AI, CRM Strategy, GTM Operations, Salesforce

What Mile High Dreamin’ Taught Us About AI: Why Agentforce Exposes Weak GTM Foundations

AUTHOR

Kayt Parrella

Published

October 8, 2026

Read Time

5 min read

Agentforce won’t fix your GTM architecture. It will expose it.

I left Mile High Dreamin’ this year with a full notebook and one thought I couldn’t shake.

The energy around Agentforce was everywhere. It set the tone from the opening keynote, where Joshua Birk and Kate Lessard talked about how AI is changing the way Salesforce professionals work and how the community can prepare for that shift. It carried through the breakout sessions, the Demo Jam, and every hallway conversation between them.

But the conversations that stuck with me weren’t about what AI can do. They were about what happens when teams try to use it. Over coffee and at the happy hour, I heard the same story from RevOps and GTM leaders again and again: we turned it on, and it didn’t work the way the demo did.

That’s the real lesson from Mile High Dreamin’, and it’s one I think every revenue leader needs to hear before their next AI investment. Agentforce won’t fix your GTM architecture. It will expose it.

Why the Community Conversations Mattered Most

What makes an event like Mile High Dreamin’ valuable is that it’s community-led. People don’t come to be sold to. They come to share what’s actually happening inside their orgs, even the messy parts. Sessions like “Salesforce Therapy: Live Troubleshooting for Real-World Hurdles” were packed for a reason. Practitioners want honest answers to real problems.

And the honest answer about AI is this: an AI agent doesn’t create order. It amplifies whatever it’s given. Point it at clean data, connected systems, and well-defined processes, and you get leverage. Point it at a fragmented GTM stack, and you get bad decisions made faster and at scale.

At 28 North, we spend our days inside the revenue engines of growth-stage companies. The pattern we see matches what I heard in Denver. The organizations getting real value from AI aren’t the ones who moved first. They’re the ones who built the foundation first.

The Three Gaps Behind "It Didn't Work Like the Demo"

When I dug into those hallway stories, nearly every one traced back to one of three gaps.

Data you can’t fully trust. Duplicate accounts. Lead sources that mean different things to marketing and sales. Opportunity stages updated on Friday afternoon before the forecast call. Your team has learned to work around these issues without realizing how much judgment they apply every day. AI doesn’t have that context. If your pipeline data is inconsistent, an agent will confidently give you the wrong answer.

A disconnected GTM stack. Most growth-stage companies don’t run on a single platform. Marketing lives in HubSpot or Marketo, sales lives in Salesforce, and intent data, engagement tools, and delivery systems sit in between. An agent working inside Salesforce only sees what Salesforce sees. If the signal that matters most lives somewhere else, your agent is making decisions with one eye closed.

Processes that were never defined. This is the gap teams underestimate most. Many GTM processes run on tribal knowledge. The marketing-to-sales handoff works because two people get along. The renewal motion works because one account manager keeps a great spreadsheet. AI can’t automate what hasn’t been defined, and if you automate an undefined process, you lock in its inconsistencies.

Bring This Checklist to Your Next Team Meeting

If you came home from Mile High Dreamin’ excited about AI, channel that energy into an honest readiness check first. Ask your team:

  • Can we agree on a single definition for our core pipeline metrics, and does our data reflect it?
  • Do we know which system is the source of truth for each stage of the customer lifecycle?
  • Are our key handoffs (marketing to sales, sales to delivery, delivery to renewal) documented and consistently followed?
  • Could a new hire understand how our revenue engine works from our systems alone, without asking around?
  • Do we have clear ownership for data quality, not just system administration?
  • Have we identified a specific business outcome we want AI to improve, rather than a feature we want to use?

If you answered “no” or “not sure” to more than two, you’re not behind. Based on what I heard in Denver, you’re in very good company. But you’re also not ready to get full value from AI yet, and that’s worth knowing before you invest.

Keep the Conversation Going

The best part of Mile High Dreamin’ is that the learning doesn’t stop when the event ends. The Salesforce community is at its best when we’re honest with each other about what’s working and what isn’t, and AI is exactly the kind of topic where that honesty matters.

That’s the spirit we bring to our work at 28 North. We look across your entire GTM stack, align the tools to your strategy, and build the architecture that makes AI useful instead of risky. And we stay involved after go-live, because adoption and iteration are where the real value shows up.

If one of those hallway stories sounded familiar, let’s pick up the conversation. Reach out to me directly or connect with the 28 North team. And if you were at Mile High Dreamin’ too, I’d love to hear what you took away from it.

Ready to Make AI Work for Your GTM Engine?
Don't let a fragmented GTM stack or untrusted data hold back your technology investments. Reach out to connect with the team and find out how ready your revenue engine really is.

AI, CRM Strategy, GTM Operations, Salesforce

Share