I've been working on something similar, tracking how things used to be. Back in maybe 2014-2017 I ran a spreadsheet tracking payout dates across three networks. The goal was simple, find who actually paid on time every time. I logged the promised date, the actual payment date and the delay in days. For one network over 24 payments, their average delay was 4.1 days with a standard deviation of only 1.2 days. That meant they were consistent even if slightly late. The trick was negotiating payment terms based on that historical data. If a network's average delay was under three days, I'd push for weekly net-7 instead of bi-weekly net-15. It worked cuz I had the numbers to back it up, not just complaints. These days everyone wants instant payouts but sometimes you have to look at consistency over speed. It all comes down to the human connection with your AM but also having cold hard data so they can't argue.