Restaurant Week is usually described as a reservations problem. It is mostly a menu comprehension problem, and the phone is where that lands.
The guest saw a listing on a promoter's site, glanced at a three-course description, and now has questions. Is the salmon one of the choices or an example. Can they swap the dessert. Does it run at lunch. Is that price per person before tax. Do they need a reservation or can four of them show up at seven. None of that is on the promoter's page in the detail they want, so they call you.
The same six questions, several hundred times
Pull a night's worth of calls during the week and the repetition is almost comic. Roughly six questions account for most of the volume, and they do not vary between guests.
What is included. What are the actual choices in each course. Is it available every day and at every service. Are substitutions allowed and do they cost extra. Is beverage, tax, or gratuity included. Do you need a reservation.
Every one of those has a fixed answer you could write down in ten minutes. Every one of them, unanswered, is a party that books somewhere else. And every one of them, answered live by your host during a seating rush, is thirty to sixty seconds taken away from guests already standing in your dining room. That trade happens invisibly a hundred times a week, which is the dynamic described in reducing host stand phone interruptions.
Write the answers down first. Not for the phone, for yourself, because the act of writing them reveals the two or three you have not actually decided yet.
The rules you have not decided are the ones that cause fights
Substitutions are the classic. The menu says three courses with two choices each. A guest asks whether they can have the pasta from the regular menu as their main and pay a supplement. If you have not decided, the answer depends on which server is on, and now two tables in the same room got different rulings on the same night.
Decide in advance and write it as a rule, not a vibe. Substitutions allowed or not. Supplement price if yes. Whether the prix fixe applies to a table where only some guests order it. Whether it runs on the busiest night of the week or is blacked out on Saturday. Whether split checks are handled differently.
These are the things that determine whether the week is profitable or a slow-motion argument. They are also exactly the answers your phone needs, because the guest asks about them before booking.
Two menus running at once is where accuracy dies
Your kitchen is running the normal menu and the prix fixe simultaneously. So is your phone, and so is your POS.
That overlap causes a specific error. A dish appears on both menus at different effective prices. Someone quotes the regular price to a prix fixe guest, or rings a prix fixe course as an à la carte item, and the ticket totals are wrong in a way nobody catches until the close-out does not reconcile. Multiply by a busy week and it is real money.
If software answers your phone, this is the case to test by hand before the week opens. Ask about a dish that appears on both menus and listen for which price comes back. An agent that has been given the prix fixe as a separate menu with its own rules will answer correctly. One that had the new dishes bolted onto the existing menu will blend them, confidently. Loading a temporary menu properly is its own small discipline, covered in seasonal menu updates for a voice agent, and it is the difference between a helpful week and a week of corrections.
Reservation pressure is real, and it is not the same as call volume
The booking surge is front-loaded. Most of it lands in the days right after the promoter's list goes live, and it concentrates on the two or three best time slots.
Two things follow from that. First, the phone rush and the service rush happen at different times, so the staffing answer for one is not the answer for the other. Second, once the prime slots are gone, a large share of calls become negotiations: what else do you have, what about the following week, can you do six at eight-thirty.
Those negotiation calls need a person, or at least need a system that can offer real alternatives rather than saying no. Larger parties especially, since a six-top during a prix fixe week has capacity implications your book needs to weigh. Large party reservations by phone goes through the handling, and the general split between what belongs on an automated line and what belongs with a human is in voice AI for reservations versus ordering.
The useful division is simple. Questions are automatable. Booking negotiations are not, at least not the interesting ones.
No-shows hurt more this week than any other
A no-show on a prix fixe night costs you a seat you had discounted and a party you turned away for it. The economics of the week are already thin, and every empty four-top makes them worse.
Confirmation calls or texts the day before are ordinary practice and they work. If your phone is automated, outbound confirmation is worth asking your vendor about specifically, because it is the cheapest intervention available on a week where the downside is concentrated.
Also decide your hold policy before the week rather than during it. Fifteen minutes, twenty, card on file for parties above a certain size. A policy applied inconsistently across a busy week creates more guest anger than a strict policy applied evenly.
Say plainly what the week is not
Some things are worth refusing out loud. Running the prix fixe on delivery is usually a bad idea: the menu is built around a seated pace and a plate, and it arrives as a disappointment in a container. Discounting it further to compete with the restaurant down the street is worse. Adding a third seating to squeeze in more covers at a price designed to be thin rarely produces the return visits the whole exercise is supposed to buy.
If a caller asks for any of those, the honest answer is no, with an alternative. Guests handle a clear no better than a reluctant yes that becomes a bad meal.
The measurement almost nobody does
Restaurant Week is a customer acquisition spend. The covers themselves are close to break-even by design. The only question that determines whether it worked is how many of those guests came back at full price in the following three months.
So set that up now, before the week, because you cannot reconstruct it afterward. Capture something identifying at booking, flag those checks in your POS, and look again in ninety days. If the return rate is decent, do it again next year and staff the phone properly. If it is near zero, you have learned something more valuable than a full dining room, and you can stop wondering.