I attended Directions NA in Orlando and DynamicsCon in Las Vegas within weeks of each other this spring. Between them, the real takeaway wasn’t about a single product announcement or product feature release.
It was about trust.
Trust between partners and customers. Trust between ISVs and VARs. Trust between organizations and the AI capabilities landing in their systems. It’s a collective questioning of whether the ecosystem can actually deliver on what it’s currently promising.
AI has raised the bar. Now, the harder conversation is whether the foundation underneath it is ready.
At Directions NA, Microsoft made its position clear: agents, Copilot, and embedded intelligence are the direction of Business Central. From sales and payables to expense tracking, the roadmap is aggressive. DynamicsCon echoed this momentum with over 300 sessions heavily focused on AI-powered automation and agent orchestration.
But the sessions that actually drew energy in the room weren’t the ones announcing what AI can do. They were the ones confronting what’s in the way.
One speaker, at Directions NA, laid out three walls organizations hit with AI:
The Reframe: Stop treating AI like Google, and start treating it like an intern doing improv. You don’t just query an intern for a one-word answer; you set the scene, give them context, challenge their output, and iterate with them. Most organizations are hitting a wall because they expect a search engine, missing the reality that AI requires a continuous, collaborative dialogue to provide real value.
Another session covered how technology is moving fast, but the knowledge layer underneath it- a deeper understanding of BC, actual business processes, and domain context is where things are breaking down.
DynamicsCon told the same story from the user side. Sessions highlighted why Copilot gives incorrect answers when underlying data is dirty, and why securing Business Central for AI before you “flip the switch” is critical. Years of inherited permission sets and overly permissive user roles suddenly become massive exposure risks the moment an AI agent starts querying across systems.
Trust can no longer be seen as a feel-good concept. It is the actual mechanism that determines whether partnerships, implementations, and AI adoption succeed or fail.
Neural Impact’s keynote during Directions NA framed this structurally: ERP funding is down globally due to gaps in customer experience, perceived value, and unresolved legacy issues.
Fixing one problem does not equal customer loyalty. The old model ended at go-live; the new model starts there. Partners who shift from “deliver and done” to “deliver and advise” are the ones keeping customers:
Another clear and direct message is that the partner business model is shifting, and partners who don’t adapt will face a significant revenue decline.
Traditional implementation isn’t going away, but new service categories are rapidly emerging around it: acceleration, adoption, ongoing optimization.
The partners winning today aren’t just delivering a successful go-live and walking away. They are staying deeply engaged throughout the entire software lifecycle, acting as long-term advisors, and helping customers continuously realize value long after the initial deployment.
This maps directly to what I see every single day. The ISVs and VARs that build real partnerships and shared accountability for customer outcomes are the ones whose implementations stick and whose customers continuously renew, expand, and refer.
Another practical session I attended was an honest rundown of implementation realities:
The session covered ten principles, but they all boiled down to a few raw truths:
Ultimately, trust isn’t built by avoiding the hard conversations. It’s built by having them early, documenting everything, and refusing to treat go-live as the last milestone.
The Microsoft Dynamics ecosystem is in the middle of a confidence pivot. AI has amplified expectations for speed and visibility, but high expectations without a strong foundation create fragility.
The common denominator across every failure and success is trust: in the tech, in the partner relationships, in the data, and in organizational readiness.
As we head into the second half of 2026, the conversation is shifting from “Are we ready?” to “Prove it.” We’ll see exactly who has built a foundation strong enough to support the future when the community gathers again later this year.
Authored by: Kayla Magnan