Over the last century, our own radio technologies have undergone a continual progression in efficiency and data rate. Imagine someone with the latest radio technology from 1920 trying to monitor our civilization's progress. By 1950, a lot of things seem to become undecodable (FM), and shortly after that a lot of carriers seem to simply disappear. Assuming they manage to lock on to the frequency-modulated edge of a suppressed carrier in the 1970s, things start to switch to a multitonal digital encoding. Then round after round of increasingly complex digital encodings with error correction and encryption, with larger and larger symbol rates and symbol set sizes. Frequencies climb into the thousands of megahertz. Hundreds of thousands of carriers are used in parallel.
By the early 21st century it's unlikely our 1920s engineer can even detect radio signals from devices like a modern cellphone even if they had it in the lab with them -- and they blatantly transmit high power beacons in the open to find other devices. Too low power, too high frequency, data encodings based on mathematics not even invented yet. Even if detected, there would be no obvious modulation, and it would look a lot like a white noise emitter.
Maybe it's not worth even attempting until the receiving civilization would be able to decode more than just a few bytes from a signal that could be affordably transmitted from a single star.