SaaS Gets Smoked: CRM, ADBE, NOW Red Because AI (And AI Capex Too)

TL;DR

• Software names are getting clipped because the market is scared AI turns parts of their product into a commodity.
• At the same time, the market is scared the AI spending binge is a money pit.
• So you get the cursed combo trade: "AI changes everything" plus "AI capex might be for nothing."
• If you are holding SaaS, the next catalyst is simple. Show real AI revenue, protect margins, stop lighting cash on fire.

What Just Happened To Software

You probably saw the red day where the usual suspects got dog-walked.

• NOW was down about 7.9%
• ADBE was down about 7.0%
• CRM was down about 6.8%

When a whole cohort sells off together, it is not because everyone read three different earnings reports and became a value investor. It is positioning and narrative.

The current narrative is AI.

AI Fear #1: "Will I Still Need This Software?"

This is the scary part for software bulls, because it is not just a macro story. It is a product story.

Here’s the bear case in plain English:

• If AI can do the work, fewer humans need seats.
• If fewer humans need seats, subscription growth slows.
• If subscription growth slows, your fancy multiple gets shaved like you joined the Marines.

CRM (Salesforce)

CRM lives on seats and workflows. AI agents are supposed to make sales reps faster, which sounds great, until the market asks:

• why do I need 100 reps if AI makes 70 reps feel like 120
• why do I need another add-on if the AI assistant becomes the interface

Bull counter:

• CRM owns the data and the enterprise relationships. The AI agent still needs the system of record.
• If they sell "AI agent" add-ons that customers actually pay for, this flips from threat to upgrade cycle.

ADBE (Adobe)

ADBE is in the crosshairs because generative tools attack the price people pay for creative software.

Market fear:

• AI makes basic design and editing cheap
• cheap tools flood the market
• pricing power gets questioned

Bull counter:

• pros want control, workflow, and rights management. The "make me a logo" button is not the same as production-grade tools.
• Adobe bundling AI into the suite can keep users locked in, even if the margin math gets messy.

NOW (ServiceNow)

NOW sells workflow and automation. AI looks like both a competitor and a turbocharger.

Market fear:

• "agentic" tools bypass traditional ticketing and workflow
• companies simplify stacks, fewer vendors, less spend

Bull counter:

• enterprises are not ripping out workflow plumbing overnight
• if AI increases automation demand, NOW sells more because it is already the system tying everything together

AI Fear #2: Inference Costs Are A Sneaky Margin Tax

Even if AI is good for user adoption, there is a cost nobody wants to talk about on bullish days.

Inference is not free.

If a software company ships AI features and bundles them into the base plan to stop churn, they can end up with:

• more usage
• higher compute bills
• worse margins

Then the market hits them with the "congrats on growth, enjoy your profit margin haircut" treatment.

The Other Side: The AI Capex Hangover

Now the second fear that is also AI, but from the opposite direction.

Everyone is building data centers like it is a competitive sport. Great for the shovel sellers, scary for anyone paying the bill.

The market worry is not that AI is fake.

The worry is this:

• capex goes vertical now
• revenue payoff is slower than promised
• free cash flow gets squeezed
• investors stop paying premium multiples for "future cash flows"

So software gets hit twice:

• disrupted by AI
• forced to spend on AI to not get disrupted

What The Market Is Pricing In Right Now

It feels like the tape is pricing:

• slower seat growth
• more pricing pressure
• higher costs to ship AI features
• less patience for "trust me bro" roadmaps

That is why the whole group can sell off together even when the companies are different.

How This Ends (Three Likely Paths)

1) AI Monetization Works

Software shows real paid adoption of AI features and keeps margins from collapsing.

This is the good ending. Multiples stabilize, the group rallies, everyone pretends they were always long.

2) AI Becomes A Feature, Not A Pricing Lever

Everyone ships copilots, nobody charges much, compute costs creep up, growth does not accelerate enough.

This is the boring ending. Stocks grind sideways, analysts fight over spreadsheets, your calls decay in peace.

3) Capex Overbuild Shows Up

If the capex binge turns into low ROI, the market will keep punishing anything that depends on a high-growth, high-multiple story.

This is the bad ending. Software stays under pressure until expectations reset.

The Setup To Watch

If you want a clean "trend is back" signal, you want to see two things:

• guidance that proves AI is a revenue driver, not just a demo feature
• cost discipline so AI does not nuke margins

Until then, software is going to keep trading like a debate, not a business.