My friend said this to me yesterday:

At BMW, all the cars that leave the factory are perfect. At Toyota, all the cars that leave the production line are perfect.

This is about as elegant a description of built-in quality that I've ever heard.

Attention to built-in quality is relatively visible in manufacturing processes, but it seems to me that it's vanishingly rare in office processes. Most corrections and rework are done at the end of a given process, when it's more difficult and more time-consuming to fix. In particular, given the ubiquity of, and reliance upon, spreadsheets, this is a real issue: a small error in only one cell propagates throughout an analysis and a business plan.

When my colleague Karen Martin teaches metric based process mapping, she emphasizes the importance of determining the "percent complete and accurate" data for each step in a process. This information reveals the (horrifyingly) low efficiency of the overall process. Equally important, though, this information helps us focus on the points in a process where we need to build-in quality.

What checks can we install to ensure that the information we send to our downstream colleagues is 100% complete and accurate? What error-proofing tools can we create? It's not that hard to insert data validation tools into a spreadsheet or a web form. Checklists and templates are other tools that can help ensure built-in quality.

It's time to raise the standards of excellence for the internal data that you manipulate to the same level as the product you deliver to your customers.