is it really that difficult to conclude support quality is tied to overall advertiser spend?