Marcelo Lima

Investor and analyst at Heller House

Mentions

Appeared in

Discussed in

Key positions and views

The market's 'SaaSpocalypse' narrative is an overreaction, creating a significant buying opportunity in high-quality SaaS companies trading at historically low free cash flow multiples.

The true moats of enterprise software are not based on code but on trust, governance, regulatory compliance, and established sales motions, which AI startups cannot easily replicate.

AI model providers like OpenAI and Anthropic will function as an infrastructure layer, partnering with and enabling SaaS companies rather than competing with them at the application level.

The next major computing paradigm will be agentic, with operating systems like Windows and Mac shipping with built-in local AI models and frameworks that will transform user interaction.

Large platforms like Salesforce are effectively future-proofing themselves by adopting headless API strategies, which decouples their powerful backend from clunky native UIs and allows for flexible, modern interfaces.

Podcast consensus on Lima

Points of consensus

▶Lima's analysis consistently posits that incumbent enterprise SaaS companies like Salesforce, ServiceNow, and Adobe have durable moats built on trust, governance, and liability management, which are not easily eroded by AI-driven code generation (Claims 2, 6, 12).May 2026

▶He consistently views the market's recent sentiment toward SaaS as overly pessimistic, citing a major sell-off in January 2024 and valuations that imply zero growth, creating a disconnect with the underlying fundamentals of strong companies (Claims 7, 9, 16).May 2026

▶A recurring point is that AI will function as an enabling infrastructure layer, augmenting existing software platforms rather than replacing them, a view he supports with statements from AI companies like Anthropic (Claims 10, 11).May 2026

Points of debate

▶Lima highlights the extreme negative market sentiment that SaaS is "uninvestable" (Claim 16) while simultaneously arguing that many of these same companies are fundamentally strong and trading at attractive, deeply discounted valuations (Claim 9), creating a tension between market perception and his fundamental analysis.

▶He points to AI coding tools increasing the capabilities of software engineers (Claim 5) but contrasts this with the catastrophic failure of Volkswagen's internal software project, CARIAD, suggesting that access to better tools does not solve for a lack of core competency in software development (Claim 19).May 2026

▶Lima describes the native Salesforce UI as "ugly and clunky" (Claim 3) yet also details how Salesforce is a highly sophisticated and evolving platform that is overcoming these limitations through advanced headless APIs and ecosystem integrations like Slack and Tableau (Claims 3, 8, 18).May 2026

Key themes

▶The 'SaaSpocalypse' Narrative vs. RealityMay 2026

Lima identifies a powerful market narrative from early 2024 that painted the SaaS sector as "uninvestable," evidenced by a 10-to-1 selling-to-buying ratio and a sharp rotation into semiconductor stocks. He contrasts this sentiment with his analysis that many high-quality SaaS companies are trading at or below 10x free cash flow, a valuation that ignores their fundamental strengths and growth potential.

This theme suggests a significant contrarian investment opportunity for those who can look past prevailing market fear and focus on the durable business models and discounted valuations of established SaaS leaders.

▶Durable Moats of Enterprise SoftwareMay 2026

According to Lima, the primary barriers to entry in enterprise software are not the cost or complexity of writing code, but rather the deep-seated requirements of trust, governance, regulatory compliance (e.g., SOC 2), and liability management. Incumbents like SAP and ServiceNow provide SLAs and guaranteed uptime, which are critical for enterprises and difficult for new entrants to replicate.

Investors should be skeptical of claims that AI startups will rapidly displace enterprise incumbents, as the competitive moats are less about technology and more about established business processes and trust.

▶AI as an Infrastructure LayerMay 2026

Lima predicts that AI model providers like OpenAI and Anthropic will not compete directly at the application layer with companies like ServiceNow or SAP. Instead, they will operate as a foundational infrastructure layer, similar to AWS, providing the core technology that SaaS companies then build upon and integrate into their specialized products.

This suggests a symbiotic future where both AI platform providers and SaaS application companies can thrive, rather than a zero-sum game, which may de-risk investments in incumbent SaaS companies threatened by AI disruption.

▶The Agentic and Headless Future of InterfacesMay 2026

Lima highlights a technical shift away from traditional graphical user interfaces. He points to Salesforce's headless API strategy, which allows for custom UIs, and predicts that operating systems will soon ship with built-in, local AI models and agentic frameworks, fundamentally changing how users interact with computers.

This points to a future where value shifts from the visual front-end to the underlying data, logic, and agentic capabilities, favoring platforms with robust, accessible APIs and integrated data ecosystems.

Source episodes

Sentiment over time

Not enough data for timeline

Changes over time

June 2019

Cites a conversation with Slack co-founder Stewart Butterfield, who stated he interacted with Salesforce entirely through slash commands, foreshadowing the move away from traditional UIs (Claim 18).

Late 2023

Notes that ServiceNow's ProPlus SKU, which incorporates its new AI features, became generally available, marking a key milestone in incumbent AI integration (Claim 17).

January 2024

Identifies a significant market event where SaaS stocks declined sharply while semiconductor stocks rose, accompanied by reports of a 10-to-1 sell-to-buy ratio and sentiment that the sector was "uninvestable" (Claims 7, 16).

Late February 2024

References an Anthropic enterprise event where the company positioned itself as an infrastructure partner for SaaS companies, supporting his thesis that AI models will not compete at the application layer (Quote).

Suggested prompts

How does Marcelo Lima's thesis on AI as an infrastructure layer account for potential vertical integration by model providers in the future? &nearr;What specific metrics or signals would invalidate Lima's contrarian bullish stance on undervalued SaaS companies? &nearr;Beyond trust and governance, what other moats could protect enterprise incumbents from disruption by AI-native startups? &nearr;What are the primary technical and business hurdles for implementing the OS-level agentic frameworks that Lima predicts? &nearr;

Key concepts

SaaS 8 ep AI Integration 7 ep Salesforce 5 ep Enterprise Software 4 ep ServiceNow 4 ep Valuations 3 ep Agentic Frameworks 2 ep Headless APIs 2 ep Atlassian 2 ep Market Sentiment 2 ep

Notable quotes

“And we heard even at the end of January from the Jefferies Trading Desk get me out at any price 10 to one volume selling versus buying and software names these things are uninvestable”

Marcelo Lima · Investing in the SaaSpocalypse with Heller House's Marcelo Lima

“I think very soon you're going to see both Windows and Mac, they're going to ship, the operating system is going to have open claw built in, and you're going to have a a model running locally built in and everything is going to be agentic on your computer and on your operating system.”

Marcelo Lima · Investing in the SaaSpocalypse with Heller House's Marcelo Lima

“to me, it makes a lot more sense that these model companies are going to be more at the infrastructure layer rather than trying to take over applications and compete directly with ServiceNow or SAP, etc.”

Marcelo Lima · Investing in the SaaSpocalypse with Heller House's Marcelo Lima

“What Salesforce did is they exposed all of the functionality on their platform, whether you're talking about Slack or Tableau or MuleSoft, etc., in a headless fashion through model context protocol interfaces”

Marcelo Lima · Investing in the SaaSpocalypse with Heller House's Marcelo Lima

Report last updated: May 5, 2026

Get started free

Back to Entities Intelligence Report