AI가 SaaS를 죽이지 않았다. SaaS가 무엇에 대해 요금을 부과하는지 바꿨을 뿐이다.

THE DAILY H.E.A.T. Torching Wall Street’s Obsolete Playbook. Delivered straight to your inbox 5 days a week. Subscribe Home Posts AI Didn't Kill SaaS. It Changed What SaaS Charges For. AI Didn't Kill SaaS. It Changed What SaaS Charges For. Wall Street buried the wrong patient. The February software selloff wasn't a funeral. It was a forced repricing of the unit of value. Jun 11, 2026 I’ve been a trader and investor for 44 years. I left Wall Street long ago—-once I understood that their obsolete advice is designed to profit them, not you. Today, my firm manages around $5 billion in ETFs, and I don’t answer to anybody. I tell the truth because trying to fool investors doesn’t help them, or me. In Daily H.E.A.T. , I show you how to Hedge against disaster, find your Edge, exploit Asymmetric opportunities, and ride major Themes before Wall Street catches on. Table of Contents H.E.A.T. The AI Buildout Has a Physical Layer News vs. Noise: What’s Moving Markets Today ETF News In Case You Missed It H.E.A.T.    $285B Enterprise software market cap erased — Feb. 3, 2026 $1.2B Salesforce Agentforce ARR — up 205% YoY, Q1 FY27 $13T U.S. labor-cost pool now in agentic software's crosshairs   WALL STREET BURIED THE WRONG PATIENT For twenty years, software companies charged by the user. More employees meant more seats. More seats meant more ARR. AI breaks that math. If one agent can do the work of ten employees, the seat is no longer the right unit of value. But that does not mean the best software companies are dead. It means the winners have to change the meter. On February 3, the market treated enterprise software like a melting ice cube. Jefferies christened it the 'SaaSpocalypse': hundreds of billions in enterprise software value erased as investors decided AI agents would hollow out the seat-based model. The diagnosis was half right. The seat model was in serious trouble. Software incumbents with the right data and the right action layer were not. Public SaaS growth rates had declined every single quarter since 2021. AI didn't murder the old model — it just finally signed the death certificate on a business model that was already in hospice. The selloff wiped approximately 25–30% from names like Salesforce, Adobe, ServiceNow, and Intuit in a matter of weeks, with the broader software ETF (IGV) falling 35% from its 2025 high to its April 10 trough. Painful and real. Also: not the whole story. The next SaaS multiple will not be awarded to the company with the most seats. It will be awarded to the company that owns the data layer, controls the action layer, and can charge for completed work. That is a larger addressable market, not a smaller one.   The best SaaS companies are trying to move from the IT budget into the labor budget. The ones that own both enterprise memory and the action layer get a second act. The ones that own only the database become utilities. THE BLOOMBERG EFFECT: ENTERPRISE MEMORY IS THE FLOOR THE BLOOMBERG PARADOX Nobody loves the Bloomberg terminal's interface — the design hasn't changed meaningfully since the 1990s. Firms pay $24,000 a year per seat not because it's beautiful, but because years of proprietary workflows, custom data models, and institutional logic are baked directly into it. Ripping it out doesn't cost money. It costs identity. That hostage dynamic is worth more than any feature roadmap.   Enterprise software incumbents have spent decades building what analysts call 'enterprise memory' — the accumulated rules, data graphs, workflows, and institutional logic that make a platform sticky in ways no startup can replicate from a cold start. Salesforce doesn't just store your CRM data. After a decade, it knows your approval chains, your deal logic, your rep-specific overrides, your customer segmentation rules. An AI-native startup might be faster, cleaner, and cheaper — but it starts at zero knowledge of your business. The data validates this. Salesforce closed Q1 FY27 (reported May 27, 2026) with $11.1 billion in revenue, up 13% year-over-year. Its Agentforce platform hit $1.2 billion in ARR, up 205% YoY, with more than 50% of Agentforce and Data Cloud 360 bookings coming from existing customer expansion. These are not new customers experimenting — they are existing hostages doubling down. Add 28.6 trillion tokens processed (up 152% QoQ) and 3.8 billion Agentic Work Units delivered to date (up 111% QoQ) — AI executing real work inside the customer's own data context — and the moat isn't weakening. It's widening. Enterprise memory is the floor. But it is not the ceiling. That is where the next section gets uncomfortable for the bulls. THE ACTION LAYER: WHERE THE NEXT MULTIPLE IS WON OR LOST Here is the scenario the optimists aren't fully pricing. Andreessen Horowitz is actively funding startups that treat Salesforce and ServiceNow not as platforms to compete with, but as backend databases to be wrapped. a16z calls this the 'system of action' layer — an AI-native interface that sits on top of the legacy system of record, handling workflows, cross-system orchestration, and user-facing automation, while the old platform persists underneath as a dumb-but-critical database. If that middleware layer captures the user relationship — and with it the agentic pricing upside — the incumbent keeps its recurring subscription revenue but loses the growth story. It becomes a toll road: valuable, defensible, and capped. The market would reprice it as a utility, and the current valuation recovery stalls permanently at a discount. The decisive question is whether incumbents can open their action layer fast enough, and price it well enough, to make middleware a feature of their platform rather than the thing that disintermediates them. Salesforce's AWU numbers suggest they are winning that race today. The others have less runway than their current valuations imply.   The risk for incumbents is not replacement. It is disintermediation. If the user relationship migrates to the middleware, the incumbent keeps the subscription floor but loses the agentic ceiling. THE $13 TRILLION PIVOT: FROM IT BUDGET TO LABOR BUDGET Here is the reframe that changes the entire valuation conversation. Traditional SaaS competed for the software budget — a $400 billion global sandbox. Agentic SaaS competes for the labor-cost pool: wages, benefits, and outsourced work that can be automated, compressed, or repriced. Private-industry wages and salaries alone ran approximately $11.0 trillion in 2025 by FRED/BEA data. Add benefits and contract labor and you are well north of $13 trillion. Deloitte's 2026 forecast projects that up to half of all organizations will allocate more than 50% of their digital transformation budgets toward AI automation this year, with agentic AI investment reaching 75% of companies. Gartner projects that by 2030, 40% of enterprise SaaS spend will shift toward usage-, agent-, or outcome-based pricing. The old model: charge per seat, grow by adding headcount. The new model: charge per outcome, grow by replacing headcount — or more precisely, by capturing a percentage of the value the agent delivers. When the software is the worker, the ceiling isn't the number of employees. It's the economic output of the enterprise. THE MARGIN ARCHITECTURE Token costs are falling roughly 10x per year. The value of the completed outcome stays high. The spread accrues to the vendor — but only if the agent is reliable, governed, integrated, auditable, and close enough to the customer outcome to price the result. The winners are not the companies that run the model cheapest. They are the companies that can charge for completed work. The cost stack is not just compute. It includes data cleanup, integration, evaluation, compliance, permissions, human review, and customer-success support. Gross margins can still be exceptional. But the margin belongs to whoever owns the outcome relationship — not the infrastructure layer.   THE SEAT-BASED EXTINCTION — AND WHO REPRICES FIRST The most legitimate fear embedded in the February selloff was pricing model risk. If AI agents can do the work of 100 sales reps, you don't need 100 Salesforce seats. Per-user revenue collapses. Multiples compress. The whole valuation architecture cracks. This is the right thesis applied to the wrong companies. The seat model is dying — but only for the laggards. The companies repricing ahead of schedule are cannibalizing their own seat revenue before a competitor can do it for them. Salesforce rolled out Agentic Work Units as an explicit operating metric alongside ARR precisely to signal the shift: from counting users to counting completed work. Salesforce CEO Marc Benioff acknowledged the friction directly — customers pushed back on pure per-conversation pricing and want more flexibility. That is a negotiation, not a rejection. The direction is irreversible. Companies that price the volume of work finished — rather than the number of employees using the tool — are entering a structurally larger market. That is the trade.   TIERED WINNERS: WHO OWNS THE METER CHANGE Companies positioned to capture value as the unit of SaaS pricing shifts from seats to completed work. TIER TICKER COMPANY THESIS CATALYST TIER 1 CRM Salesforce Cleanest public proof set: $1.2B Agentforce ARR (+205% YoY), AWU metric live, 50%+ bookings from existing customers. Owns memory + action layer simultaneously. FY27 guidance raised to $46B; Q2 AWU acceleration vs. seat-revenue trend TIER 1B NOW ServiceNow Deep workflow automation in IT/HR/Finance; cRPO +22.5% YoY; management framing platform as 'AI control tower.' Less clean AWU equivalent metric than CRM, but moat is comparably deep. AI Pro SKU adoption; Q2 workflow automation attach rate TIER 2 MSFT Microsoft Copilot embedded across M365/Azure; owns the enterprise desktop and cloud substrate. Baggage: vast existing seat base creates internal cannibalization drag. Copilot seat-to-consumption conversion rate; Azure AI revenue mix TIER 2 ADBE Adobe Firefly >$250M ending ARR; >45% QoQ generative credit consumption growth (Q1 FY26). Credits are an early template for metered AI consumption. Risk: standalone gen-AI tools commoditize creation. Firefly credit consumption acceleration; enterprise contract renewal rates TIER 2 ORCL Oracle Cloud ERP + database layer is AI's preferred substrate. Multi-cloud positioning and OCI capacity additions strengthen lock-in. Less direct action-layer exposure than CRM/NOW. OCI capacity additions; Fusion ERP AI agent attach rate TIER 2+ PLTR Palantir Not classic SaaS — but arguably the cleanest public 'action layer over messy enterprise data' name. AIP platform operationalizes AI on proprietary data in defense and commercial settings. Commercial AIP deal velocity; U.S. government contract expansion TIER 3 WDAY Workday HR/Finance data moat is defensible; Illuminate AI platform is strategically correct but early. Needs proof that agentic consumption offsets seat-count pressure before re-rating. Illuminate customer pilots converting to production; NRR trajectory TIER 3 SNOW Snowflake Data cloud as substrate for enterprise AI agents; Cortex AI gaining attach. Consumption model already in place. Needs Cortex to become a material NRR driver. Cortex product contribution to NRR; net revenue retention trend     PRESSURE POINTS — INCLUDING ONE CONTESTED INCUMBENT TICKER COMPANY RISK FACTOR WHAT TO WATCH SAP SAP SE CONTESTED INCUMBENT: Deepest enterprise memory in ERP, but action layer is nascent. a16z notes SAP will likely persist as system of record while a user-facing AI layer emerges above it. Risk: SAP becomes the database that everyone wraps. BTP (Business Technology Platform) AI agent adoption; action-layer capture vs. middleware encroachment INTU Intuit Consumer-facing AI tools (TurboTax, Mint) face direct competition from GPT-native tax and finance apps. Less enterprise-memory protection than B2B peers. Paid user retention in core franchises; Assist AI engagement depth DDOG Datadog Monitoring spend gets rationalized as AI ops teams shrink. Pricing remains seat-adjacent. Risk: the observability layer gets absorbed into the hyperscaler stack. NRR stability; AI workload monitoring as % of new ACV ZS Zscaler Security platform consolidation risk. If Microsoft or CrowdStrike win the bundle, standalone zero-trust erodes. No clear agentic consumption story yet. Platform NRR vs. point-product churn; MSFT Defender overlap in enterprise accounts     CREDIBILITY FIREWALL Separating what is confirmed from what is directional inference. SOURCED / REPORTED DIRECTIONAL INFERENCE Salesforce Q1 FY27 revenue: $11.1B, +13% YoY (reported May 27, 2026) The February selloff marked the effective bottom of the SaaS de-rating cycle Agentforce ARR: $1.2B, +205% YoY as of Q1 FY27 Incumbents with both data and action layers will capture outcome pricing before middleware rivals scale 3.8B Agentic Work Units delivered to date; +111% QoQ AWU metric will become a standard enterprise software KPI within 2-3 years 28.6 trillion tokens processed on Salesforce platform; +152% QoQ Token cost deflation sustains gross margin expansion even as seat revenue declines 50%+ of Agentforce and Data Cloud 360 bookings from existing customers SAP's action layer will be captured by third-party middleware unless BTP accelerates significantly Salesforce FY27 full-year guidance raised to $46B PLTR's AIP is better positioned than traditional SaaS for near-term agentic adoption IGV ETF fell 35% from 2025 high to April 10, 2026 trough Companies that lose the action layer to middleware face re-rating to utility multiples Software names fell ~25-30% in weeks following Feb. 3 selloff (per a16z commentary) The $13T labor-cost framing will anchor software re-ratings when agentic ARR hits material scale Adobe Firefly ending ARR >$250M; generative credit consumption +45% QoQ (Q1 FY26)   ServiceNow cRPO +22.5% YoY, Q1 2026; management framed platform as 'AI control tower'   FRED/BEA: private-industry wages and salaries ~$11.0T in 2025   Deloitte 2026: up to 50% of orgs to allocate >50% of digital transformation to AI   Gartner: 40% of enterprise SaaS spend to shift to usage/agent/outcome pricing by 2030   a16z actively funding 'system of action' middleware layer startups (per public a16z posts)       BEAR CASE THREE WAYS THIS THESIS BREAKS 1. Enterprise adoption stalls at the pilot stage. CFOs approve AI experiments, not AI transformations. If Agentforce and its analogs hit a wall of procurement conservatism, the consumption revenue growth story disappears and the stocks re-rate back toward utility multiples on seat-revenue alone. 2. Middleware wins the user relationship. If a16z-backed orchestration layers capture the agentic interface before Salesforce and ServiceNow lock it down, incumbents become expensive databases with capped growth. The subscription floor holds. The premium evaporates. 3. Pricing model transition drags three-plus years. Benioff already flagged customer pushback on per-conversation pricing. If the AWU model takes longer to land than the street expects, earnings estimates built on consumption acceleration are too aggressive — and the stocks are already pricing the optimistic scenario.    FIVE THINGS TO TAKE AWAY 1. The February wipeout was a forced repricing, not a death sentence. The market de-rated SaaS from ~35x cash flow to ~15x because the seat-based logic was broken. That repricing is largely complete. The next re-rating — upward — is driven by whether consumption and work-unit revenue materializes at scale. 2. Enterprise memory is the floor. The action layer is the ceiling. Incumbents with decade-deep data moats are harder to displace than the bears believe. But memory alone is not enough. The valuation premium belongs to whoever captures the action layer — the AI-native surface where work gets done and outcomes get priced. 3. The addressable market expanded by roughly 30x. Traditional SaaS competed for $400B in software budgets. Agentic software competes for the $13 trillion labor-cost pool. Companies that price outcomes instead of seats aren't just growing into a bigger market — they're redefining what a software company is worth. 4. SAP is a contested incumbent, not just a laggard. The deepest enterprise memory in ERP is an asset. The question is whether SAP builds its own action layer or watches a middleware startup capture the user relationship above it. The answer determines whether SAP re-rates as a platform or gets stranded as an expensive database. 5. Watch AWUs, consumption revenue, gross retention, and seat cannibalization. The thesis is confirmed when work-unit revenue grows faster than seat revenue decays. Until that crossover is visible in the numbers, treat the recovery trade as a high-conviction thesis, not a fait accompli.   The seat was the unit of value for twenty years. The work unit is the unit of value for the next twenty. Own the companies that changed the meter before the market priced it in. The AI Buildout Has a Physical Layer Many of today’s data centers are still using copper wiring. The same metal we’ve been using for a hundred years. At the speeds AI demands with data moving between thousands of GPUs, billions of times a second, copper doesn’t just slow down. It turns that data into heat. The more you push through it, the worse it gets. There’s no software for fix for that. So what’s the answer? Explore the Photonics Layer….. Tuttle Capital Pure Play Photonics ETF (FOTO) Distributor: Foreside Fund Services | Investing involves risk including possible loss of principle. News vs. Noise: What’s Moving Markets Today CPI came in at 4.2%, the highest since May 2023. Ignoring energy, core CPI still increased 2.9%. Higher inflation leads to higher interest rates. Higher interest rates mostly impacts high flying companies who’s peak earnings are out in the future. ARKK is the poster child for those names, and it looks like a short here…. Continue to watch rates, that will be the tell…. 4.7% on the 10 year would be a warning signal, anything above 5% would likely be a real problem. Where Does the Money Go When AI Hits a Wall? When capital chases a tech theme, it tends to pile into the most obvious layer and miss the one underneath. AI spending is now bumping hard against memory. Hyperscalers — the big cloud builders like Amazon, Google, and Microsoft — have shifted memory from 8% of their build budgets to an estimated 30% in a single cycle. That capital has to go somewhere. If the constraint is memory, and the build can't move without it, shouldn't an investor own the layer AI runs on? View HBMX fund holdings → Distributor: Foreside Fund Services | Investing involves risk including possible loss of principal. <Link = http://www.hbmxetf.com/> ETF News $UFOD ( ▲ 4.66% ) $PCPP ( ▲ 0.84% ) 1980s Permanent Portfolio Gets A Bitcoin Upgrade In Tuttle's New ETF - Porter & Company Porter Portfolio The PCPP ETF modernizes the &quot;permanent portfolio&quot; strategy, combining Bitcoin, precious metals, insurance and other assets in an all-weather investment. Benzinga • Chandrima Sanyal A Stock I’m Watching $AAPL ( ▲ 1.39% ) AAPL is not what it once was, but the selloff appears a bit overdone. Could use the low yesterday as a stop. In Case You Missed It We invited Dan Ferris to the Financial Heat Podcast….. The H.E.A.T. (Hedge, Edge, Asymmetry and Theme) Formula is designed to empower investors to spot opportunities, think independently, make smarter (often contrarian) moves, and build real wealth. The views and opinions expressed herein are those of the Chief Executive Officer and Portfolio Manager for Tuttle Capital Management (TCM) and are subject to change without notice. The data and information provided is derived from sources deemed to be reliable but we cannot guarantee its accuracy. Investing in securities is subject to risk including the possible loss of principal. Trade notifications are for informational purposes only. TCM offers fully transparent ETFs and provides trade information for all actively managed ETFs. TCM's statements are not an endorsement of any company or a recommendation to buy, sell or hold any security. Trade notification files are not provided until full trade execution at the end of a trading day. The time stamp of the email is the time of file upload and not necessarily the exact time of the trades. TCM is not a commodity trading advisor and content provided regarding commodity interests is for informational purposes only and should not be construed as a recommendation. Investment recommendations for any securities or product may be made only after a comprehensive suitability review of the investor’s financial situation.© 2026 Tuttle Capital Management, LLC (TCM). TCM is a SEC-Registered Investment Adviser. All rights reserved. Keep Reading Jun 4, 2026 • 1 min read The memory trade most investors trade wrong This is a long-term hold Jun 9, 2026 • 14 min read The Second Phase of the Optics Trade The market knows the ticker. It doesn't know the supply chain. That gap is where the next round of returns gets made. Jun 5, 2026 • 2 min read Porter Stansberry's life’s work — now in an ETF View more 6