The data-model test

The data-model test is a two-week, time-boxed proof run before buying or building quoting software: take your most difficult manufacturer, two product series, one historical price list and one current update — photos, options, and missing values included — and push them through the candidate system's data model. The acceptance test is what comes out: a clean export you can open elsewhere, uncertain and missing values surfaced rather than hidden, and a reversal, with the update undone.

Why it matters

The strongest predictor of a quoting-software decision going wrong is not the price of the software — it is a data model imposed on a catalog that does not fit it. A model that fits shows it in two weeks. A model that does not shows it in month three of a rollout, at a very different price.

Common confusion

The test is not a demo. A demo runs on the vendor's clean data and always passes. The test runs on your worst data, and its most important output is how the system treats the values it is not sure about.

Where we use it

Step three of the calculation in Enterprise CPQ vs custom — and it applies unchanged to an ERP's built-in offer module.