AI Is Not Boosting Big Software — It’s Eating It

Let’s look at what actually happened on the stock market last week.

The large enterprise software players published their usual quarterly earnings. And on the surface, everything looked fine. 

Strong Results, Brutal Market Reaction

• Revenue was solid

• Growth was present

• Profits beat expectations

And then the market destroyed them. 

• Microsoft: good results, stock down 11%, roughly $400 billion erased 

• Salesforce: still growing, stock down 5.6% 

• ServiceNow: record profits, stock down 11% 

• SAP: beat expectations across the board, stock down 16% 

Taken together, the enterprise software sector is now in a bear market — down 22% from the peak.  Which is absurd, if you listen to the old narrative. These were supposed to be the AI winners. They spent the most on AI. Acquisitions. Massive teams. Exploding HR costs. New workflows. New roadmaps. Everything screamed “AI-first”. And yet the market did something very different. 

Because it finally priced in the uncomfortable truth that everyone had been dancing around: 

AI doesn’t strengthen traditional software companies. It replaces them. 

The Real Problem Is the Business Model

Nothing is “wrong” with the numbers. 

What’s wrong is the assumption that the old enterprise software business model still makes sense in an AI-native world. 

If an AI can: 

• write production-ready code, 

• assemble applications on demand, 

• Automate entire workflows,

then why would a company keep paying $50,000 per year for a single software license?  An AI agent replaces 10 users. few prompts replaces months of development.  A phone-based LLM can wipe out entire operational departments.  At that point, you’re not optimizing software spend — you’re questioning why the software exists at all.  This is already happening. Companies for example started to remove Salesforce from their stack entirely and rebuilt around AI-driven systems. 

That’s not efficiency. That’s displacement.

What the Market Is Actually Pricing

This is the key shift many people miss. Markets no longer price revenue, growth, or even profit in isolation. They price whether a business model has a future. And right now, investors are asking a brutal question: Does this company’s core value still exist five years from now?  That puts enterprise software companies in a trap with no clean exit: 

• If they don’t invest in AI, they become irrelevant. 

• If they do invest in AI, they cannibalize their own products. 

This isn’t a classic innovator’s dilemma. It’s closer to a structural dead end. 

Why Hardware Is Winning While Software Is Bleeding

There’s a reason software stocks are falling while hardware and infrastructure are flying. Capital is pouring into: 

• GPUs 

• semiconductors 

• memory 

• storage 

• cloud capacity 

• energy 

AI needs compute and electricity. It does not need enterprise subscriptions. Just like in every gold rush, the money flows to the people selling the pickaxes and shovels, not the ones claiming to own the land. And in this cycle, traditional software licenses are neither. 

The New Power Layer: AI Consultancies

Those who understand organizations, not just agents. This is where the real opportunity emerges.The next winners won’t just be “AI-native” companies. They’ll be AI consultancies that understand how companies actually work. Knowing agentic models is table stakes. Everyone will claim that.  What actually matters is understanding: 

• how a company is structured, 

• how decisions flow, 

• how roles, incentives, and accountability are designed, 

• and how all of that must change once AI agents are introduced. 

Because agentic AI doesn’t just automate tasks — it reshapes organizations. When you introduce agents: 

• some roles disappear, 

• others merge, 

• decision-making moves closer to execution, 

• management layers collapse, 

• and responsibility shifts from humans to systems. 

Most enterprises are not ready for this. Replacing a CRM or ERP with agents isn’t a technical migration. It’s: 

• organizational redesign, 

• change management, 

• role redefinition, 

• and cultural transformation. 

This is where classic software vendors fail — and where serious AI consultancies win. 

The new “golden era” belongs to consultancies that can: 

• redesign workflows around agents, 

• re-map roles and responsibilities, 

• help leadership understand what humans should still do, 

• and safely integrate agentic systems into real, messy organizations. 

Not slides. Not demos. Not copilots bolted onto old tools. Real structural change.

This Is a Reset, Not a Cycle!

The AI shock isn’t limited to creatives. 

It’s not limited to developers. It hits: 

• coders, 

• the code itself, 

• and the corporations built on owning and licensing that code. 

We’re watching the subscription software model being repriced in real time. 

And on the other side of this reset, value won’t sit with the biggest platforms — it will sit with the people who know how to rebuild companies around agentic AI. 

That’s the shift. 

And it’s already underway.