Running valet for a 300-guest wedding: staffing, timing and the arrival curve

The shape of the demand is the whole problem

A restaurant doing 120 cars on a Saturday spreads them across four hours. Some guests arrive at six, some at nine, and some are leaving while others are still parking. Your crew is busy but the work is continuous, and a crew of three can absorb it because the intake and the retrieval overlap and cancel each other out.

A wedding does the same 120 cars, or twice that, in two violent bursts separated by five hours of nothing. Everyone arrives before the ceremony because arriving after it is socially unacceptable. Everyone leaves when the music stops because the venue has a contractual end time. In between, your crew stands around.

That shape is the reason a wedding costs more per car than a restaurant night, the reason published staffing ratios are useless without a time window attached, and the reason the single most common way to fail a wedding is to staff it from a guest count instead of from a clock. We wrote separately about what to charge for valet, including the event bands by guest count. This piece is about the other question: once you have won the job, how do you actually run it.

Question one: how many cars is 300 guests?

This is the number your entire quote and your entire staffing plan rest on, and the published guidance spans a range of more than 60%.

Open Door's figure comes from its 2026 wedding and event valet guide, which states the ratio as "0.6 to 0.8 vehicles per guest (accounting for couples arriving together and rideshares)". So the honest planning range is 150 to 240 vehicles, and where you land inside it is not random. It is driven by four things you can ask about:

• How many guests are from out of town. Out-of-town guests staying in a hotel block frequently share a shuttle or a car. A wedding where 40% of guests flew in produces far fewer vehicles than a wedding where everyone drives from twenty minutes away.
• Whether there is a hotel shuttle. If the couple booked one, a large block of guests arrives on a bus, in one movement, at one moment. That is fewer cars but a sharper spike at the drop-off lane.
• How walkable and rideshare-friendly the venue is. A downtown hotel ballroom and a vineyard forty minutes outside town produce completely different ratios. The vineyard is closer to one car per two guests with almost no rideshare.
• Whether guests expect to drink. An open bar until eleven at a rural venue means more designated drivers and more carpooling, but it also means some cars will still be in your lot in the morning. Ask the venue what happens to a car left overnight, and put the answer in writing before you agree to be responsible for it.

If you cannot get a straight answer, staff to the top of the range, not the middle. Being over-staffed at a wedding costs you a few hundred dollars of wages. Being under-staffed at a wedding produces a photograph of a queue, a review, and the loss of the planner who referred you.

Question two: how many attendants? The published ratios disagree by more than double

Here is what the industry actually publishes, side by side. Read this table before you trust any single one of them.

Two of these say seven people and two say thirteen. That is not a rounding difference, it is a different business. And the reason for the gap is that every one of those ratios is missing its denominator. "One attendant per thirty vehicles" is not a rate. One attendant can handle thirty vehicles over an evening, or three vehicles in ten minutes, depending on how far away the cars are parked and how compressed the window is. A ratio without a time unit is not a staffing model, it is a rule of thumb that happens to be true for whatever operation the person writing it had in mind.

Open Door's own guide is a clean example of this. It says one valet per 30 to 40 vehicles at arrival peaks, and then gives the worked case: "For a 150-guest wedding (roughly 100 vehicles), plan for 3 to 4 valets." Three people, a hundred cars. That works if the cars trickle in over an hour and the lot is thirty seconds away. It does not work if 100 cars show up in twenty minutes and the lot is a four minute walk.

Do it as throughput instead

The only ratio that survives contact with a real curb is cars per attendant per hour, and it is almost entirely determined by one variable: the round trip to the parking area. An attendant's cycle is roughly the walk out, plus the drive back, plus the fixed overhead of the key handoff, the ticket and the log entry.

There is one published figure that anchors the top end, and it is worth knowing. Open Door's guide to high-volume amphitheatre valet states that "a well-coordinated team of eight can process 40 to 50 vehicles in the first ten minutes of post-show" using pre-pull staging and radio-coordinated retrieval. That is roughly 5 to 6 cars per attendant in ten minutes, or 30 to 37 per attendant per hour, but read the condition carefully: it is only achievable because the cars were already moved to exit-adjacent positions before the show ended. That number is a ceiling reached with preparation, not a rate you can assume.

So for a 300-guest wedding with 180 to 240 cars and a lot within a two minute walk, the throughput math lands at 9 to 12 people on the floor for the two peaks: roughly 8 to 10 running vehicles plus 2 at the stand handling greeting, keys and tickets. That is above what the wedding-specific ratios suggest and roughly in line with the corporate event ratio. We would trust the throughput number, because it is the only one that names its assumptions and can be checked against your own venue.

The arrival curve, hour by hour

Wedding etiquette is remarkably consistent about when guests show up, which is a gift to an operator: this is one of the few events where you can predict the curve before you see it. Guidance from the major wedding publications converges on guests arriving 15 to 30 minutes before the time printed on the invitation, with 30 minutes recommended for larger guest lists. Meanwhile couples are advised to start the ceremony 15 to 30 minutes after the printed time, and some deliberately print a start time earlier than the real one to absorb stragglers.

Put those two behaviours together and the practical arrival window for a large wedding is roughly 40 minutes wide, with the bulk of it in the final 20. Nobody wants to be the person walking in during the processional, so lateness has a hard social deadline that restaurant guests simply do not have.

Note what the arrival peak permits that the departure does not: a queue. A car waiting three minutes in the drop-off lane contains a seated guest with the air conditioning on, listening to the radio. It is not comfortable but it is survivable, and it means your drop-off lane functions as a buffer that absorbs the mismatch between arrival rate and your crew's absorption rate. Design the lane deliberately: know how many cars it holds before the tail spills onto the public road, because that is the moment the venue's neighbours start calling somebody.

The departure is the entire job

Everything above is the easy half. Here is the part that decides whether the planner calls you again.

Why the exit is compressed, and why you cannot negotiate it

Wedding receptions do not end when the energy fades. They end at a contractual time, and frequently that time is set by a municipal noise ordinance rather than by the venue. Coverage of the issue is blunt about it: many outdoor venues sit under county ordinances that end amplified music at 9:00 or 10:00pm, a constraint that is rarely mentioned on the venue tour but legally caps the timeline. A typical published end-of-evening sequence runs last call at 9:15, bar closed at 9:30, end-of-event announcement at 9:55, music off at 10:00, with the DJ or band required to have equipment out by 11:00.

So the departure is not a curve at all. It is a step function triggered by a legal deadline, and unlike arrivals it has no social pressure spreading it out. Nobody feels rude for leaving at the same moment as everybody else.

The send-off makes it worse, and it happens exactly where you are standing

The modern grand exit compounds the problem in a way that is specific to weddings. A sparkler send-off is normally the final item on the schedule, and it works by getting every remaining guest outside and lined up simultaneously, on the same driveway where your stand is. Published advice to couples recommends budgeting one or two sparklers per guest and lighting in sections rather than all at once, with the sparklers themselves burning for three and a half to four minutes.

Which means that at roughly 10:05pm you have 200 people already standing outside, already holding their phones, already at your podium, and the ceremony that was keeping them inside has just finished. There is no trickle. Plan the send-off position with the planner in advance so that the tunnel of guests is not standing across the lane your runners need.

The arithmetic of the exit

Take 200 cars leaving inside 40 minutes. That is 5 vehicles per minute, sustained, in the dark, with guests who have been drinking, some of whom cannot find the ticket. Against the throughput table above, a two minute lot gives you about 0.27 cars per attendant per minute, so five per minute needs roughly 18 attendant-equivalents of retrieval capacity, which you almost certainly do not have and cannot afford to bill.

This is why pre-staging is not an optimisation, it is the plan. The published amphitheatre figure exists for exactly this reason: moving high-demand vehicles to exit-adjacent positions during the last 20 to 30 minutes, which Open Door says cuts average retrieval time in half. Halving the round trip doubles your effective crew without adding a single wage hour.

• Stage during the last dance, not after it. The final 20 to 30 minutes of the reception is dead time for your crew and the single highest-leverage window of the night. Move cars forward then.
• Stage in departure order where you can infer it. You cannot know who leaves first, but you can know some things: parents of young children, guests who asked for their car at 9:15, anyone who told you at drop-off that they were leaving early. Park those shallow at intake.
• Run tickets ahead of guests. The published pattern is runners calling ticket numbers ahead by radio so cars are already moving before the guest reaches the podium. If the guest requests the car from their phone while still inside, you gain the entire walk from the dance floor to the door.
• Accept that a queue on the exit is a visible failure. Standing in the cold at 11pm in formalwear is what people describe afterwards. The whole design goal is to keep the guest moving toward a car that is already moving toward them.

Two venues means two arrival curves, and the second one is worse

Plenty of large weddings split the ceremony and the reception: a church, a synagogue or a garden for the ceremony, a hotel ballroom or an estate for the reception. If you are covering both, you are quoting two separate operations that happen to share a guest list, and the naive mistake is to price it as one long shift.

What actually happens is this. The first arrival curve is the sharp one described above, because the ceremony has a real start time. Then, within about 15 minutes of the ceremony ending, you run a full departure surge at site one and a full arrival surge at site two, back to back, with the same crew if the sites are close enough to share it and with two crews if they are not. And the site-two arrival is unusually compressed even by wedding standards, because everyone leaves the ceremony at the same instant and drives the same route.

Practical consequences worth pricing:

• Shuttles change the arithmetic in your favour. Wedding transportation guidance recommends shuttle service when venues are more than about 30 minutes apart. Every guest on a shuttle is a car that stays parked at site one through the whole reception, which removes them from your second curve entirely. Find out whether a shuttle is booked before you quote, because it can move your car count by a third.
• Cars left at the ceremony site are still your problem. If guests shuttle from site one to site two, somebody has to be there at midnight when the shuttle brings them back. That is a second, later, thinly-staffed shift and it belongs on the invoice as its own line.
• The crew cannot be in two places. If the sites are ten minutes apart you can move people, but the last car out of site one and the first car into site two overlap. Size that overlap honestly or split the crew.

We could not find published operator guidance or pricing conventions for the two-site wedding specifically, which is a real gap in the available material. Treat it as two events on one invoice with a stated transfer window, and say so plainly in the proposal rather than absorbing it.

Five wedding-specific things that break an otherwise good plan

1. The getaway car is a vehicle you never park

If the couple booked a classic car for the send-off, it is not part of your inventory and it must not be treated as such. Classic car hire guidance describes the vehicle arriving early, being warmed up and staged, and staying available without pressure from another booking. Operationally that means a reserved position near the exit, occupied all night, that your crew must not fill, and a departure that happens at the worst possible moment: inside the sparkler tunnel, with 200 guests on the driveway, immediately before the exit surge.

Get three things in writing from the planner: where the car stages, who has its keys (usually the hire company's own driver, not you, and that changes your liability), and what the clear path out is. Do not let the send-off route and your retrieval lane be the same strip of asphalt at the same minute.

2. Photography owns the driveway at unpredictable times

The cocktail hour and the golden hour before sunset are when the couple is being photographed, and the photogenic parts of a venue are frequently the front steps, the driveway, the tree-lined approach and the porte-cochère. All of which are your operating area. This is not in any published operator guide we found, so treat it as judgment rather than fact, but it follows directly from the shape of a wedding timeline: ask the planner for the photography schedule, not just the event schedule. Your dead hours and their portrait window overlap almost perfectly, which is convenient if you know about it and a confrontation if you do not.

3. Weather at an outdoor venue is an operational variable, not a mood

Published wedding valet guidance is consistent that outdoor events need a written rain contingency: covered podiums, umbrellas for escorting guests, staging repositioned to covered areas, and weather gear for staff. Open Door's wedding guide adds the seasonal end of it, noting summer requires shade and hydration for staff and winter requires discussing snow removal and de-icing.

The one that costs real money at a barn, farm or vineyard wedding is mud. Grass overflow parking after rain produces stuck vehicles, ruined turf and a venue asking who pays for it. Walk the overflow area on the day, not on the site visit, and if it is soft, say so before the first car goes on it. That conversation is uncomfortable at 3pm and catastrophic at 10pm.

4. Vendor vehicles are already in your best spaces

The band, the DJ, the photographer, the florist and the caterer all arrive before your guests, park close because they are loading in, and leave at completely different times from everyone else. The published end-of-night sequence has equipment out by 11:00, which is during or just after your exit surge. Ask the planner for the vendor list and load-out times, and mark those positions on your lot map before the first guest arrives, so you are not discovering at 10:15 that the band's van is boxed in by twelve guest cars.

5. Somebody will not be able to drive

An open bar and a car in your custody is a combination with a defined and unpleasant ending. Decide the policy before the night rather than at the podium, agree it with the planner and the venue, and know who makes the call. This is a genuine legal exposure that varies by state, and it is worth a conversation with your own attorney and carrier rather than a paragraph on a blog.

Season: you are bidding on about twenty Saturdays

Wedding demand is concentrated to a degree that shapes your whole staffing model. Carats and Cake's analysis of 2,181 weddings from 2025 gives this national distribution:

Five months carry more than two thirds of the year. The Knot's own national data puts October first at 17% of US weddings in 2024, which disagrees with the ranking above but agrees entirely on the shape: fall and late spring dominate and midwinter is empty. And within those months, Saturday is overwhelmingly the day. The Knot's most popular 2026 dates are 6 June, 20 June, 19 September, 26 September and 10 October, every one of them a Saturday.

Three consequences for an operator:

• Your capacity constraint is Saturdays, not months. Roughly twenty Saturdays carry most of your annual wedding revenue, and two weddings on the same date means you turn one down or you find a second crew. Every declined date is a planner you do not get to work with again this year.
• You cannot staff this with full-time employees. A 300-guest wedding needs ten to twelve people for about six hours, twenty-odd times a year, all on the same day of the week. That is a bench of trained casual staff, which means recruiting, training and retention are your actual bottleneck, not equipment or software.
• Book far ahead and say so. Published guidance tells couples to book valet 4 to 6 months out for peak-season weekends, and 6 to 12 months for peak Saturdays. If planners hear from you in March about October, you are early, not pushy.

What the planner needs from you, and what you need from them

Wedding planners are not the same buyer as a restaurant owner. They are booking on behalf of a client who will remember this day forever, they carry the blame for anything that goes wrong regardless of whose fault it was, and they refer repeatedly if you make them look good. The mechanics of winning and structuring the account are covered in our guide to bidding a valet contract with the actual numbers, which applies to planners as much as to restaurants. What is wedding-specific is the information exchange.

Ask for, in writing, before you commit to a crew size:

1. Final guest count and the date it becomes final.
2. Printed ceremony time and the real intended start time. These differ, sometimes deliberately, and only the planner knows both.
3. Whether a hotel shuttle is running, from where, and for how many guests.
4. The hard end time and what sets it: contract, noise ordinance, or the couple.
5. Whether there is a grand exit, what kind, and where the tunnel forms.
6. Whether a getaway car is booked, who drives it, where it stages.
7. The photography schedule, specifically anything happening at the entrance or driveway.
8. The vendor list with load-in and load-out times and vehicle count.
9. The rain plan, and what happens to overflow grass parking if it rains the day before.
10. Who your single point of contact is on the day and how you reach them when the venue wifi is saturated by 300 phones.

Give them, in writing, before the day: your crew size and call time, your lot map with capacity and overflow, your rain contingency, your insurance certificate with the venue named as additional insured if they asked, and the name and mobile number of your supervisor on site. A planner who has that document stops worrying about parking, and a planner who stops worrying about parking calls you for the next one.

Where the software fits, and where it does not

If a wedding is a one-off for you rather than a weekly account, the honest problem with most valet software is that it is priced and sold for a permanent operation. The market floor is around US$99 a month with a mandatory demo call, which is a strange thing to buy for one Saturday in September.

ParkingPro is self-service, which is the part that matters here: you can create an account the week of the event, set up the operation and run the shift that night without an installation, a sales call or a contract. Plans start at US$19 a month with 14 days free, so a single event can be run inside the trial. Guests receive their ticket as a link over WhatsApp or SMS rather than a paper stub, which matters at a wedding because no guest is going to install an app on the way to a ceremony. Every active vehicle sits in a live list you can search by plate or description, which is what you actually need at 10:15pm when somebody cannot find their ticket. It keeps recording entries, exits and charges when the connection drops, which is a real consideration at a barn or a vineyard.

At the end of the night the shift report shows what was recorded: which attendant was on, what came in, what went out and what was collected. For a wedding that is genuinely useful as a record afterwards, because when a planner or the couple asks what happened with a particular car, you have a timestamped answer instead of a recollection.

If you are still setting the price rather than planning the operation, the companion piece on what to charge for valet parking has the published event bands, the staffing ratios and the cost floor you should never bid below.

This article is general operational guidance, not legal, insurance or employment advice. Liability for vehicles in your custody, obligations around intoxicated guests, municipal valet permits and noise ordinances all vary by state and by city. Confirm your own position with your attorney and your insurance carrier. Every external figure cited is from a public source consulted on 17 August 2026 and is presented with its origin. Published valet staffing ratios come from operator marketing pages rather than from an industry study, and they disagree with each other by more than a factor of two, which is why this article shows the disagreement rather than picking a winner. The throughput estimates are arithmetic from stated assumptions, not measured data. ParkingPro Cloud is a product of Abalon LLC.