Yes, that's -- I think when you look at the Lab and go back to, sort of, when we started this initiative I think it last about, I think it reported $7 million-odd of losses in Q3 2010, probably about 4.5 and 4 of those were run rate. And since then, it's been a question of improving efficiency while trying to drive revenue through the cost base. And the lab team has done a very good job in both of those fronts. So as we look at it now you kind of get to the point where Lab costs can -- they sort of go in steps, so there's been a lot of work done in taking out cost, improving efficiency, in things like outsourcing, kit building and using certain services in some of the offshore labs. And so we've got a good handle on the cost base, I think it'll be fair to say, but there's been a fairly spectacular level of business wins in some of those quarters. I mean, the backlog was $150-odd million when we started this, so in a period of restructuring a lab fairly significantly and restructuring the cost base, the guys have also grown the backlog by 31% to over $200 million. So I think when you look at where it is, going forward, that's a fairly significant level of revenue growth. And historically, we know that at times when you're on board in that revenue, if it's around, assays that are perhaps not on your panel, there'll new work there with new customers. That level of business wins has been driven by being successful and expanding the customer base and dealing with some new people. That can tend to increase set up cost, so it's kind of a step level. So I think for the balance of this year, we're probably looking at continuing to drive the revenue through and concentrate on growing the backlog, which probably means that the margin -- that, that 5% where we're at year-to-date will become the target for the balance of the year. I think it's when we go beyond that and you get revenue that's up into the mid-20s that you'll start to see another kick on the leverage. But at the moment, I think it's fair to say, the turnaround is still in progress. So we don't have the same level of predictability that we might have in a steady state.