Holdout choice tasks are usually used to test the external validity of your model. Having said that, you need to carefully consider how to build your alternatives in the tasks.
Maybe this paper is also intersting for you: https://sawtoothsoftware.com/resources/technical-papers/including-holdout-choice-tasks-in-conjoint-studies
And I can also recommend this book:
Mariel, Petr; Hoyos, David; Meyerhoff, Jürgen; Czajkowski, Mikolaj; Dekker, Thijs; Glenk, Klaus et al. (2021): Environmental Valuation with Discrete Choice Experiments. Cham: Springer International Publishing (SpringerBriefs in Economics).
On p. 120, you can find the following regarding holdout tasks: "An alternative strategy is to randomly drop a certain percentage of the observations, and estimate the model without the dropped observations (hold-out sample). The estimated parameters are then used to predict the choices of the excluded observations. The number of correct predictions can be used as an indication of how well the model performs outside of the sample. This procedure is less computationally intensive than the leave-one-out test, because each model is estimated only once. The hold-out sample approach is therefore more adequate for larger samples and computationally intensive models."