I don't get how counting cars in Wal-Mart and comparing rate of change over time can really help predict profit margin. The average size of basket (retail sort

I don't get how counting cars in Wal-Mart and comparing rate of change over time can really help predict profit margin. The average size of basket (retail sort-of-equivalent of ARPU), profit margin on goods, etc are all equally important but easily fluctuating factors which also make up profit projection. Even in the article says that the 4% increase in cars parked in June this year over last was due to extra cost-slashing WalMart was doing... well that can lead to lower profits per customer so even with more customers coming in the store it could lead to a net 0 profit gain (but more merchandise shifted) It's like saying I can guess the value of a in a = x * y * z by just monitoring the change in z, when all 3 are variables.