The old system dealt with the data in a very different way, and it is inevitable that some small differences in counts may have emerged. But during our transition period we carried out extensive testing and found that the counts were often exactly the same and if not were very similar, certainly well within the COUNTER margin of error.