Wedding budget · the arithmetic

What each wedding guest actually costs you

The Knot's 2026 study puts the average American wedding at $34,200 for 117 guests. Divide one by the other and you get $292 a head. That number is real, and it is almost useless for deciding anything.

Here is the problem. $292 is an average, not a price. It is what you get when you take everything you spent and split it by everyone who showed up. It is not what the next guest costs you, and it is not what you save by uninviting one.

Those are three different numbers, and the gap between them is where most wedding budgets go wrong.

Half your wedding does not care how many people come

Walk down The Knot's own list of average vendor costs and sort it by one question: does this change if twenty more people RSVP?

Mostly, no. A photographer is $3,000 whether he shoots 80 guests or 200. The videographer is $2,300. The dress is $2,100. The DJ is $1,800. The band is $4,500. The officiant, the license, the hair and makeup, the planner's fee, the bulk of the venue's site fee — all flat. None of it moves.

Then there is the other pile, the one that moves with every RSVP: catering at roughly $80 a plate, the bar, chairs and linens and place settings, invitations and postage, favours, cake slices, and the service charge sitting on top of all of it.

Where the per-guest number comes from

The $172.92 is not just the plate. It is the plate plus the drinks that guest drinks, the chair they sit on, the linen under their plate, the invitation you mailed them, the favour they leave behind — and then the service charge levied on the catering and rental portion of it.

That last part is the piece people miss. In this model, roughly $133 of each guest's cost sits on lines that carry a 24% service charge and tax. So a $133 line item is really costing you $164.92. The charge is invisible on the quote and very visible on the final invoice.

So what does cutting the guest list actually save?

Three ways to estimate it, and only one is right:

The left column is what you get if you use the average — it overstates the saving badly, because it pretends a slice of the photographer walks out the door with each guest. The middle column is what you get if you only count the plate, which is the assumption behind most "cut ten guests, save a thousand dollars" advice. It understates it.

The real number sits between them, and for most couples it is the single largest lever in the entire budget. Cutting twenty names saves more than dropping the band and the videographer together.

The whole curve

Run the same model across guest counts and you get the shape of the thing. Note what the middle column is and is not: several cuts on the list are alternatives to each other — you can cap the bar or go cash-bar-after-dinner or bring your own alcohol, but not all three — so this is the best combination that can actually coexist, and each one is capped at what that category really costs. "Floor" is what is left after taking all of them. It is a ceiling on what cutting can do, not a plan anyone would follow.

Two things fall out of this table. First, the floor rises with guest count and never gets near zero — there is no number of cuts that makes a 250-person wedding cheap. Second, the cuttable amount grows much more slowly than the total. At 100 guests you can take about $15,540 off. At 300, with a bill more than twice the size, you can only take $25,220. Scale works against you.

What this model gets wrong

Being straight with you: this baseline runs above The Knot's national average — roughly $39,500 at 117 guests against their $34,200. That is deliberate. It assumes a full-service wedding with a catered plated dinner, a bar, a photographer and rentals, in a mid-to-high cost-of-living area. A backyard wedding with a taco truck breaks every assumption in it.

Regional spread is also enormous. The same wedding is a different number in Manhattan and in Boise, and no national model fixes that. Real vendor quotes beat any model, always. What a model is good for is telling you which questions to ask before the quotes arrive, and how much a decision is worth before you make it.

Vendor averages and the $34,200 / 117-guest / $292-per-guest figures: The Knot 2026 Real Weddings Study, surveying 10,474 US couples married in 2025. Fixed/per-guest split and all derived figures are from The Wedding Math model described above.