The honest answer is somewhere between 66 and 220 drinks, and that spread is the actual lesson. Two events with 200 guests on the invite list can differ by more than 3x in what they actually pour. Guest count alone does not predict drink count, which is why so many carts either run out at hour two or drive home with forty wasted cups of milk.
Below are the real participation rates our estimator runs on, what they produce for a 200-guest event, and the two adjustments that move the number more than anything else.
The number that predicts demand is event type, not headcount
A captive wedding reception and an open-air festival are not the same event with different decorations. One has a room of people with nothing else to do and a three-hour window. The other has foot traffic that walks past you. These are the baseline rates, expressed as drinks per guest over a standard two-hour service:
- Wedding: 0.90. Highest of any type. A captive room over a long reception comes back for seconds.
- Corporate: 0.85. Nearly as high, and far more predictable. People take the free coffee.
- Private party: 0.75. Relaxed setting, guests drift, not everyone orders.
- Nonprofit: 0.65. Longer events, more program time, less standing around.
- Festival: 0.40. This is a share of nearby foot traffic, not of ticket holders. Do not quote it against gate numbers.
If no event type is set at all, plan on 0.75. That is the private-event rate, which sits in the middle of the five and is the safest guess when a client has told you a headcount and nothing else.
Public events swing on the clock, not the type
Public events get scored by start time instead, and the spread is wider than the gap across all five typed events above. A 7am commuter crowd converts. A 7pm one does not:
- Morning (before 11am): 0.60
- Midday (11am to 2pm): 0.40
- Afternoon (2pm to 5pm): 0.35
- Evening (after 5pm): 0.30
Morning to evening is almost exactly a 2x swing on identical foot traffic. If you quote a public event flat regardless of time slot, you are wrong in one direction half the time.
Two adjustments that move the number
The early start bump. Any non-public event starting before 10am adds 0.10 to the base rate. Coffee demand is highest before people have had any. An 8am corporate breakfast runs at 0.95 per guest, not 0.85. Public events are excluded from this on purpose, because their morning band already prices the hour in and bumping them would count the same effect twice.
The hours multiplier. Past the two-hour baseline, every additional hour adds a repeat-order rate: 0.10 per hour for weddings, 0.08 corporate, 0.06 private, 0.04 nonprofit, 0.03 festival and 0.03 public. Weddings carry the highest hourly for the same reason they carry the highest base. Nobody is leaving.
The full formula is simple enough to do on a phone: drinks = guests × (base + extra hours × hourly).
What 200 guests actually means
Running that math across real scenarios, all at 200 guests:
- Wedding, 4 hours, 5pm start: 220 drinks. 0.90 + (2 × 0.10) = 1.10 per guest. More drinks than guests.
- Corporate breakfast, 3 hours, 8am start: 206 drinks. The morning bump does most of that work.
- Corporate, 2 hours, 1pm start: 170 drinks. Same headcount, same client, 36 fewer drinks than the breakfast.
- Private party, 3 hours, evening: 162 drinks.
- Nonprofit gala, 3 hours: 138 drinks.
- Public event, 3 hours, 8am: 126 drinks.
- Festival, 5 hours: 98 drinks against 200 people of nearby foot traffic.
- Public event, 3 hours, 6pm: 66 drinks. The floor.
Same 200 guests. 220 drinks at one end, 66 at the other. Any vendor who quotes you a flat "plan on one drink per guest" is guessing, and on the evening public event they are guessing three times too high.
Then add 10%, and not because the math is uncertain
Stock and staff to 110% of the estimate. The reason is not that the estimate is shaky, it is that the two failure modes cost wildly different amounts. Running out of milk at hour two of a wedding costs you the reference, the review, and every booking that would have come from that room. Carrying 10% spare costs you a few dollars of stock you will use next weekend.
The failure modes are not symmetrical, so the estimate should not be centered. For the 220-drink wedding, prep for 242.
Staffing: the machine is the ceiling, not the person
One barista on a single-group espresso setup sustains about 60 drinks per hour. That number is the equipment, not the human. Putting a second barista on the same single group buys you almost nothing, which is the most expensive staffing mistake in this business: paying two people to be bottlenecked by one machine.
For the 242-drink wedding over 4 hours, that is 242 ÷ 240 = just over one barista-hour of capacity per hour, so you need 2 baristas and a second group head, not 2 baristas on one machine. If you cannot add the second group, the honest move is to tell the client the service window needs to be longer.
The takeaway
Stop quoting off headcount. Ask three questions before you price anything: what type of event, what time does it start, and how many hours of service. Those three move the drink count by more than 3x, and the guest count is only the multiplier at the end.
VenVen's pricing calculator runs exactly this model against your own menu and costs, and the all-in cost per drink is the other half of knowing whether a 200-guest event is worth taking.