We have that we are given 4 formulas for the products; there are split in 2, one for each company and they are also split in 2 on whether they are predictions or date. We need to substitute in x=4 days for all of them to calculate the value after 4 days after the launch (initial day counts for x=0). For company A, the actual average downloads per day were: for x=0 (day 1), x=1, x=2, x=3, x-4: f(0)=450, f(1)=562.5, f(2)=703, f(3)=879, f(4)=1099 To find the average download per days, we have that we divide the final downloads f(4) by the number of days, 4. This yields 275 downloads per day. For the prediction formula f(x)=400*1.5^x, if we substitute x=4 we get 2025 downloads. This yields on average 506 downloads per day. Thus, the difference is 231 downloads per day and we have a decrease in the data (from predicted to actual downloads).
For company B, the same procedure for the actual data yields: g(4)=2621 downloads after 4 days, hence 655 downloads on average; the prediction data yields 500*1.4^4=1921 and on average 480 downloads per day; the difference is around 175 (174 due to rounding errors) downloads per day and it is still a decrease (predictions are lower).
