Do autonomous agents make incumbent software companies and their systems of record more or less valuable?

Varun Mohan · Jun 2, 2025

Hyperscalers have enormous distribution in AI coding yet ship mediocre products because the binding constraint is marshalling talent to execute quickly, not reach

It is hard to assemble talent that learns fast; Windsurf ships a major release every one to two weeks and no large company learns at that pace

Scope: specific to the AI coding category

24:46 20VC: Windsurf Founder on Will Model Companies Own the App Layer | Why Moats Do Not Exist in a World of AI | Why the Notion of Single Person $BN Companies is BS | Lovable vs Bolt & Cursor vs Windsurf: How Does it All End with Varun Mohan

Harry Stebbings · Sep 19, 2025 · hedged

Agentforce has not had the success Salesforce wanted it to have — presupposed by the speaker in a question, who adds that Benioff himself would accept this

Scope: claims Benioff himself would accept this

34:49 20VC: Why 90% of Founders Build Startups Wrong | Why AI Growth Rates are Sustainable & Remote Work is BS and the AI Talent War | Competing with Brett Taylor and Sierra: Who Wins the Customer Service War with Jesse Zhang, Decagon

Jesse Zhang · Sep 19, 2025 · hedged

Salesforce's agent efforts likely underperform because their huge customer base pulls them in too many directions and their scale and risk profile prevent them from moving fast.

Too many customers create competing demands, and large companies have risk considerations that slow shipping.

Scope: speaking only from an outside view; declines to judge authoritatively

34:58 20VC: Why 90% of Founders Build Startups Wrong | Why AI Growth Rates are Sustainable & Remote Work is BS and the AI Talent War | Competing with Brett Taylor and Sierra: Who Wins the Customer Service War with Jesse Zhang, Decagon

Harry Stebbings · Apr 20, 2026

No public company has created a genuinely good agent product except Palantir; everyone else ships 60%-quality agents

34:30 20VC: Everyone is Wrong; We Will Have More Developers in Five Years | Why Frontier Labs Will Be Way More Valuable Than They Are Today | Are SaaS Companies Cooked: Which Thrive & Which Die with Aaron Levie, Founder at Box

Jerry Murdock · Feb 28, 2026

Whether autonomous agents make a system of record more or less valuable depends entirely on the management team's execution in capturing the coming trends

Carta shows both branches: if stock tokenization arrives and runs through Carta's cap table, Carta becomes infinitely more valuable; if tokenization bypasses Carta and a new system of record is built, the old one isn't worth much

Scope: he is a long-time Carta investor; both outcomes held as live possibilities

18:27 20VC: Why Cursor is Dead | An AI Tsunami is Coming & You Need to Prepare | Systems of Record Become Valueless Databases with Agents | Is This The End of Tech Private Equity with Jerry Murdock, Co-Founder of Insight Partners

Jerry Murdock · Feb 28, 2026

Divergence in software company value will be driven by management adaptation speed, not by the technology shift itself — companies that put their system-of-record context to work with autonomous agents will increase in value while slow adapters decline

Prices change first out of fear, but the underlying quality of the business then changes based on how well the management team adapts; the tsunami is a wake-up call to move to higher ground rather than get caught on the beach

20:26 20VC: Why Cursor is Dead | An AI Tsunami is Coming & You Need to Prepare | Systems of Record Become Valueless Databases with Agents | Is This The End of Tech Private Equity with Jerry Murdock, Co-Founder of Insight Partners

Eran Zinman · Mar 2, 2026

The public debate about which incumbent has the more robust system of record or the deeper moat is the wrong argument — every company has to change, and speed of change is what matters

Moat arguments are just arguing about who gets to change last; the real question is whether you play defense or offense, and the winners will be those who transform fastest

45:04 20VC: Monday.com CEO on Is SaaS Dead: Will Everything Be Vibe Coded | Will Systems of Record Become Valueless Databases in an Agentic World | Will LLMs Own the Value in the Application Layer with Eran Zinman

Patrick Forquer · May 11, 2026 · hedged

All traditional SaaS companies are under threat right now, but Salesforce is an exception worth respecting because it executed an early, full agentic pivot ahead of other big tech companies

They did a break-glass pivot last year at enormous scale and got there before peers; a survey of hyperscaling AI companies found most of them are using Salesforce

Scope: admits he doesn't know how it will shake out

56:54 20VC: Inside Legora: $100M ARR in 18 Months | Jude Law Generated $50M in Sales Pipeline: The Economics Broken Down | Competing Against Harvey, the 800 Pound Gorilla | Why Legora is Undervalued at $5.5BN with Patrick Forquer, CRO @ Legora

Alex Rampell · Jan 12, 2026

SaaS companies fall into three groups under AI: those impervious or tailwinded because they hold hostages and can add features (Workday, NetSuite), those whose seat-based revenue could go to zero (Zendesk), and middle cases like Adobe

If every support ticket can be answered automatically you need zero Zendesk seats, whereas incumbents with distribution and lock-in can simply add AI features

63:29 20VC: a16z's $15BN Fundraise with Alex Rampell | The Best Companies Have Hostages Not Customers | The Best Founders Materialise Capital, Customers and Labour | Mid-Sized Funds with Die and The Future of Venture Capital

Anish Acharya · Feb 9, 2026

Systems of record vary widely in disruption risk: an on-prem database with no engagement layer or human workflows is at real risk, while a bank's core transactional system with high throughput, many human users and accuracy demands remains as good as gold

Defensibility comes from the engagement layer and human workflows built around the system of record, not from being a system of record per se

24:47 20VC: Is SaaS Dead in a World of AI | Do Margins Matter Anymore | Is Triple, Triple, Double, Double Dead Today? | Who Wins the Dev Market: Cursor or Claude Code | Why We Are Not in an AI Bubble with Anish Acharya @ a16z

Harry Stebbings · Aug 3, 2026

Entrenchment in the enterprise determines survivability: deeply integrated incumbents like ServiceNow and Salesforce are safe, while less sticky products like Wix are exposed

It's all about how integrated and sticky you are within the enterprise

53:59 20VC: 70% of Neolabs Will Die | There Will be a $100BN US Open-Source Model | Data is a Trillion $ Market | Governments Cannot Regulate Models: It is Too Late | The Cyber Attacks to Come Will be Insane with Anastasios Angelopoulos @ Arena

Jonathan Siddharth · Dec 1, 2025

Incumbents that fail to adopt AI will lose to competitors operating at a fraction of the headcount with better pricing and customer experience

A competitor running at one hundredth the headcount while pricing insurance deals better, earning more premium and paying out fewer claims, will eat the laggard's lunch — the cost of not adopting is too high

Scope: illustrated via the hypothetical insurance company

18:58 20VC: Scale, Surge, Turing, Mercor: Who Wins & Who Loses in Data Labelling | Is Revenue in Data Labelling Real or GMV? | Why 99% of Knowledge Work Will Go and What Happens Then? | Why SaaS is Dead in a World of AI with Jonathan Siddharth @ Turing

Nikesh Arora · Jun 22, 2026

The bear case for an incumbent is failing to make the AI-first transition within three years, because agents will become real and do work humans did manually, and that must be embodied in the product or a competitor builds a better mousetrap

False positives will keep reducing and agents will genuinely take over manual human work, so products that don't absorb this get displaced

Scope: conditional on execution over the next three years

52:53 20VC: Nikesh Arora on the Frontier Model Problem: Breadth vs Depth | The Future of Token Costs | Memory Becoming the Moat | Where Value Accrues: Infra, Models, or Apps? | Why Enterprise AI is Not Ready & Systems of Record vs Systems of Intelligence

Jerry Murdock · Aug 22, 2026

Companies at $400–500M revenue that haven't already integrated AI compellingly won't be able to do much of anything with the company in two years

The question isn't current revenue but whether that revenue persists, which depends on having a thoughtful AI product and strategy

Scope: Applies to companies without an existing AI strategy; Two-year horizon

42:06 20VC: The AI Bubble Will Burst: Half the Neoclouds Will Die | China: Should We Ban Chip Exports & Be Fearful of Chinese Open-Source | Mag7: Who Dies and Who Thrives: Why Meta is Meh and Microsoft is Mega

Varun Mohan · Jun 2, 2025

Complex applications like Salesforce and Workday will not be dethroned in the short term by a company that is just a database with an agent on top

Enormous inertia from custom workflows built around these tools and from humans already trained on them; and agents today are not trusted enough to write to databases at scale without human supervision, they mostly read

Scope: short term; concedes a company's state can in principle be distilled to a database

41:04 20VC: Windsurf Founder on Will Model Companies Own the App Layer | Why Moats Do Not Exist in a World of AI | Why the Notion of Single Person $BN Companies is BS | Lovable vs Bolt & Cursor vs Windsurf: How Does it All End with Varun Mohan

Mitchell Green · Mar 7, 2026

Most software incumbents — Oracle, Microsoft, SAP — will not disappear because of AI; some will innovate and become far bigger, some will grow 10% a year, and he would bet any amount of money on it.

Mainframes launched in 1950 and are still a $5.5B market running most banks; legacy technology persists for decades, and high gross-dollar-retention businesses (90–98%) are especially durable.

Scope: new giant companies will also be created; some incumbents will go to zero, but nowhere near all

16:47 20VC: Why the SaaS Apocalypse is BS | Why China Will Win the AI War | Why 50% of VCs Should Not Exist and are Tourists | Why Stock-Based Comp is the Hidden Sin of the Valley with Mitchell Green, Lead Edge Capital

Chris Degnan · May 23, 2026

He doesn't fully buy that SaaS is dead — legacy technology dies a slow death and incumbents like Salesforce will remain systems of record even as the market changes

When he joined Snowflake the enterprise was still in the data center and IBM was still selling mainframes — and IBM still sells mainframes

Scope: acknowledges the market is changing

45:10 20Sales: The $100M CRO Bubble: Why Anthropic Are Causing a Comp Crisis | Why You Should Never Hire From Salesforce or Service Now | How to Hire, Train and Forecase in a World of AI with Chad Peets and Chris Degnan

Jonathan Siddharth · Dec 1, 2025

SaaS as we know it is over, so if you believe in AGI you cannot invest in SaaS apps

Many SaaS companies existed because software was hard and capital-intensive to build; LLMs collapse that bar so companies will build custom software themselves, and incumbents also risk being 'sonic boomed' by foundation model companies moving into the app layer

46:36 20VC: Scale, Surge, Turing, Mercor: Who Wins & Who Loses in Data Labelling | Is Revenue in Data Labelling Real or GMV? | Why 99% of Knowledge Work Will Go and What Happens Then? | Why SaaS is Dead in a World of AI with Jonathan Siddharth @ Turing

Chad Peets · Jan 9, 2026

The most important unanswered question is whether AI eats SaaS — whether SaaS companies survive, whether LLMs absorb categories like code and customer service, or whether independent AI companies emerge

Public market multiples show it: SaaS multiples are compressing because people question SaaS's future, while AI multiples like Palantir's are through the roof on assumed endless upside

Scope: says nobody knows the answer; allows others may already be discussing it

55:07 20Sales: Why the Best Sales People are F**** in the Head | Why Remote Work is BS | Why Europe is a Nightmare for Recruiting Sales Teams | How to Build a Sales Machine | What Everyone Gets Wrong Today in Sales with Chad Peets

Sebastian Siemiatkowski · Feb 16, 2026

Software is becoming a much riskier investment category and the future of SaaS is unclear

56:01 20VC: SaaS is Dead: Why Systems of Record Will Die in an Agentic World | What Revenue Multiple Will Software Companies Trade At? | From 7,000 to 3,000: We Need Less People Than Ever with Sebastian Siemiatkowski

Jesse Zhang · Sep 19, 2025

AI customer service platforms create lock-in as a 'system of intelligence' — the stored business logic and way the business works is what's hard to port, unlike conversations or procedure outlines.

Traditional systems of record lock you in through hundreds of hours of configuration; AI solutions instead accumulate business logic that updates over time and can't easily be moved.

Scope: some assets like conversations and operating procedure outlines can be ported

39:08 20VC: Why 90% of Founders Build Startups Wrong | Why AI Growth Rates are Sustainable & Remote Work is BS and the AI Talent War | Competing with Brett Taylor and Sierra: Who Wins the Customer Service War with Jesse Zhang, Decagon

Aaron Levie · Apr 20, 2026

The 'SaaS is just a database' reduction is wrong for software with substantial proprietary business logic above the database — an ERP system's supply chain and accounting logic doesn't go away.

All software has a database behind it, but the durable value sits in the business logic layer — security, permissioning, and domain workflow encoded in the APIs.

15:08 20VC: Everyone is Wrong; We Will Have More Developers in Five Years | Why Frontier Labs Will Be Way More Valuable Than They Are Today | Are SaaS Companies Cooked: Which Thrive & Which Die with Aaron Levie, Founder at Box

Maor Shlomo · Nov 24, 2025

Smaller CRM companies — any software that is essentially a front end on a database with team features and integrations and no moat — will be eliminated

They have no moat, whereas Salesforce's sheer size lets it make big bets and move in many directions while everyone adapts to the AI era

13:22 20VC: Base44's Maor Shlomo on How Vibe Coding Will Kill SaaS and Salesforce | Why it is BS that Vibe Coding Platforms Do Not Have Defensibility and Bad Margins | Why He Worries About Google, Not Replit and Lovable | Why Long Anthropic, Not OpenAI?

Jerry Murdock · Feb 28, 2026

Cursor will survive and pivot because the team is smart and has capital, customers and time to embrace autonomous agents

In AI you have to skate to where things are going, and Cursor has the resources and runway to make that move

Scope: conditional on them quickly embracing autonomous agents

7:32 20VC: Why Cursor is Dead | An AI Tsunami is Coming & You Need to Prepare | Systems of Record Become Valueless Databases with Agents | Is This The End of Tech Private Equity with Jerry Murdock, Co-Founder of Insight Partners

Christian Lanng · Sep 27, 2023

Enterprise applications should not define how people work; a human-centered layer should let people work the way their brain works while systems like Salesforce, Workday and SAP sit in the background

Salesforce doesn't have a human at the center — it turns you into a robot clicking buttons in a certain order, and storing that useless procedural knowledge wastes your brain

43:39 20VC: "How Being a Founder Almost Killed Me"; We Have Lied to a Generation of Founders | The Hardest Truths About Being a Founder Revealed | Why AI Co-Pilot is BS, Seat Pricing is Over & User Interfaces are Stupid with Christian Lanng

Maor Shlomo · Nov 24, 2025 · hedged

Most existing SaaS products will ship built-in vibe coding capabilities that let customers customize the software to their needs, with the underlying product becoming a system of record on top of which users generate their own UI

This makes software more liquid — instead of configuring features, customers get extensibility, and you can already see it with Monday Vibe and Microsoft building vibe coding on top of Office

Scope: speculating about Salesforce specifically — 'I don't know if Salesforce is planning to do that at all'

17:45 20VC: Base44's Maor Shlomo on How Vibe Coding Will Kill SaaS and Salesforce | Why it is BS that Vibe Coding Platforms Do Not Have Defensibility and Bad Margins | Why He Worries About Google, Not Replit and Lovable | Why Long Anthropic, Not OpenAI?

Anish Acharya · Feb 9, 2026

The main way coding agents show up in enterprise software is by dramatically lowering the cost, speed and risk of switching between SaaS providers, which is a net positive for the ecosystem

Systems integration is what locks customers in — migrating from SAP to Oracle has been a multi-year, high-risk, likely-to-fail project, and coding agents collapse that complexity, turning hostages into customers

Scope: an underdiscussed dynamic

9:35 20VC: Is SaaS Dead in a World of AI | Do Margins Matter Anymore | Is Triple, Triple, Double, Double Dead Today? | Who Wins the Dev Market: Cursor or Claude Code | Why We Are Not in an AI Bubble with Anish Acharya @ a16z

Gokul Rajaram · Mar 16, 2026

Switching costs will go to essentially zero over the next one to two years as data becomes trivially portable between ecosystems and products can be cloned almost pixel by pixel.

With easy data portability and near-identical clones popping up everywhere, there is nothing left holding a customer in place — like asking whether you cheer for the player or the team when they switch.

Scope: over the next one to two years

17:25 20VC: The 8 Moats of Enduring Software Companies: How to Analyse for Durability and Defensibility in a World of AI | Why Dropouts are "AI Maxing" the World & Remote Early-Stage Companies are Dying with Gokul Rajaram

Christian Lanng · Sep 27, 2023

The disruption risk to enterprise applications is under-discussed relative to Google search; Salesforce is essentially a database and LLMs query databases well, so the 15 layers in between are unnecessary

LLMs can talk directly to databases, removing the need for the application layer

58:34 20VC: "How Being a Founder Almost Killed Me"; We Have Lied to a Generation of Founders | The Hardest Truths About Being a Founder Revealed | Why AI Co-Pilot is BS, Seat Pricing is Over & User Interfaces are Stupid with Christian Lanng

Saam Motamedi · Jul 15, 2024

In ten years sales reps won't even know what the Salesforce UI looks like because agents will navigate underlying systems on their behalf, with dynamically generated UI

Generative AI can change or remove the interface entirely; reps already hate the clunky Salesforce UI and are only constrained by it

19:53 20VC: Why We Are in a Bubble & Now is Frothier Than 2021 | Why $1M ARR is a BS Milestone for Series A | Why Seed Pricing is Rational & Large Seed Rounds Have Less Risk | Why Many AI Apps Have BS Revenue & Are Not Sustainable with Saam Motamedi @ Greylock

Nikesh Arora · Jun 22, 2026

In technology, missing one platform shift is survivable, missing two leaves you badly damaged, and missing three makes you obsolete

Technology transitions compound; failure to catch them stacks up

0:00 20VC: Nikesh Arora on the Frontier Model Problem: Breadth vs Depth | The Future of Token Costs | Memory Becoming the Moat | Where Value Accrues: Infra, Models, or Apps? | Why Enterprise AI is Not Ready & Systems of Record vs Systems of Intelligence

Nikesh Arora · Jun 22, 2026

In technology you can survive missing one trick, are partly impaled missing two, and risk obsolescence missing three — which is why leaders must force a point of view on fast-moving tech

The stuff is moving so fast that without a point of view and team attention, you miss shifts cumulatively

49:55 20VC: Nikesh Arora on the Frontier Model Problem: Breadth vs Depth | The Future of Token Costs | Memory Becoming the Moat | Where Value Accrues: Infra, Models, or Apps? | Why Enterprise AI is Not Ready & Systems of Record vs Systems of Intelligence

Mitchell Green · Mar 7, 2026

The software selloff is a buying opportunity and the incumbent software companies are not going away

Incumbents have distribution, data, and balance sheets, so it's a fool's errand to think they all disappear

Scope: concedes big disruptions also create new winners and blow up some incumbents

0:34 20VC: Why the SaaS Apocalypse is BS | Why China Will Win the AI War | Why 50% of VCs Should Not Exist and are Tourists | Why Stock-Based Comp is the Hidden Sin of the Valley with Mitchell Green, Lead Edge Capital

Anastasios Angelopoulos · Aug 3, 2026 · hedged

The SaaSpocalypse has been overstated and Salesforce will not go downhill

Salesforce has a strong AI strategy, and people underestimate how hard it is to replicate what incumbents built from a network and data perspective

Scope: 'we'll see'

53:34 20VC: 70% of Neolabs Will Die | There Will be a $100BN US Open-Source Model | Data is a Trillion $ Market | Governments Cannot Regulate Models: It is Too Late | The Cyber Attacks to Come Will be Insane with Anastasios Angelopoulos @ Arena

Aaron Levie · Apr 20, 2026

For Box specifically, the agent shift is all upside — the headless, API-driven version of the business has always been larger than end-user interaction and more agent-generated content means more data to secure and govern.

API call volume already vastly exceeds end-user interactions because content flows through ERP systems, wealth management portals and invoice processing where users never see Box; Box already knows how to monetize that.

18:44 Agent generated content increases demand for governance platforms

Miles Grimshaw · Sep 18, 2023

The LLM architecture is a total break from the database-query mental model of how applications work, which large incumbents struggle to internalize — and that break is what makes it exciting

People at big companies keep asking which database an answer was pulled from, revealing how deeply the old model is embedded

60:20 Llm architecture breaks the database query mental model incumbents struggle to internalize

Anish Acharya · Feb 9, 2026 · hedged

The market is shifting from products focused on execution, which coding agents are subsuming, to products focused on thinking — and Figma is an N-of-1 network-effects product well positioned for that shift

Much of the thinking work will be done in products like Figma, while execution-oriented products get absorbed by agents

47:34 Execution products get subsumed thinking products gain

Daniel Khachab · Oct 28, 2024 · hedged

Incumbents like Salesforce should become the prompt-driven application layer themselves, with the underlying system reduced to a database, rather than ceding the interface to a layer above them

Salespeople hate manual data entry; given the choice between filling forms and pasting notes and letting the tool do the rest, they choose the latter, so the prompt-native tool wins adoption and makes reps more efficient

11:58 Incumbents should become the prompt driven interface themselves reducing system to database

Aravind Srinivas · Jun 15, 2026

Public enterprise SaaS companies can survive the AI transition only by cutting costs and acquiring new value-producing products; those just selling the same software will not be around

Salesforce endured because it kept buying the next thing, and IBM survived via Red Hat, HashiCorp and Confluent acquisitions

64:11 Only cost cuts and acquiring new products save incumbents

Jerry Murdock · Feb 28, 2026

Salesforce will survive for a long time regardless, but the right way to value it is to watch the health of the ecosystem of companies built on top of it — if those get knocked off one by one, Salesforce's underlying value declines

Dozens of companies like nCino sit on top of the Salesforce system of record, so their fate is the leading indicator of the platform's worth; Salesforce is like an 8,000 meter peak that won't melt overnight

19:16 Ecosystem health is the leading indicator of platform value

Jerry Murdock · Aug 22, 2026

We have entered a 'co-work' era of autonomous agents that will threaten SaaS companies lacking a system of record or AI strategy

As co-work becomes more stable and broadly used, SaaS incumbents without AI or a system of record have no defense

42:51 System of record is the defense against co work agents

Jerry Murdock · Aug 22, 2026

Microsoft is a buy even without a frontier model, because it controls human communication for global enterprises and that is a money machine that cannot be displaced

Enterprise email/Exchange isn't going away and can't be cancelled like a cable subscription; they won't control agent communication, but human communication is monetizable forever

57:13 Control of human communication is undisplaceable even without a frontier model

Sebastian Siemiatkowski · Feb 16, 2026

Tech and financial services are heading for a brutal awakening: high switching costs created money-printing machines and perk-laden cultures, and those firms will now have to work daily to earn customers like a normal retail business

Unlike restaurants or retailers who must wake up every day and fight to put the right product in front of the customer, tech and fin lived off entrenched switching costs

37:06 Lost switching costs force tech firms to earn customers daily like retailers

Eran Zinman · Mar 2, 2026

Large M&A is not how Monday will win; betting the company's future on acquiring a startup would be a failure of his job as CEO

They already have everything needed to execute internally, and the public-private valuation inversion means a $5M ARR startup is priced at $2B while Monday is at $3.7B, so any meaningful acquisition would require heavy use of undervalued stock

48:14 Incumbents must build internally not acquire their ai future

Eran Zinman · Mar 2, 2026 · hedged

The AI disruption may turn out to be Monday's biggest opportunity, letting the company reaccelerate and grow far larger than it otherwise would have

If they do the right things, the shift resets the field in their favour rather than just threatening the existing path

53:11 Platform shift is a reacceleration opportunity for incumbents

Eran Zinman · Mar 2, 2026 · hedged

Being a bigger company makes AI-era change harder, and serving SMBs may be an advantage rather than a weakness because SMBs adopt new technology sooner

The hardest part of change is moving humans, and on offense the traits that look like disadvantages when playing defense — smaller size, SMB customer base — become advantages

56:14 Smaller size and smb customer base make incumbents more adaptable

Amjad Masad · Apr 25, 2026

Enterprises are not ripping out system-of-record SaaS like Salesforce and Workday; instead they build on their APIs — but a growing cohort skips SaaS tools entirely and builds on the data warehouse, making the warehouse the real system of record

That's what Replit is seeing in enterprise; MCPs and hooks into HubSpot/Salesforce on one side, Databricks-based builds on the other

20:18 Warehouse native building quietly becomes the real system of record

George Sivulka · Jan 22, 2025

Traditional enterprise B2B applications like ServiceNow, Salesforce and UiPath are not attractive businesses — he'd kill all three

Traditional enterprise B2B software isn't sexy

59:11 Traditional enterprise b2b incumbents are unattractive businesses vulnerable to ai

Jean-Denis Greze · Jul 21, 2023

Salesforce will not stay number one over the next decade

Salesforce is increasingly just a database and store rather than where the UX happens — users increasingly live in tools like Gong — and it remains the base purchase only because everyone still has to store data there

30:57 Salesforce loses ux centrality becoming just a database as tools like gong capture user attention

Mitchell Green · Mar 7, 2026

Workday's slowing growth reflects the law of large numbers rather than AI disruption, and comparing it to Anthropic or OpenAI is unfair because those companies grow without profits

Workday is a $10B revenue business throwing off $3B of free cash flow, and its AI business is growing fast inside it; huge companies decelerate

4:45 Deceleration is the law of large numbers not ai disruption

Avishai Abrahami · Jul 13, 2026

Salesforce's core value is enterprise trust to hold customer data, not the CRM software, so it cannot be vibe-coded away

Huge banks and companies like JPMorgan trust no one outside their own secure environment except Salesforce; that trust is extremely hard to displace and the CRM functionality is a small part of the value

8:21 Enterprise trust to hold the data not the software is the moat

Varun Mohan · Jun 2, 2025

Large companies are slow at development velocity because their employees face no existential dread — shipping a weak product costs them nothing personally, whereas a startup dies

An engineer inside Amazon working on a competing product doesn't fear losing their job in six months; if Windsurf had shipped hyperscaler-quality product it would have been out of the market a year ago

25:15 Lack of existential pressure not distribution explains incumbent slowness

Saam Motamedi · Jul 15, 2024

A horizontal incumbent's core can only be attacked when the data model, the delivery model, or the interface profoundly changes

Incumbent data models become the standard that users, managers, channel partners and adjacent tools all adapt to, so only a shift in one of these three dimensions creates an opening

18:08 Incumbents core is only attackable when data model delivery model or interface changes

Gokul Rajaram · Mar 16, 2026

Systems of record like Salesforce and NetSuite are more attractive than most software companies, but only survive if they decide where their profit pool sits and commoditize the complement — giving away data storage and pricing workflows on outcomes, or giving workflows away and monetizing data.

They cannot wait around for others to build agentic products on top of them; they have to use their own data to build better free products and squeeze the layer above.

19:31 Systems of record must commoditize the complement to survive

Jake Saper · Mar 10, 2025

IBM is the stock to short, because AI is the critical enabler for migrating legacy COBOL mainframe code to the cloud and IBM's billions in mainframe maintenance and server revenue depend on that code staying trapped

75% of the Fortune 500 still run core applications on mainframes in COBOL that nobody writes anymore; if companies like Mechanical Orchard succeed at migrating that code, IBM's revenue stream is in trouble

58:06 Ai enabled mainframe migration threatens incumbent legacy maintenance revenue like ibms

Amit Bendov · Sep 12, 2025

Revenue software is moving from a data-entry world to a system of intelligence that runs the entire organisation, and Gong's vision is to be that Revenue AI Operating System across a much more diverse set of industries.

If every company benefits from AI in revenue, selling becomes easy and competition shifts to simply having the best product.

61:35 Revenue software evolves into an intelligence layer running the whole org