Your ERP Knows What Happened. But Can It Tell You What’s Coming?

Here’s what I’ve been thinking about.

The Boston Manufacturing Group’s 2025 supply chain resilience report confirmed what most of us already know. Last year was defined by disruption. Supply networks reshaped faster than they could stabilise.

“Predicting disruption is impossible, but preparing for it is not”

I’ve sat in too many boardrooms where leadership teams fixate on price fluctuations. Meanwhile, their supply chains fracture beneath them. The conversation tells me everything I need to know about their readiness.

When you’re focused first on tracking cost per unit, and not equally considering on-time, on-quality delivery to your customer, you’re only looking at part of the picture. Here’s the simple truth: price becomes irrelevant the moment you realise you won’t — or worse, don’t — have the materials you need when you need them.

The Measurement Problem Manufacturers Face

Let’s start with what your ERP does well. It tracks transactions brilliantly. Purchase orders, invoices, and delivery dates are all recorded with precision.

But the question your business needs answered has changed:

“Will I have what I need when I need it, and what is the impact if I won’t? Oh my, what can I do about it?”

Most systems can’t answer those questions. They tell you what happened; they can’t tell you what’s coming.

That gap between historical data and predictive, actionable intelligence? That’s the competitive separation emerging through 2026 and into 2027. Supply Chain Management Review states it clearly: “2026 is the year AI agents stop assisting and start operating.”

BCG reports early adopters see 20% to 30% faster workflow cycles through agentic AI native to ERP platforms. That’s the difference between reactive scrambling and intelligence-led anticipation.

Two Layers That Separate Data From Intelligence

You already have the data. Delivery records. Quality metrics. Consumption patterns. It’s buried somewhere in your systems. So why can’t you use it? The problem has two parts.

First: aggregation and accessibility. When operational knowledge is spread across fragmented systems, physical records, and loosely integrated databases, you can’t access it when decisions need to be made. The data exists. The intelligence doesn’t.

Second: accuracy and trust. If you can’t trust the data without manual extraction, consolidation, and validation, you can’t respond fast enough. And you definitely can’t stay ahead of complex processes like production planning or supplier qualification.

Native Intelligence Versus Bolt-On Analytics

McKinsey’s research confirms that AI use cases rely on data and applications housed within ERP systems. But here’s the critical distinction: the ERP platform architecture determines what’s possible.

Native agentic AI works completely differently than a “bolt-on” layer. It inherently understands your business objects and their relationships. It respects the data security, segregation, and processing limits you’ve defined.

Example: Imagine demand shifts and you need to source a component through an RFQ process. Native agentic AI already knows which suppliers are approved. It understands the forms, logic, and workflows. It identifies recipients, creates documentation, automates distribution, and interprets responses, surfacing everything to the right person at the right time.

The Readiness Timeline Through 2027

Research from JustDoers reveals that 78% of IT Leaders expect at least some ERP functionality to be replaced or augmented by agentic AI within three years. 44% anticipate an impact on more than 10% of ERP functionality.

The barrier is twofold:

• Technical reality: Legacy architecture can’t support this level of execution.
• Awareness gap: Leaders still think of AI only as meeting summaries, not as something that reduces stock shortages or maximizes shop floor capacity.

What Readiness Actually Looks Like

BDO’s 2026 manufacturing predictions state that operational agility will shift from competitive advantage to survival requirement. 2026 is the year you need to make yourself ready.

AI amplifies what you’ve systematized. It can’t create intelligence from chaos. The manufacturers who embed their operational knowledge into integrated systems now will have the foundation for intelligence-led operations through 2027.

The Valuation Angle Nobody’s Talking About

If you plan to seek investment or sell your business, having a solid AI strategy is fundamental. It directly affects valuations. Maximizing productivity and quality adds EBITDA and maximizes agility. Native manufacturing agentic AI can shorten process execution timeframes by a factor of 10 while making them more accurate.

The 12-18 Month Window

The competitive window is closing faster than most manufacturers realize. IDC predicts that by 2026, 55% of G2000 OEMs will redesign their service supply chains around AI. They’ll use predictive models to pre-position parts, schedule technicians, and prevent disruptions.

You have 12 to 18 months to get this right.

The risks aren’t slowing down. The Boston Manufacturing Group’s analysis warns that manufacturers face compounding challenges from tariff volatility, reshoring pressures, and fragmented regional supply networks. J.S. Held’s report documents a surge in cyber-attacks targeting logistics infrastructure, with 2025 seeing unprecedented threats to supply chain continuity.

“Last year was about managing disruptions. Right now, it’s about redesigning your global network.”
— CEO of the Association for Supply Chain Management

Disruption is the norm now. Preparation is the only strategy that works. So, here’s the choice in front of you.

Your ERP can remain a record of what happened. Or it can become the intelligence infrastructure that tells you what’s coming and helps you respond before the problem arrives. But only if you’re on a platform architected to make that transformation possible.

The manufacturers who understand that distinction in the next 12 to 18 months will separate themselves from the ones still waiting for supply chains to stabilize.

They’re not going to stabilize. And deep down, you already know that.