THE BOX MODEL

When was the last time you saw a conference session on technology — topics like blogging, podcasting, and predictive modeling?

Right.  Probably the last time you attended a conference.  Tech sessions are everywhere.

But when was the last time you saw a session about process improvement?  Probably never.

That’s a huge problem.  Here’s why:

Case in Point

A couple years ago, the University of Rochester moved to an online application review system.  The first thing I did, before speaking with vendors or writing a single line of code, was call a meeting to get a better sense of the entire process from end-to-end.

The operations manager discussed mail processing, folder creation, completion criteria, unmatched item handling, completion sweeps, and folder checkouts.

The application review manager discussed first review, second review, committee criteria, school equity review, merit/scholarship review, and special population review (early decision, combined-degree applicants, VIP students, legacies, athletes, etc.).

With very few changes, that had been the standard file handling and review process for years.

And, obviously, it worked.  But because it worked, it was hard for anyone to imagine there could be flaws.

Yet when I drew everything out on a whiteboard, the gaps, redundancies, and inefficiencies became fairly apparent.

For example, 5 people might review a file when only the last reviewer’s decision mattered.  We had cycles in the graph where a file could theoretically remain for eternity.  We had multiple points of school and merit review throughout the process when a thorough analysis could only be made near the end.

And there were obvious efficiency improvements we could make: moving certain reviews later made more sense given information availability.  We could replace second review with special population review.  We could combine merit review with school review and move it later in the process.  And so on.

When the dust settled, we had a new, more efficient file handling and review process.  What’s more, I had buy-in (even enthusiasm and cautious optimism) from the key stakeholders.

If we had jumped straight into vendor selection or building our own system, the gains we would have received through electronic review would have been reduced or even negated by our inefficient process.

The Takeaway

Despite what vendors or session presenters claim, technology is rarely the solution to your woes.

Before you build a predictive model, you should improve your communication process.  Before you blog, improve your customer service.

Technology is like cement: if you pour it onto a bad foundation, you’re only hurting yourself.