There are 2 approaches I take. Either I try to explain to them the outcomes that will potentially happen and see of they will budge ( which usually doesn't happen) or I just go ahead and try to achieve their unrealistic expectation, knowing full well that it will end up in a disaster... but then have a contingency plan to save the day. Because some stakeholders will not see reason until it collapses. So you will need to let it collapse or at least right the the verge of collapsing. Then let your stakeholders know that it's going to end up badly, but you have an out. Usually they will see it doesn't work and then in the future you can always reference the "almost" failure and push back. They will usually follow through with your decision.

When dealing with stakeholders who have unrealistic expectations, manage their outlook by setting clear, achievable goals from the start. Use data and past event performance to illustrate what is feasible within the given constraints. Open and honest communication helps align their expectations with the realities of the event planning process.

Set realistic boundaries by discussing the budget, timeline, and logistical constraints early on, so they understand what’s feasible. If their expectations are unrealistic, offer alternative solutions that still align with their vision but are more practical. If pushback happens, be tactful but firm about what can and can't be done. Manage expectations proactively by reminding them that quality, safety, and attendee experience are the priority, and sometimes compromises are necessary to achieve those goals.

I love analogies - I often use the "juggling Glass and Plastic balls" - in this instance it would be knowing what expectations are Glass Balls (ones I can't afford to drop) and what are Plastic (can be dropped). Ultimately, what's important is the end result, what's trying to be achieved - can I deliver on this with a few absences along the way. It's like putting on a great show, the actors might make mistakes on stage but the audience is none the wiser unless it's made obvious.