CRM Is Not Dead (and Other Enterprise POVs from CRTV Episode 128)
In ConstellationTV episode 128, co-hosts Liz Miller & Holger Mueller tackle the biggest myths in enterprise tech: from "CRM is dead" to Wall Street's AI obsession. Here are the top six stories covered this week:
#1. What Instant Coffee and CRM Have in Common
"CRM is as dead as instant coffee and paper."
That's how Liz Miller opened our latest episode, and it's the perfect analogy for what's happening in enterprise software right now.
When instant coffee launched, manufacturers took out full-page newspaper ads declaring that traditional coffee was dead. The future was instant. The future was convenient. The future didn't require a coffee maker.
Spoiler alert: Traditional coffee survived.
Today, we're seeing the same pattern with CRM. Every week, another think piece declares "CRM is dead" because of AI. Another vendor announces they're "reinventing" CRM with agents. Another analyst predicts the end of Salesforce.
But here's what's actually happening: CRM isn't dying. It's being right-sized.
The center of gravity is shifting. CRM is returning to what it does best: managing transactional conversations. Meanwhile, flexible data platforms (CDPs, data clouds, AI-native systems) are handling the contextual intelligence that AI needs. This isn't death. This is evolution.
#2: Wall Street's AI Problem
During episode 128, Liz asked a question that's been on every enterprise leader's mind:
"Can we stop letting Wall Street define AI success?"
Here's the issue: The current definition of "AI success" is entirely driven by investor sentiment, stock prices, and quarterly earnings calls. If your AI vendor's stock is up, they're winning. If it's down, they're failing.
But that's not how enterprise technology works.
Real AI success looks like:
- Reduced manual work in customer service
- Faster time-to-insight in analytics
- Better predictions in supply chain planning
- Improved employee productivity in day-to-day tasks
Wall Street's version of AI success looks like:
- Revenue growth percentages
- Speculative future valuations
- Hype cycle positioning
- Investor conference soundbites
The problem? When enterprise leaders chase Wall Street's metrics rather than operational metrics, they end up with AI strategies that optimize for press releases rather than productivity.
#3. The OpenAI Warning Sign
Case in point: OpenAI's decision to shut down Sora.
Sora was one of the most hyped AI products of the past year as a video generation tool that promised to revolutionize content creation. Then it launched. And within days, deepfake cartoon creators overwhelmed the system, forcing OpenAI to shut it down.
As Holger pointed out on the show: "As an enterprise CEO, that means: fine that you're doing video or cartoon animation, but I need my level of service here, guys."
This is the enterprise wake-up call. When your AI vendor's infrastructure can be taken down by consumer hobbyists creating cartoon deepfakes, what does that say about their enterprise readiness?
The answer: They need adults in the room.
#4. Canva's Stealth Marketing Automation Play
While everyone's watching the AI model wars between OpenAI and Anthropic, Canva just made a move that could reshape SMB marketing forever.
What happened:
- Canva acquired a marketing automation platform
- Canva acquired a Customer Data Platform (CDP)
- Canva is now positioned to challenge Salesforce, HubSpot, and ActiveCampaign in the SMB market
Why this matters:
For the past decade, marketing automation has been CRM-centric. You started with Salesforce or HubSpot, then bolted on email tools, content tools, and design tools.
Canva is flipping the script: Start with content and design, then add automation and data.
For small and mid-sized businesses, this is a game-changer. Instead of paying enterprise prices for CRM-first platforms they barely use, they can start with tools they already love (Canva) and grow into marketing automation.
The center of gravity is shifting from CRM-first to CDP-first. And Canva just positioned itself at the center of that shift.
#5. Anthropic's "Freeze the Market" Strategy
Here's a model release strategy we haven't seen before:
Anthropic announced a new AI model for cybersecurity. But instead of launching it immediately, they're freezing the market and working directly with cybersecurity vendors to test the model before general availability.
Why?
The model is so powerful that releasing it without proper testing could create security risks. So Anthropic is taking a controlled approach: partner with vendors, run tests, establish guardrails, then release.
As Holger noted on the show, this could become the standard for releasing powerful AI models—not just in cybersecurity, but in any high-stakes domain:
- Self-driving cars
- Financial trading systems
- Healthcare diagnostics
- Critical infrastructure
The "move fast and break things" era of AI might be coming to an end. The "test carefully, release responsibly" era might be beginning.
#6. Oracle's Agentic AI Bet
While the AI model wars dominate headlines, Oracle is making a different bet: agentic AI packages for specific enterprise functions.
At their London and New York events last week, Oracle announced AI agent packages for:
- HR (recruiting, onboarding, performance management)
- Finance (invoice processing, reconciliation, forecasting)
- Supply chain (demand planning, inventory optimization)
- CRM (lead scoring, pipeline management, renewal prediction)
The strategy: Instead of selling you a general-purpose AI model and letting you figure out how to use it, Oracle is packaging pre-built agents for specific business processes.
This is enterprise AI done right. No coding required. No prompt engineering. No "figure it out yourself" documentation. Just: "Here's an agent that handles invoice approvals. Turn it on."
But here's the catch (and Holger called this out): When your AI agents succeed, they stress your backend systems.
Oracle is ahead of the curve here—they're already building database infrastructure that can handle sub-second availability for agent-driven workloads. But most enterprises aren't ready for this.
The timeline:
- 2024: Data foundation for AI
- 2025: Agentic frameworks
- 2026: Backend systems stressed by successful agents
Start planning now.
#7. HubSpot's AEO Tools Go Live
Buried in this week's news: HubSpot's AI-enabled operations (AEO) tools are coming out of beta and going generally available.
What this means:
Every HubSpot user—from free tier to enterprise—will have access to AI-powered automation tools that used to require custom development or third-party integrations.
This is the democratization of marketing automation. What used to cost six figures in Salesforce Einstein licenses is now baked into HubSpot's standard product.
For SMBs, this levels the playing field against enterprises with larger budgets.
For enterprises: This raises the bar for what "table stakes" looks like in marketing automation.
The Bottom Line
Enterprise technology is in a weird transitional moment. The old playbooks don't quite work, but the new playbooks aren't fully written yet.
Here's what is true:
- CRM isn't dead—it's being right-sized for the AI era
- Wall Street shouldn't define your AI strategy—operational metrics matter more
- OpenAI needs to prove enterprise readiness—Sora was a warning sign
- Canva is a serious marketing automation player now—SMBs have a new option
- Responsible AI release strategies are emerging—Anthropic is setting a precedent
The companies that will win in this transition are those that focus on operational outcomes, not on investor sentiment. The ones who right-size their technology stack instead of declaring everything dead. The ones who demand enterprise-grade SLAs from their AI vendors rather than accept consumer-grade infrastructure.
As Liz said: "We need some adults in the room."
📺 Watch the Full Episode
ConstellationTV Episode 128 is available now on YouTube. Subscribe to catch weekly episodes with enterprise tech analysis you won't find anywhere else.
Featured in this episode:
- Liz Miller, VP & Principal Analyst, Constellation Research
- Holger Mueller, VP & Principal Analyst, Constellation Research
- ShortList Spotlight: Chirag Mehta on Elastic's autonomous IT platform
- Event Recap: R "Ray" Wang, Mike Ni, and Holger Mueller recap Infor's Analyst Summit
Resources:
About ConstellationTV
ConstellationTV is a bi-weekly web series hosted by Constellation Research analysts, covering enterprise technology news, trends, and analysis. Episodes air live every other Wednesday at 9:00 AM PT / 12:00 PM ET.