How to Plan a Dinner When Everyone Eats Differently

Everyone knows this dinner. Someone suggests a place. Someone else replies, gently, that they cannot eat there. A third person sends a menu screenshot. Somebody proposes a compromise nobody likes. Forty messages later the plan is either a restaurant that half the group is quietly unhappy with, or no plan at all.

The frustrating part is that this is not a hard problem in principle. It is a hard problem because the information needed to solve it is spread across five people who are all being polite about it.

Why the group chat fails at this

A group chat is a terrible constraint solver, for three specific reasons.

• Constraints surface late. Nobody opens with "I am coeliac." It comes up after someone has already got excited about a place, which makes raising it feel like an objection.
• Nobody wants to be the difficult one. People with restrictions routinely under-report them to avoid derailing the group, and then have a bad meal.
• Suggestions are serial. You evaluate one restaurant at a time against constraints you only partly know, which is the slowest possible search.

The result is predictable: the group optimises for social comfort instead of for the meal, and the person with the strictest needs pays for it.

Collect constraints before you suggest anything

The single biggest improvement you can make is to reverse the order. Do not suggest a restaurant and then find out who cannot go. Ask everyone what they need first, privately, and then search.

This matters more than it sounds. When people state a restriction into a form rather than into a conversation, they state it accurately. There is no social cost to being honest with a checkbox. You end up with the real constraint set instead of the polite one.

What to ask for

• Hard restrictions: allergies, coeliac, vegetarian, vegan, kosher, halal. These are non-negotiable filters, not preferences.
• Soft preferences: cuisines people are tired of, things they dislike, what they ate recently.
• Budget. People will not say this out loud in a group and it wrecks more plans than dietary needs do.
• Where they are coming from, so the location is fair rather than convenient for the organiser.

Filter hard, then rank

Treat the two categories differently. Hard restrictions remove restaurants from consideration entirely — there is no score high enough to make a steakhouse work for a vegan. Soft preferences reorder whatever survives.

Collapsing the two into one score is the classic mistake. It produces a "best match" that technically ranks highly and that one person in the group still cannot eat at.

Give the group a small, real shortlist

Once you have filtered, do not present twenty options. Present about five. A short list of places that all genuinely work turns the conversation from "is this okay?" into "which of these?", which is a much faster question and a much friendlier one.

Then vote. Voting is not just faster than discussion — it removes the social cost of preferring one option, because a vote is anonymous in a way that "actually I would rather not" is not.

Doing this without the spreadsheet

All of the above is what Pixx does by default. You create a meal and invite the group with a link. Everyone adds their own dietary restrictions and preferences to their own profile, so the vegetarian never has to announce it again for the next dinner. Pixx filters on the hard restrictions across the whole group, ranks what is left, and proposes a handful of restaurants that actually work for everyone.

The group votes, chats in a thread attached to the meal, and gets live updates until it is settled. Friends can RSVP and vote from the link in a browser without installing anything, which matters when the group includes the one person who will not download an app for a dinner.

And the next time this group eats together, the constraints are already there. That is the real win: the vegetarian, the kosher friend, and the person with the nut allergy each say it once, not every single time.