Clarification: the Poisson coefficient beta is the difference between the log expected counts beta=log(mu_(x+1))-log(mu_x). Reworking this expression the percent change in counts equals (mu_(x+1)-mu_x)/mu_x=exp(beta)-1. That's why at 2:38 and 5:44 in the video, I interpreted the coefficient approximately as a percent change in counts.
Clarification: the Poisson coefficient beta is the difference between the log expected counts beta=log(mu_(x+1))-log(mu_x). Reworking this expression the percent change in counts equals (mu_(x+1)-mu_x)/mu_x=exp(beta)-1. That's why at 2:38 and 5:44 in the video, I interpreted the coefficient approximately as a percent change in counts.