March Madness, AI Agents, and Why Most ERP Teams End Up in the “First Four Out”

Every March, teams across the country (and most of us) obsess over brackets, bubble teams, and one painful distinction: the First Four In vs. the First Four Out.

The difference between those two groups isn’t raw talent. It’s preparation, fundamentals, and execution when it matters most.

That same dynamic is playing out right now with AI agents in ERP, especially in Microsoft Dynamics 365 Business Central.

Everyone wants to “be in the tournament.” Very few are building the foundation required to make a deep run.

The Bracket Isn’t the Problem, its the fundamentals…

I read that almost 95% of AI initiatives fail before the first tip‑off, not because the technology is immature, but because organizations skip the early rounds:

• Data foundation
• Use‑case definition
• Change management

They jump straight to demos and copilots.

That’s how teams end up on the bubble… and then watching from home.

Round 1: The Data Foundation (Win Here or Go Home)

In March Madness, teams that can’t rebound or defend don’t survive long. In AI, bad data is the equivalent of bad defense.

For Business Central customers, the fundamentals matter more than ever:

• Inconsistent dimensions
• Duplicate vendors and customers
• Manual journal workarounds
• Inventory data that “mostly ties out”

AI agents don’t fix these problems. They amplify them.

First Four Out mistake:

“We’ll clean up the data after the AI agent is live.”

First Four In discipline:

“We define ownership/accountability, normalize dimensions, and trust the data before the agent touches it.”

If your ERP data can’t answer basic operational questions confidently, an AI agent won’t magically do it for you.

Round 2: Defining the Right AI Use Case (Stop Taking Half‑Court Shots!)

Not every process deserves an AI agent. And not every agent needs to be autonomous.

Strong teams don’t jack up low‑percentage shots. They run plays they’ve practiced.

The best Business Central AI use cases are:

• Narrow
• Measurable
• Tied to a clear business outcome

Examples:

• Exception handling in AP
• Forecast variance explanations
• Inventory risk alerts
• Sales order anomalies
• Margin erosion detection

First Four Out mistake:

“We want an AI agent that understands our entire business.”

First Four In discipline:

“We want an AI agent that solves one painful, repeatable problem and proves value in 30–60 days.”

Winning teams don’t try to win the championship in the first round.

Round 3: Change Management (This Is Where Upsets Happen)

Every tournament has a favorite that gets upset. In AI projects, that upset is almost always people, not the technology.

AI agents change:

• How work gets done
• Who makes decisions
• What “good performance” looks like

Ignoring that reality is how promising initiatives stall.

First Four Out mistake:

• No role clarity
• No training plan
• No explanation of why the agent exists
• Quiet resistance from users who feel replaced or bypassed

First Four In discipline:

• Clear ownership and escalation paths
• Human‑in‑the‑loop design
• Training focused on trust, not features
• Leadership reinforcing that AI augments judgment as it doesn’t replace accountability

Teams that manage change well don’t just survive the first weekend. They build momentum.

The Final Four Mindset: AI as a System, Not a Feature

The organizations that make deep runs with AI in Business Central treat it as:

• A data strategy, not a plugin
• A business capability, not a demo
• A journey, not a launch event

They invest in:

• Strong ERP fundamentals
• Thoughtful use‑case sequencing
• Change management that respects how people actually work

That’s how you stop chasing hype and start winning consistently.

Final Thought

In March Madness, talent gets you noticed. Discipline gets you wins.

AI in Business Central is no different.

If you’re serious about being First Four In and not another early exit, start with the fundamentals. The bracket will take care of itself.