You need the multiple to compress. A financial data company that traded at 15x revenue when it had pricing power might trade at 6x revenue when the market believes both are eroding.

The Real Threat

The real threat isn't the LLM itself. It's a pincer movement that vertical software incumbents didn't see coming.

"From below, hundreds of AI-native startups are entering every vertical. When building a credible financial data product required 200 engineers and $50M in data licensing, markets naturally consolidated to 3-4 players. When it requires 10 engineers and frontier model APIs, the market fragments violently," Bustamante explains.

From above, horizontal platforms are going deep into vertical territory for the first time. "Microsoft Copilot inside Excel now does AI-powered DCF modeling and financial statement parsing. Copilot inside Word does contract review and case law research," he notes. "The horizontal tool becomes vertical through AI, not through engineering."

What Comes Next

Bustamante sees a bifurcation emerging:

"The companies that survive this transition are the ones that moved from 'we organize public data better' to 'we own data you can't get anywhere else,'" he suggests. "Bloomberg's real-time pricing data from trading desks? Can't be scraped. Can't be synthesized. Can't be licensed from a third party. In an LLM world, this data becomes the scarce input that every agent needs."

For investors and entrepreneurs, the key is understanding which vertical software companies are truly at risk. Bustamante offers a simple framework:

1. Is the data proprietary? If yes, the moat holds. If no, the accessibility layer is collapsing.
2. Is there regulatory lock-in? If yes, LLMs don't change the switching cost equation. If no, switching costs are primarily interface-driven and dissolving.
3. Is the software embedded in the transaction? If yes, LLMs sit on top of you, not instead of you. If no, you're replaceable.

Zero "yes" answers: high risk. One: medium risk. Two or three: you're probably fine.

As Bustamante concludes, "The vertical SaaS reckoning isn't about all vertical software dying. It's about the market finally distinguishing between companies that own something genuinely scarce that is safe from LLM agents."