Access is the precondition.
You can’t do analytics on data you can’t reach.
Every manufacturer we meet wants better numbers. Most assume the obstacle is analytics — the right tool, the right dashboard, the right hire. It usually isn’t. The obstacle is earlier and more boring: they cannot reach their own data. Analytics is not the first problem. Access is.
Where the data actually sits
In a 50–500 employee shop, the operating truth is spread across the ERP and the systems around it — MES on the floor, a quality system, a CRM, and the spreadsheets that quietly hold the rest together. The advantage is in combining them. The problem is that most of them sit behind something you don’t control:
| The barrier | What it means in practice |
|---|---|
| A VAR’s portal | Your reports live inside a reseller’s tool. You see the number; you cannot see the query, export the model, or change what it means. |
| A proprietary schema | The data is technically yours, but structured so that only the vendor’s tools can read it without a translation you don’t own. |
| A refresh path you don’t control | Someone else scheduled the job that moves the data. When it breaks, or when they leave, you wait. |
| The surrounding systems | MES, quality, and CRM each hold a piece of the picture behind their own login. Nothing conforms them into one place you can query. |
Why the trap tightens
Here is the part that keeps good operators stuck. The two parties best positioned to help — the ERP vendor and the VAR who implemented it — both make money the same way, and both point in the same direction. The fix is always another module, another seat, another year on the platform. The advice may even be sincere. But when the ERP vendor and the VAR iterate the same message, you have two voices and no independent counsel.
That is the alternative we are. We don’t sell the ERP. We don’t sell the VAR’s roadmap. We have nothing on the other side of the recommendation, which is what lets us say the thing neither of them will: get access to your own data first, and own it. Then the analytics become possible — and so does the multiplier at exit.
Access first, then the rest is achievable
The ordering is not a preference; it is a dependency. Each step is impossible without the one before it:
- Reach the data — the ERP and the systems around it — in a place you control.
- Define the numbers once, so the reports stop disagreeing.
- Build the trusted model and the operating views on top of those definitions.
- Then, and only then, point AI at it — because AI on data you can’t reach just automates the guess.
Skip the first step and every later one inherits the lock-in. This is the same argument as Own the platform, not the portal, stated from the front end: there, the point is that analytics welded to the ERP deepens a dependency your next owner inherits. Here, the point is that the dependency blocks you from ever starting.
Lock-in caps the multiple
The cost isn’t only a slower Monday meeting. The data you can’t reach is also the data a buyer can’t verify. A business whose numbers live behind a vendor’s portal presents to diligence as a story to be taken on faith; a business that owns its data and its definitions presents as a system a buyer can inspect. One of those earns a turn on the multiple, and the turn is worth more than the earnings it sits on. The full arithmetic is in the exit math — but the headline is short: lock-in doesn’t just slow reporting. It caps the multiple.
The audit starts exactly here. Before any model gets built, it establishes what you can and cannot reach today, who controls each refresh path, and what it would take to own them. You keep that map either way.
The audit prices this before you spend a dollar.
Two weeks, $3,000, fixed scope. A graded report on how your ERP is actually used, where the numbers disagree, and a priced recommendation — you keep the report either way.