Well, I think that's one of the core questions that we and all of you have been struggling with for the last couple of years. I feel like our -- the outcome of net revenue margin compression over the last couple of years, one of the reasons why I highlighted the 10-year trend is that, if you go back 4 years to late 2008, early 2009 and you see our all-time highs and the 10-year highs on that net revenue margin, we were saying then, and it was not at all surprising to us, that those margins would cycle back to more longer-term averages. Now the fact that they've cycled down for all of 3 years and have cycled to lows and all the questions around secular impacts versus cyclical, and kind of sorting out how much of it's the balanced market, how much of it is competition, how much of it is mix shift, how much of it is change in shipper's attitude, those are all real factors that we've acknowledged before, that I think they all play into the mix in one variety or another. And I guess, part of the prepared comments was really saying, because you can't sort them out, we have to make the choice of how we want to respond to the environment in aggregate, and our choice is clearly that in this balanced market, we want to stay near the top of that route guide, we want to have committed relationships to our customers and go after that market share. So again, it's just impossible to isolate the impact from any one of those variables. It's a combination of all those market factors.
Christopher J. Ceraso - Crédit Suisse AG, Research Division: Okay. And then as a follow-up, you mentioned in -- at the end of your closing comments, John, about productivity being a critical factor. How do you measure that? And what are the benchmarks, kind of targets that you're looking at? And we can't really use the historical metrics for Robinson now, with the acquisition that those numbers get kind of thrown off. So what should we look at to gauge your performance in terms of productivity?