How did they measure the number of legitimate customers that they lost because they blocked those IPs? It's not enough to estimate the change in unwanted traffi

How did they measure the number of legitimate customers that they lost because they blocked those IPs? It's not enough to estimate the change in unwanted traffic, you also need to estimate the change in desirable traffic.