If a signal uses the daily close, the earliest realistic execution is the next open, not the same close. Assuming you traded the close you used to compute the signal is a free look into the future worth, on many daily strategies, more than the entire claimed edge.
Intraday the lag is a real engineering number: data arrival, computation, network hop, broker acknowledgement, exchange matching. A strategy on one-minute bars might genuinely be able to act 200 milliseconds after the bar closes, but not at the bar close itself.
Make the lag an explicit parameter and test sensitivity to it. If a strategy earns 18% a year with zero lag, 6% with one-bar lag, and loses money with two, it is a latency business you are not equipped to run rather than a signal you discovered.
Related: look-ahead-bias, next-bar-execution, latency, fill-assumption