Here are side by side comparisons of 2007 and 2008 – I think now that we are looking at a second years worth of data this is more interesting than the time series. I think the percet change figure is still better plotted as an overall time series though.
NOTE: I plotted this as 52 weeks – for last year we had 3 weeks that sdr dared to be out on vacation – for those weeks I averaged the number from the preceding and following week to get a comparative value.