This post was originally published by Saanya on her Substack, 'The Change Constant' – subscribe HERE!

When the macro is a bloodbath, focus on the micro.

We cannot control the fact that the market is down year-to-date and looks determined to keep testing everyone’s pain tolerance. But we can control how we adapt to what enterprise buyers are actually thinking right now. Here’s what I’ve picked up in the last few weeks from conversations with decision-makers:

One-year contracts only. No one’s committing long-term.

Tech is evolving too fast, prices are dropping too quickly, and no enterprise wants to be the sucker locked into an overpriced, obsolete contract. If you're selling, expect pushback on multi-year deals unless you can prove long-term defensibility.

Indemnity is king.

Turns out, when an AI agent hallucinates, enterprises really don’t want to be the ones on CNBC explaining why their chatbot recommended Tide Pods as a balanced breakfast. If an AI agent goes rogue, companies want someone else to take the heat – preferably a trillion-dollar scapegoat, not a Series A startup. A major reason enterprises pick Google/Microsoft over a hot new AI startup? CYA.

If an agent hallucinates, they want someone to point the finger at. If you’re an emerging AI vendor, selling accuracy isn’t enough - you need a plan for how enterprises can trust you without betting their reputation on it. Consider how you de-risk adoption beyond “trust us.”

Fragmentation is a deal-breaker.

In a world where AI agents are hitting parity, the differentiator isn’t just performance – it’s integration and completeness of platform. Big Tech wins here: They bundle agents with analytics, automation, and workflow tie-ins.

If you're an emerging player, the short-term wedge is outperforming them, but the long game? Build integrations. Race to platform completeness.

Price for quality, not just volume.

Enterprise buyers aren’t just measuring usage; they’re pricing based on performance. One example: If an AI agent’s NPS score drops below 3.5, the vendor doesn’t get paid. Automatic quality control is being baked into pricing models. If you're building, make sure you're tracking (and optimizing for) more than just scale.

TL;DR - Enterprises are cautious, risk-averse and optimizing for flexibility. Build your GTM and sales strategy around this reality.