In a recent blog post, Reinventing the Business Intelligence Process, Michael Vizard highlights the importance of adding ntext to data as it is llected rather than after the fact. And preferably in a reasonably automated way. So it’s easy to fet that this is something that XBRL, and indeed any taxonomy-based data tagging paradigm, can be used for.

This number - 100 – is merely data. It means whatever you want it to mean because it has no ntext associated with it. Tagged with a bunch of XBRL tags our ‘100′ uld now be clearly defined as the pre-tax profit in a specific currency for a specific mpany, in a specific period of a specific fiscal year. Problem is, as Michael points out, this ntextual tagging is being added after the fact. By XBRL tagging software like ours, for example. Adding ntext to data is the hidden burden of gaining insightful information.

But what if ntext was being generated as the data is being generated? Say you run a simple financial statement summarizing the inme and expenses of your business. And say these rollup totals are already ntextualized with XBRL tags. Then whatever calculates the ‘profit’ from the of the two can automatically tag the resulting profit or loss with the appropriate ntextual tags – assuming it is XBRL-savvy, taxonomy-aware and rule-driven that is.

One of the possibilities of XBRL is precisely to facilitate the automated generation of data ntext so that business intelligence can beme just that, without depending on burdensome ‘after-the-fact’ tagging.