The B2B markets bifurcated in 2025 — and the laggards all share one thing in common.
2025 has been the year of the Great B2B Bifurcation.
- On one side: Palantir up 142%. Cloudflare up 80%. MongoDB up 70%. Security stocks printing money.
- On the other side: HubSpot down 51%. ServiceNow down 26%. Monday.com down 36%. Even though they all still had decent years of revenue growth.
Same macro environment. Same interest rates. Same customers.
Completely different outcomes.
What happened? And more importantly, what does it tell us about where B2B is headed?
The 2025 B2B Stock Leaderboard
Let’s look at the data through mid-December 2025:
The Winners:
- Palantir (PLTR): +142% — AI-native, mission-critical, government + commercial
- Cloudflare (NET): +80% — Edge computing + security + AI inference infrastructure
- MongoDB (MDB): +70% — Data platform positioned for AI workloads (was down mid-year, massive recovery after blowout Q3)
- Shopify (SHOP): +51% — E-commerce platform riding AI and merchant growth
- CrowdStrike (CRWD): +51% — Security is non-discretionary
- Snowflake (SNOW): +45-50% — Data infrastructure = AI infrastructure; growth re-accelerated to 29-32% YoY
- Zscaler (ZS): +34% — Zero Trust security
- Palo Alto (PANW): +30% — Cybersecurity platform consolidation
- Veeva (VEEV): +36% — Vertical SaaS (life sciences)
- Guidewire (GWRE): +27% — Vertical SaaS (insurance)
- Rubrik (RBRK): +17% — Data security/backup (newer IPO, holding up well)
The Middle:
- Oracle (ORCL): +14% YTD — but the real story is the AI rollercoaster: hit all-time high of $346 in September, now ~$188 (down 45% from peak)
- Toast (TOST): -4% — Restaurant vertical holding steady, essentially flat
- Datadog (DDOG): ~Flat — Observability solid but not breaking out
The Beaten Down:
- HubSpot (HUBS): -51% YTD — SMB CRM/marketing automation
- Bill.com (BILL): -43% — SMB financial operations
- Monday.com (MNDY): -36% — Work management (was ~$342, now ~$150)
- GitLab (GTLB): -32% — DevOps platform
- Monday (MNDY): –36% YTD — Work/project management platform
- Adobe (ADBE): -35% — Creative/marketing software
- Atlassian (TEAM): -34% — Project management/collaboration
- Salesforce (CRM): -31% — Enterprise CRM
- Asana (ASAN): -31% — Work management
- Klaviyo (KVYO): -29% — Marketing automation
- ServiceNow (NOW): -26% — IT service management
HubSpot Stock Was Unloved in 2025 — Despite a Strong Business
HubSpot has fallen from ~$880 to ~$370 — a 51% decline that makes it the roughest performing major B2B SaaS stock of the year.
The irony? The business is actually performing fine:
- 18-20% revenue growth
- Expanding margins
- Beat Q3 estimates
- Still the go-to platform for SMB go-to-market
So why the destruction?
The market is pricing in maximum AI disruption risk for SMB-focused horizontal SaaS.
HubSpot launched Breeze AI in September 2024, including customer agents, prospecting agents, and content agents. They’ve announced 200+ AI features. Their Breeze Customer Agent reportedly resolves 50%+ of support tickets.
But here’s what the market is worried about:
- The SMB customer is most vulnerable to AI disruption. A 50-person company that used to need 3 SDRs and 2 customer support reps might need 1 of each with good AI tools — whether from HubSpot or not.
- Seat-based pricing is under pressure. HubSpot’s business model still fundamentally relies on “more seats = more revenue.” AI’s core value proposition is “do more with fewer seats.”
- The AI narrative isn’t translating to numbers. Unlike Salesforce’s Agentforce ($500M+ ARR, 330% growth), HubSpot hasn’t disclosed AI-specific revenue metrics that move the needle.
- Competition is intensifying. AI-native CRM startups are emerging that could eat HubSpot’s lunch in the SMB segment before HubSpot can reinvent itself.
The Breeze platform is legitimately good. But the market isn’t convinced it changes the fundamental economics fast enough.
Salesforce: A Tough Year, But Growth Reaccelerating
This one is puzzling because Salesforce has actually shown real AI traction, and Q3 FY26 (reported December 3rd) showed growth reaccelerating:
- Q3 revenue hit $10.3B, up 9% YoY (vs. 8% in Q1)
- Subscription & support revenue up 10% YoY
- Q4 guidance of 11-12% YoY growth — finally back to double digits
- Full year raised to $41.5B at 9-10% growth
- cRPO up 11% YoY to $29.4B — strong forward indicator
The Agentforce numbers are legitimately impressive:
- Agentforce ARR surpassed $500M, up 330% YoY
- Combined AI/Data 360 ARR hit $1.4B (114% YoY growth)
- 18,500+ total Agentforce deals, 9,500+ paid (up 50% QoQ)
- 3.2 trillion tokens processed through their LLM gateway
- Agentforce accounts in production up 70% quarter-over-quarter
Marc Benioff has gone all-in on “Agentforce” as the company’s future. They acquired Informatica (closed November 2025), Regrello, and Waii to accelerate AI and data capabilities.
So why is the stock still down 31%?
- The hole was deep. Growth decelerated to 7.7% in Q1. Even with Q3’s improvement, investors got burned earlier in the year.
- Enterprise adoption is still early. Only ~8% of Salesforce’s 150K+ customer base has adopted Agentforce. The “innovation-adoption gap” is real, even if it’s closing fast.
- The AI cannibalization question lingers. Benioff acknowledged that Agentforce could be “a victim of its own success” — if agents reduce the need for human users, Salesforce sells fewer seats.
- The stock ran up in late 2024. CRM hit $367 in early 2025. Some of the 2025 decline is just mean reversion from AI hype.
The good news: Q3 may have been the turning point. Growth is reaccelerating, Agentforce adoption is inflecting, and the $60B revenue target by FY30 suddenly looks achievable. If Q4 delivers that 11-12% growth, 2026 could be a very different story for CRM.
Adobe: AI Everywhere in Its App, But The Markets Don’t Believe in the AI Revenue
Adobe is down 35% despite legitimately impressive AI numbers:
- Record revenue of $23.77B in FY25 (up 11%)
- Firefly generated 24B+ AI content creations
- AI-influenced ARR now exceeds 1/3 of the business
- Generative credit consumption tripled QoQ in Q4
- GenStudio exceeds $1B ARR (25%+ growth)
Adobe even admitted their own Firefly model wasn’t keeping pace with Google, OpenAI, and others — so they opened their platform to third-party models. That’s strategically smart.
So why is the stock cratering?
The market believes AI commoditizes creative tools.
If Canva, Figma, and dozens of AI-native startups can generate images, videos, and designs at near-Adobe quality for a fraction of the price, Adobe’s moat erodes.
Adobe’s defense is integration — Firefly embedded in Photoshop, Illustrator, Premiere, and the entire Creative Cloud suite. But the market isn’t sure that’s enough when standalone AI tools are “good enough” for 80% of use cases.
The other concern: AI is expensive. Adobe’s profits are up 9%, but the cost of running AI inference at scale will only grow. How long can they absorb those costs without raising prices?
Bill.com: The SMB AI Risk in Concentrated Form
Bill.com (BILL) is down 43% YTD.
Bill provides financial operations software (AP/AR automation) primarily to SMBs. The thesis was that SMBs would continue digitizing their back-office operations.
The problem: AI makes this category look commoditizable.
If AI agents can handle invoice processing, payment approvals, and vendor management, why do you need specialized Bill.com software? QuickBooks + AI agents might be “good enough.”
Bill’s challenge is similar to HubSpot’s: their SMB customer base is the most vulnerable to AI-driven seat compression. The companies that need 3 people in the finance function today might need 1 person + AI tomorrow.
What the Beaten Down Stocks Have in Common — Besides All Still Being Very Strong Businesses
Looking at HubSpot, Salesforce, Adobe, Bill, and Asana, clear patterns emerge:
#1. Horizontal > Vertical = Risk
Every major loser serves broad horizontal use cases (CRM, marketing, creative, work management, finance). The winners (Veeva, Guidewire) serve specific verticals where domain expertise creates defensibility.
Horizontal tools are easier to disrupt because the problem space is general. AI can attack general problems more easily than specialized ones.
#2. SMB Focus = Risk
HubSpot and Bill.com serve primarily SMBs. SMBs have smaller teams, simpler processes, and less switching cost inertia. An AI-native competitor can win an SMB customer with a good demo. Enterprise deals require years of relationship building.
SMBs are also more price-sensitive, so they’ll switch faster if AI enables cheaper alternatives.
#3. Seat-Based Pricing = Risk
The entire SaaS industry was built on the assumption that more users = more value = more revenue. AI inverts this: more value often means fewer users.
Companies that can pivot to consumption-based, outcome-based, or agent-based pricing will adapt. Companies stuck on seats will struggle.
#4. “AI Features” ≠ “AI Business”
Every loser has shipped AI features. HubSpot has Breeze. Salesforce has Agentforce. Adobe has Firefly.
But adding AI to existing products isn’t the same as building an AI-native business. The winners in 2025 — Palantir, the security companies — didn’t “add AI.” They are AI companies that solve specific, mission-critical problems.
The AI Divide: What Separates Winners from Losers
The data tells a clear story:
2025 Winners share these traits:
- AI-native or AI-first architecture
- Mission-critical positioning (security, data infrastructure, government)
- Vertical specialization
- Clear AI differentiation in the product
- Enterprise or government customer base
2025 Losers share these traits:
- Traditional SaaS with AI features bolted on
- Horizontal tools serving general use cases
- Seat-based pricing models
- SMB customer concentration
- AI narrative without AI-specific revenue proof
What This Means for Founders
If you’re building a B2B company in 2025, the market is sending a clear message:
#1. Go Vertical
Horizontal tools are under pressure. Vertical solutions with deep domain expertise are defensible. Veeva for life sciences. Guidewire for insurance. Procore for construction. The best opportunities are in verticals that haven’t been AI-transformed yet.
#2. Build AI-Native. For Real. Not To Check More Boxes.
Don’t build a SaaS product and add AI. Build an AI-native product from day one. The architecture is different. The economics are different. The customer expectation is different.
#3. Rethink Pricing
Seat-based pricing made sense when software helped humans do their jobs better. Consumption-based or outcome-based pricing makes sense when AI does the job instead.
HubSpot is moving toward credits-based pricing for Breeze Customer Agent. Salesforce is experimenting with Agentforce pricing models. The transition is happening.
#4. Solve Mission-Critical Problems
The market is rewarding companies whose software customers literally cannot live without. Security. Data infrastructure. Compliance. Healthcare. Defense.
“Nice to have” software is getting crushed. “Must have” software is thriving.
#5. Target Enterprise or Government
SMB SaaS is harder in 2025. The customers are price-sensitive, AI-curious, and have low switching costs. Enterprise and government customers take longer to close but are stickier and less vulnerable to AI commoditization.
The Contrarian View: These Stocks Might Be Screaming Buys
One analyst note from Morningstar estimates:
- HubSpot is 40% undervalued
- Salesforce is 28% undervalued
- Adobe is 40% undervalued
The median SaaS stock they cover is 30% undervalued. “That is pretty extreme.”
If you believe:
- AI features will eventually translate to AI revenue
- These companies have the distribution and customer relationships to win
- The market is overreacting to AI disruption fears
- Seat-based models will successfully transition to consumption/agent-based models
…then 2025’s losers might be 2026’s winners.
Salesforce, HubSpot, and Adobe aren’t going away. These are great names with extremely strong logo retention. They have cash, distribution, and customer relationships. The question is whether they are adapting fast enough. They’re already adapting.
AI Isn’t a Feature, It’s a Wedge
2025 exposed a brutal truth about B2B software: AI isn’t a feature, it’s a wedge.
Companies that built AI-native, mission-critical, vertical products are winning.
Companies that added AI to existing horizontal SaaS products are struggling.
The market isn’t being irrational. It’s pricing in a future where:
- Seats decline
- Horizontal tools commoditize
- SMB customers switch
- AI-native competitors emerge
HubSpot, Salesforce, Adobe — these are great companies with great products and great teams. They’re not going anywhere.
But the market is telling them: “Show us the AI numbers. Show us the new business model. Show us that AI makes your customers pay more, not less.”
Until they do, the stocks stay under pressure.
The SaaS playbook that worked from 2010-2023 — land, expand, seats, seats, seats — is being rewritten.
2025 is the year we learned what the new playbook looks like.
Now Let’s Talk About the Biggest Stock Price Winners of 2025
It’s easy to focus on the carnage. But the market cap winners of 2025 are equally instructive — and they share a playbook that every B2B founder should study.
Palantir: +142% — The AI Operating System for the Enterprise
Palantir has been the undisputed winner of 2025, and it’s not close.
The Numbers Are Absurd:
- Q3 2025 revenue: $1.18 billion (+63% YoY)
- U.S. commercial revenue: +121% YoY
- Total revenue growth accelerated for nine consecutive quarters
- Rule of 40 score: 114% (that’s not a typo)
- Adjusted operating margin: 51%
What happened? Palantir’s Artificial Intelligence Platform (AIP) became the enterprise standard for deploying AI in mission-critical environments.
Why Palantir Won:
- They were ready when AI hit. While others scrambled to add AI features, Palantir launched AIP in mid-2023 and bet the company on it. When enterprises needed to deploy AI in production — not just demos — Palantir was the only game in town.
- AIP Bootcamps changed the GTM motion. Palantir’s “bootcamps” let customers build working AI prototypes on their own data in hours or days. This “try-before-you-buy” approach collapsed sales cycles and created viral adoption. Walgreens deployed AI-powered workflows to 4,000 stores in eight months. AIG says their Palantir-powered underwriting process is 5x faster.
- Mission-critical positioning. Palantir doesn’t help you write marketing copy. They help you make decisions that matter — military operations, supply chain optimization, fraud detection, clinical trials. When AI hallucinations cannot be tolerated, enterprises choose Palantir.
- Government + Commercial flywheel. The $10 billion U.S. Army contract (over 10 years) provides a stable base. But U.S. commercial is now growing 121% YoY and represents 31% of revenue — up from 23% a year ago.
The Lesson: Palantir proves that AI-native architecture + mission-critical use cases + enterprise/government customers = market leadership. They didn’t add AI to existing products. They built an AI operating system from scratch.
The valuation is insane (70x+ forward P/S). But the growth justifies the multiple — for now.
CrowdStrike: +51% YTD — Security Is the New Mission-Critical
CrowdStrike has quietly become one of the best-performing B2B stocks of 2025, up 51% YTD despite a major outage in July 2024 that rattled customers and investors.
The Numbers:
- Q3 FY26 revenue: $1.17 billion (+22% YoY)
- ARR: $4+ billion
- Falcon Flex ARR: $1.35 billion (3x last year)
- Net dollar retention: 134%
Why CrowdStrike Won:
- Security spending is non-discretionary. CIOs told Morgan Stanley they expect cybersecurity spending to grow 50% faster than overall software spending. When budgets get cut, security doesn’t.
- Platform consolidation is working. CrowdStrike’s Falcon platform now includes endpoint, cloud workload, identity, and SIEM. 49% of customers use 6+ modules. Customers who start with endpoint keep adding.
- The July 2024 outage became a moat. Counterintuitively, the outage accelerated customer commitment. Net retention hit 115% as clients signed commitment packages. Instead of leaving, customers locked in.
- AI-native positioning. CrowdStrike positions Falcon as “AI-native cybersecurity” — and the market believes it. Wedbush calls them a “second-derivative beneficiary of the AI Revolution.”
The Lesson: CrowdStrike proves that mission-critical + platform consolidation + non-discretionary spending = pricing power. Security is the one category where enterprises will pay more, not less, even in a tough macro.
Snowflake: +45-50% YTD — Growth Re-Accelerated
Snowflake had a rough 2024 but staged an impressive recovery in 2025 as the market recognized that data infrastructure is the foundation for enterprise AI.
The Numbers:
- Q3 FY26 revenue: $1.21 billion (+29% YoY)
- Q2 FY26 revenue: $1.14 billion (+32% YoY) — growth actually accelerated
- 7,300+ accounts using Snowflake AI weekly
- AI driving 50% of new bookings
- Net revenue retention: 125%
Why Snowflake Won:
- Growth re-accelerated. After decelerating through 2024, Snowflake’s growth rate inflected back up — from ~26% to 29-32% YoY. The market loves acceleration.
- AI is driving consumption. Cortex AI, Snowflake Intelligence, and AI-powered analytics are now responsible for 50% of new customer acquisitions. AI isn’t a feature — it’s the growth driver.
- Data = AI moat. Enterprises realized you can’t do AI without clean, governed data. Snowflake has the data. Therefore, Snowflake has leverage.
The Lesson: Snowflake proves that data infrastructure companies benefit from AI rather than being disrupted by it. AI needs data. Companies with the data win.
Cloudflare: +80% YTD — The Infrastructure Play Nobody Saw Coming
Cloudflare has quietly become one of the biggest B2B winners of 2025, up 80%+ YTD.
Why Cloudflare Won:
- AI inference at the edge. Cloudflare’s Workers AI lets developers run AI models at the edge — closer to users, faster responses, lower latency. As AI applications proliferated, edge compute became essential.
- Security + performance bundled. Zero Trust security combined with CDN, DDoS protection, and edge compute creates a sticky platform that’s hard to displace.
- Developer-first GTM. While enterprise SaaS companies struggle with AI adoption, Cloudflare’s developer community is already building AI-powered applications on their platform.
The Lesson: Cloudflare proves that infrastructure positioning beats feature positioning. They’re not selling AI features — they’re selling the pipes that AI runs on.
MongoDB: +70% YTD — The Comeback Story
MongoDB had the most dramatic turnaround of 2025. Down mid-year, then two consecutive blowout quarters sent the stock soaring 70%+ YTD.
The Numbers:
- Q3 FY26: Beat estimates massively, stock jumped 23% in a single day
- Atlas cloud revenue: +30% YoY
- Raised full-year guidance significantly
- Over 70% of Fortune 500 now use MongoDB
Why MongoDB Won:
- AI needs document databases. MongoDB’s JSON-based architecture is ideal for storing unstructured AI data — embeddings, vectors, model outputs. As AI applications proliferated, MongoDB became essential.
- Atlas momentum accelerated. The cloud-native version is now 75% of revenue and growing faster than the company average.
- Execution matters. Two consecutive “crushed it” quarters reset expectations and proved the business model is working.
The Lesson: MongoDB proves that data layer companies can be AI beneficiaries — if they position correctly. AI doesn’t threaten databases. AI creates more data that needs to be stored and queried.
Shopify: +51% YTD — The E-Commerce Platform Keeps Winning
Shopify’s 51% YTD gain shows that not all horizontal SaaS is struggling.
Why Shopify Won:
- Mission-critical for merchants. Shopify isn’t a nice-to-have. It’s the operating system for millions of businesses. You can’t just turn it off.
- AI features drive merchant success. Shopify’s AI tools (Sidekick assistant, Magic for store setup) actually help merchants sell more — which means Shopify makes more.
- Platform stickiness. Payments, fulfillment, capital, POS — Shopify owns the full merchant stack. Switching costs are enormous.
- BFCM dominance. Record $14.6 billion in merchant sales over Black Friday/Cyber Monday 2025 (+27% YoY) proved the platform’s continued momentum.
The Lesson: Shopify proves that platform companies with true merchant-level embedding can still win — even in a tough SaaS environment. When you’re the operating system, not just a tool, you have pricing power.
Oracle: +14% YTD — The AI Rollercoaster of 2025
Perhaps the most instructive story of 2025 is Oracle — the 47-year-old database giant that somehow became an AI infrastructure play, hit an all-time high of $346 in September, and then crashed 45% to ~$188 by December.
The Rollercoaster:
- Started 2025 at ~$166
- Hit all-time high of $346 on September 10, 2025 (+108% peak)
- Crashed to ~$188 by mid-December (down 45% from peak)
- YTD return: +14% (still a win, but barely)
Why Oracle Won (Until It Didn’t):
- AI infrastructure positioning. Oracle didn’t try to become an AI company. They positioned themselves as the essential infrastructure backbone for AI workloads. Their database and cloud services became indispensable for companies building AI applications.
- Cloud migration momentum. Oracle’s aggressive cloud transformation finally gained enterprise traction. Customers migrating mission-critical workloads that competitors couldn’t easily displace.
- Larry Ellison’s AI vision. Oracle has been building AI-optimized infrastructure (OCI) that now competes with AWS and Azure for AI training workloads — at better price/performance.
Why Oracle Crashed:
- Expectations got ahead of reality. At $346, Oracle was priced for perfection. Q2 FY26 revenue beat by 38% on EPS but missed on revenue by less than 1%. The market punished mercilessly.
- AI infrastructure is competitive. AWS, Azure, and Google Cloud aren’t standing still. Oracle’s window of AI infrastructure arbitrage may be narrower than bulls hoped.
- The AI trade rotated. December 2025 saw a broader AI infrastructure selloff as investors questioned whether capex spending would translate to revenue fast enough.
The Lesson: Oracle proves both that legacy doesn’t mean left behind AND that AI momentum can reverse fast. The 47-year-old company is still up 14% YTD and positioned in AI infrastructure — but the September euphoria was clearly overdone.
What the Winners Have in Common
Looking across Palantir, Cloudflare, MongoDB, Snowflake, CrowdStrike, and Shopify, clear patterns emerge:
1. AI-Native or AI-Infrastructure (Not AI-Features)
The winners either built AI-native products from scratch (Palantir) or positioned as essential AI infrastructure (MongoDB, Snowflake, Cloudflare). They didn’t just “add AI” to existing products.
2. Mission-Critical Positioning
Every winner solves problems that enterprises literally cannot ignore:
- Security (CrowdStrike, Zscaler, Cloudflare): You can’t turn off cybersecurity
- Data infrastructure (Snowflake, MongoDB, Oracle): You can’t do AI without data
- Operational AI (Palantir): You can’t run a modern enterprise without AI-powered decisions
- Commerce operations (Shopify): You can’t run an online store without your platform
3. Enterprise and Government Customers (or Sticky Platform)
The winners serve enterprises and governments — not SMBs. Or like Shopify, they have such platform stickiness that SMB churn is still manageable. These customers have longer sales cycles but stickier relationships and less price sensitivity.
4. Platform, Not Point Solution
CrowdStrike’s multi-module strategy, Palantir’s operating system approach, Snowflake’s data cloud — winners built platforms that expand within accounts, not point solutions that get replaced.
5. Consumption or Outcome-Based Pricing Potential
The winners either already have consumption-based pricing (Snowflake) or are positioned to move there. They’re not dependent on seat counts that AI might compress.
The 2025 Playbook: What Separates Winners from Losers
Final Thoughts: The New B2B Playbook
2025 taught us that the SaaS playbook from 2010-2023 is being rewritten.
- The old playbook: Land with a point solution → expand with more seats → cross-sell adjacent products → win on integration.
- The new playbook: Build AI-native infrastructure → solve mission-critical problems → serve enterprise/government customers → win on outcomes, not seats.
The ones who took a market cap hit in 2025 — HubSpot, Salesforce, Adobe, ServiceNow, Monday.com, Atlassian — are great companies that are adapting. Breeze, Agentforce, and Firefly are real products with real adoption.
But the markets aren’t patient. Especially in the Age of AI.
The market wants proof that AI makes customers pay more, not less. That AI creates new revenue streams, not just features. That seat-based models can transition to consumption-based models without destroying economics.
The winners — Palantir (+142%), Cloudflare (+80%), MongoDB (+70%), Shopify (+51%), CrowdStrike (+51%), Snowflake (+47%) — already have that proof.
The losers are still building it.
If you’re a founder, the lesson is clear: Don’t build SaaS and add AI. Build AI and add SaaS economics.
If you’re an investor, the contrarian opportunity might be massive: HubSpot at -51%, Monday.com at -36%, Atlassian at -34%, Salesforce at -31%, ServiceNow at -26% — if they successfully navigate the transition, these could be the buys of the decade.
But “if” is doing a lot of work in that sentence.
