2025 Revenue AI in Review: What changed, what stayed human, and what I’m taking into 2026

As we hit “send” on 2025, I’ve been thinking about how quickly the Revenue AI conversation matured this year.

We’ve actually had AI in go-to-market for a while, albeit mainly in the form of predictive AI. Think intent signals, scoring, and prioritization. Tools like 6sense brought that to many teams.

But here’s the thing: it didn’t always feel like AI. It felt like… a number. A ranking. Another dashboard.

The tech didn’t suddenly become useful in 2025. The interface became understandable.

Generative AI made it click because it made AI visible. You can ask it a question and get something back that looks like real work, like an email, a summary, a talk track, or a point of view.

Predictive AI quietly helped us choose where to spend time. Generative AI is trying to help with how we spend time.

And regarding “humans in the loop,” if anything, the human part continued to grow in importance: judgment, trust, and execution still decide outcomes.

But overall, in 2025, the question shifted to something much more practical (and much more interesting): How do we operationalize AI, so it actually helps our teams win ...without compromising trust?

Here are a few milestones I saw across the Revenue AI landscape this year, plus a moment of gratitude for the people who make go-to-market work.

1) Revenue AI moved from “insights” to “actions”

The early wave of AI in sales was incredible at summarizing (calls, emails, deal notes, account research).

In 2025, the bar moved. The market started expecting AI to do more than narrate reality. The expectation became: turn signal into next-best action.

That shift sounds subtle, but it changes everything:

• “Here’s what happened” became “Here’s what to do next.”

• “Here’s a forecast” became “Here are the deals at risk—and the intervention that typically saves them.”

• “Here’s a playbook” became “Here’s the playbook, applied to this account, this buyer, this moment.”

If you felt your tech stack getting more proactive this year, you weren’t imagining it. Revenue AI started showing up as workflow, not just analytics.

2) “Agents” started to get real (and so did the hype)

This was also the year “agent” became everyone’s favorite word.

At its best, agentic Revenue AI means an AI system can take a goal (“move this opportunity forward”), break it into steps, use the tools you authorize (CRM, enablement, engagement, support systems), and execute in a controlled way – with a human in the loop where it matters.

At its worst, “agent” was just a new label slapped onto the same old features.

The big milestone in 2025 wasn’t that agents solved everything. It was that the industry began separating 'agent theater' from 'agent utility'. The teams that made progress did two things consistently:

1. kept agents scoped to specific revenue tasks (not “run my entire business”), and

2. built guardrails around data, approvals, and accountability.

3) Data quality stopped being a back-office topic

If 2024 was the year everyone added AI tools, 2025 was the year many teams realized: AI doesn’t fix messy revenue data — it amplifies it. 😧

A lot of the most meaningful “Revenue AI wins” I saw this year didn’t start with a new model. They started with unglamorous work:

• 🧹 cleaning opportunity stages and definitions

• 📞 improving call/meeting capture

• 🤝 tightening handoffs between SDR → AE → AM/CS

• 🛜 aligning buying intent and product usage signals with commercial workflows

• 📜 documenting what “good” looks like in a repeatable way

🤔 💭 In other words, the teams that treated data and process as first-class citizens got better results from AI – often with fewer tools.

4) RevOps and Enablement became AI operators

A quiet but significant shift: the people who run revenue operations, enablement, and systems increasingly became the “product managers” of Revenue AI inside the business.

In 2025, the best operators I know weren’t asking, “What can AI do?”
They were asking:

• “Where do we have friction in the revenue motion?”

• “What decision is made too late, too inconsistently, or with incomplete data?”

• “What should be automated vs. what must stay human?”

• “How do we measure if this is actually working?”

This mindset turned AI from a novelty into an operating capability.

5) Trust, governance, and AI literacy moved to the center

Buyers got sharper. Legal and security reviews got deeper. Internally, teams started asking smarter questions about:

• permissioning and data access

• hallucinations and overconfidence

• audit trails and explainability

• what “human review” should look like (and when)

The milestone here is simple: trust became a feature, not a footnote. And honestly, that’s a good thing. Revenue AI is too important to be treated casually.

A moment of gratitude for the humans behind the motion

Go-to-market can be hard. Not in a dramatic way – just in the steady, daily way that comes from juggling targets, buyers, process, pressure, and change.

💪💫 So on this last day of the year, I want to say thank you to:

• our reps doing the real work of building trust with customers

• our leaders coaching through ambiguity

• the RevOps teams who keep the engine running

• the enablement leaders who turn strategy into behavior

• marketing, SI and tech partners, customer success, professional services, solution consultants, and partners who make the motion coherent

• and the broader community of peers who share what’s working (and what isn’t)

🏆 Revenue outcomes are always a team sport. 2025 reinforced that for me.

Looking ahead to 2026: a few bets I’m making

If 2025 was the year Revenue AI became operational, I think 2026 will be the year it becomes organizational.

🔮 A few forecasts I’m watching:

1. AI will become the default interface for revenue work (not a separate tool).

2. The competitive advantage will shift from “having AI” to “running AI well” (think clean and connected data).

3. Agent ecosystems will push teams to simplify stacks and standardize workflows.

4. Measurement will mature: more controlled experiments, clearer baselines, and fewer vanity dashboards.

5. Human judgment will matter more, not less – because trust, strategy, negotiation, and creativity don’t automate cleanly.

If you’re heading into 2026 planning your roadmap, my simplest suggestion is this: start with the motion. Map where decisions happen. Map where work stalls. Then decide what AI should accelerate.

🎆 Happy New Year to everyone building, selling, implementing, enabling, operating, supporting, and leading. 💪 I’m grateful for the community, and excited for what we’ll create next. 🙌

❓ So tell me, what was your biggest Revenue AI lesson in 2025, and what are you prioritizing for 2026?