Hi @talhahussain484 ! Great question.
Each brand is a bit different, so the best way to land on the right delay is to A/B test. I'd test your control (20 seconds) against 10 or 15. If 15 wins, test that against 8. Time delay is one of my favorite things to test - you don't always get conclusive results, but popups tend to give you a clearer winner than most tests do.
The 14-day re-trigger is reasonable. Once your main tests wrap, you can test 14 vs 7, just don't expect a big signal - that traffic is a much smaller sample.
For PDPs: yes, show it there and on your main LPs, but exclude sensitive pages like /cart and /support. On PDPs specifically, you might get better results from a scroll-depth or exit trigger than a pure time delay, since a timed popup can interrupt someone mid-evaluation. Longer term, a PDP-specific popup with its own offer is worth exploring.
You’re thinking about this in a smart and strategic way! Happy to help talk through results once you've got some tests running.