All of these people could work behind

All of these people could work behind the scenes as advisors if your internal people can take the advice and run with it, saving a lot of consulting fees. Maybe you even have one or more of these internal resources available and need no line item at all.

Either way, the return will be huge relative to the investment, because you're focusing very specific and limited resources to address an essential business need. In this particular situation, we're not even really talking about solving a problem, we're talking about raising the bar and enabling significant change, which is an exciting place to be. Who knew you could actually have exciting organizational development during a recession?

2. Identify the root causes of any significant problems before agreeing to a budget number. Do not settle for, "Cut the technical training budget by 20%," which is a silly request because it has no bearing on the business. Figure out if your internal training is solving -- or at least helping to solve -- the most critical problems of the business. If it's not, chuck all of it, not just 20%. Believe me, you'll be the CFO's new best friend and employees who have been stuck in irrelevant training classes will cheer. Then figure out which training, if any, will help address critical business needs, and rebuild your budget from the bottom up. It's never too late to get this right. If you're too late in the game for this year, start the process with an eye on setting next year's budget based on real need, not just the previous year's number.

3. Get creative about continuing your business-critical organizational development investments, with the goal of sustaining and moving forward until better times allow you to take bigger leaps. For example, it's essential to develop your high-potential leaders on an ongoing basis, during good times and bad, but do you really need to fly them all to Tuscany, or would an easy-in/easy-out business center like Frankfurt do? How about focusing on local rotational assignments instead of bigger development "events" that increase opex? How about utilizing other technologies to develop them? Even something as dirt-cheap as a quarterly book club is a way to keep the process moving forward until better times. You could host it via Webex, or a conference call, or a blog, or a wiki (o.k., maybe a wiki is a stretch...).