We have the anonymized data from millions of Americans' online purchases. The data isn't perfect: the data extraction technology can have errors, and the users themselves are not perfectly representative of the overall online shopping population in the US (bias). We do a number of things to correct/improve the data, and need you to weigh in on whether or not the final analytics are credible or not. Lets say we're interested in Home Depot. Do some research into Home Depot's financials (and/or other sources you may identify) and deliver an estimated dollar value for the four quarters of 2014 that is apples-to-apples with our data.
Some hints: We observe total dollars, which might not be the same what Home Depot reports on their financial reports (have a look at their 10Qs). The average tax and shipping on Home Depot orders is 10% of the subtotal.