A number of metrics would be used. First is looking at what we call “escape infection” and the rates of escape infection: people who have been vaccinated and the rate at which you get new infections. If that rate goes above what is the expected efficacy of the vaccine, then it may be a strong signal that immune levels are waning.
The second, as I mentioned earlier, is to identify immune measures, which, if they go below certain thresholds, would suggest that you don't have adequate immune protection. Those are called “correlates of protection”.
On both fronts, it's too early to tell. We'll need a significant amount more time, I think, before we have evidence that suggests definitively what the appropriate time for a booster is.