Chapter 5: The Guest List — The Number That Drives Every Other Number

"Everything in the wedding is downstream of the number of people." — a working rule in the trade; the phrasing here is my own

Chapter Overview

The worst hour I have ever spent in a client's kitchen was about eleven names.

They were $8,000 over. The list was at 186. Every option we had discussed — cheaper flowers, a DJ instead of a band, dropping the videographer — produced savings in the hundreds, and we needed thousands. There was exactly one lever that would work and both of them knew it, and neither of them would say it, so I did.

"We need to be at about a hundred and sixty."

Twenty-six names. The bride's mother, who was not paying and had contributed forty-one of the invitations, was going to have to lose some of them. The groom's college roommate — someone he had seen twice in nine years but who had been in the room the night they met — was on the bubble. There were four coworkers whose presence was entirely a matter of office politics.

It took four hours over two evenings. There was crying, and there was a phone call to a mother that I could hear from the next room, and at one point the groom said, quietly, "this is the worst part of the whole thing" — which is true, and which almost every couple discovers.

We got to 158. The budget closed with $600 to spare.

Here is what I want you to notice: nothing else we could have done would have worked. Not one of the other levers moved enough. The list was not one option among several. It was the only option, and it was the only option because of an arithmetic fact that most couples never have explained to them.

In this chapter, you will learn to:

• Classify every line in a budget by how it responds to guest count — and discover that this is a three-way split, not a two-way one
• Calculate what one more guest actually costs, which is not what the client thinks and not what the caterer's per-head price says either
• Find the step thresholds in a specific event, where removing one guest saves nothing and removing four saves $1,400
• Build a list with explicit ownership of every single name, which is what makes cutting possible
• Forecast who will actually come, defensibly
• Have the conversation

🏃 Fast Track: If you have run events with headcount-driven budgets — catering, venue, corporate meetings — skim §5.2 and §5.3 and read §5.4 (the politics), §5.5 (the three policies), and §5.7 (the A/B list) in full. Those are the wedding-specific parts and they are where operationally competent people get caught. Then do exercises B.2, B.4, and D.1.

🔬 Deep Dive: Work case-study-02.md, which computes the full marginal-cost curve for a single wedding and identifies every step threshold between 100 and 200 guests. It is the most quantitative material in Part I.

5.1 The Master Variable

🚪 Threshold Concept: guest count is the master variable

Ask a couple what determines the cost of their wedding and they will name things: the venue, the food, the flowers, whether you have a band. All of those are real. None of them is the answer.

The answer is how many people are in the room, and it is the answer by a very large margin.

Here is the demonstration, and it is worth doing on paper with a client. Take a $42,000 wedding for 140 people and change exactly one thing at a time:

Change	New total	Saving
Band → DJ	$40,900 | $1,100
Videographer removed	$40,400 | $1,600
Floral reduced 40%	$40,824 | $1,176
Plated → buffet	$40,600 | $1,400
Bar: full → beer and wine	$40,000 | $2,000
All five of the above together	$34,700** | **$7,300
Guest count 140 → 110	$34,850** | **$7,150

Thirty names produced the same saving as five substantial compromises to the wedding itself. And the wedding at 110 guests with a band, a videographer, full floral, plated service, and a full bar is — by almost any measure a guest would apply — a better wedding than the one at 140 with none of them.

Before this clicks: you treat the guest list as a given and try to build a budget around it. You spend months negotiating vendors down by three and five percent while the single largest variable sits untouched because nobody wants to discuss it. You find $2,000 of savings across eleven painful conversations and remain $6,000 over.

After this clicks: the guest count is the first number you interrogate and the one you return to whenever the budget will not close. You know that a couple who says "we can't cut anyone" is making a choice — a legitimate one — and you make sure they know what it costs, in specific dollars, before they make it. And you learn to raise it early, in month two, when it is arithmetic, rather than in month nine, when it is a fight in a kitchen.

The corollary that clients find hardest: you cannot have a big wedding and a nice wedding on a fixed budget. You can have either. Which one you want is a real question with no wrong answer, and the planner's job is to make sure it is asked out loud rather than decided by default.

5.2 The Three Kinds of Cost

Almost everyone — including many working planners — thinks of wedding costs as splitting two ways: per-person costs and flat costs. That model is wrong in a way that matters, because it misses the category where all the useful savings hide.

There are three.

Fully variable

Costs that move smoothly with every single guest.

One more guest costs one more of each of these, immediately. These are the costs the client's mental model captures.

Step-function

Costs that do not move at all until a threshold is crossed, and then jump.

This is where the money is. A step-function cost means the 121st guest may cost $28 and the 122nd may cost $1,340 — and it means that cutting four names sometimes saves nothing while cutting six saves $1,600.

Fixed

Costs indifferent to guest count within any plausible range.

Photography. Videography. Music. The officiant. The ceremony floral. The venue site fee. Attire. The planner. Stationery design (though not printing or postage). Lighting, mostly. The marriage license.

💰 Run the Numbers: the $42,000 wedding, decomposed

Reyes–Whitfield, 140 guests, Wildrye Farm.

Category	Total	Fully variable	Step	Fixed
Catering (food + service + staff)	$14,700 | $11,900	$2,800	—
Venue	$5,040 | — | — | $5,040
Photo + video	$4,200 | — | — | $4,200
Rentals	$3,780 | $1,540	$2,240	—
Bar	$3,360 | $2,660	$700	—
Floral / décor	$2,940 | $1,890	—	$1,050
Music	$2,100 | — | — | $2,100
Attire / beauty	$2,100 | — | — | $2,100
Stationery	$840 | $520	—	$320
Cake	$630 | $560	—	$70
Transport	$840 | — | $840	—
Officiant / license / gratuities	$630 | $180	—	$450
Contingency	$840 | — | — | $840
Total	$42,000** | **$19,250	$6,580** | **$16,170
Share		46%	16%	38%

Sixty-two percent of this budget responds to guest count. That is the number to put in front of a client who says the list is fixed.

Note also what the composition tells you about this specific wedding: the fixed share is 38%, which is high, because photography, music, and the venue are all substantial and none of them care how many people attend. At a larger wedding the fixed share would be a smaller fraction, and the guest-count lever would be even more dominant.

🔄 Check Your Understanding

1. Name the three cost categories and give two examples of each.
2. Why does the step-function category matter more to a planner than the fully variable one, even though it is usually smaller in total?

Verify

1. Fully variable (food, bar, place settings, chairs, favors) — move with every guest. Step-function (tables, servers, bartenders, restroom trailers, tent size, shuttles, dance floor, generator) — flat until a threshold, then jump. Fixed (photography, music, venue fee, officiant, attire, ceremony floral, planner) — indifferent to count.
2. Because steps create free savings and free additions. Between two thresholds, adding a guest is cheap; crossing one is expensive. A planner who knows where the steps are can tell a client "cutting four saves you $180, cutting six saves you $1,780" — which is actionable in a way that a smooth per-head figure never is.

5.3 The Marginal Guest

Now the arithmetic that changes conversations.

Clients — and, honestly, most vendors — reason with the average guest cost: total budget divided by headcount. $42,000 ÷ 140 = **$300 per guest**. That number is misleading in both directions and should almost never be used for a decision.

What you need is the marginal cost: what does the next guest cost, or what does removing this guest save?

💰 Run the Numbers: what one guest actually costs

Guest 141, arriving between step thresholds:

Component	Cost
Food	$85
Service charge and tax on food (28%)	$24
Bar	$24
Place setting (china, glass, flatware, napkin)	$9
Chair	$6
Menu, place card, favor	$5
Cake slice	$4
Invitation and postage (shared household — half)	$3
Marginal cost	$160

$160, not $300. The average includes the photographer and the band, and neither of them notices guest 141.

Now guest 145, who requires a 19th table:

Component	Cost
All of the above	$160
Table rental	$18
Linen	$34
Centerpiece	$95
Additional server (this is the 8th)	$260
Marginal cost	$567

And guest 151, who crosses the restroom-trailer bracket:

Component	Cost
Base	$160
Restroom trailer, next size up	$700
Second shuttle run	$420
Marginal cost	$1,280

Three guests. $160, $567, $1,280. Same wedding, same week.

Why this matters for how you talk to clients

The single most useful thing you can hand a couple is not a per-head number. It is a step map: a list of the guest counts at which something expensive happens.

⚡ Quick Reference: the Reyes–Whitfield step map

At guest count	What happens	Cost of crossing
128	17th table	+$147
136	18th table	+$147
138	8th server	+$260
144	19th table	+$147
150	Restroom trailer bracket	+$700
150	Second shuttle run	+$420
152	20th table	+$147
160	2nd bartender	+$310
160	21st table	+$147
168	Tent frame width	+$1,900

What a client does with this. They are at 140 and wondering whether to add the eight people they cut from Sam's work. The answer is not "that's $1,280 more" (8 × $160). It is:

"Eight more takes you to 148. That's $1,280 in food and drink plus one more table at $147 — so about $1,430. But if it becomes ten more, you cross 150, and that's another $1,120 for the restroom trailer and a second shuttle. So eight is fine and ten is expensive. If you're going to add, add eight, and if there's any chance of it becoming ten, add none."

That is a genuinely useful answer, it takes fifteen seconds to give once you have the map, and no client has ever received one before.

🧩 Productive Struggle

Before reading on, try this without the tables above.

A couple is at 156 guests and $3,100 over budget. Their catering is $92/head all-in including service charge. They want to know how many people they need to cut.

Write down your answer, and then write down what is wrong with it.

The naive answer is $3,100 ÷ $92 = 34 guests. That answer is wrong in a way that will cost your client a great deal of pain, and it is wrong in two directions at once. Spend three minutes working out both errors before continuing.

The two errors

Error one: the per-head figure is too low. $92 is catering only. The marginal guest also costs bar, place setting, chair, stationery, cake, and favor — perhaps $50 more. Using catering-only per-head over-estimates the number of cuts needed by roughly 35%.

Error two, and this is the expensive one: it ignores the steps. Cutting from 156 downward, the first crossings back are the 20th table (at 152), then the bartender and the shuttle and the restroom bracket clustered around 150. **Cutting six names — from 156 to 150 — might save $1,590**, not $552. Cutting 34 names is very possibly 20 more than necessary.

The correct method: build the step map first, then work backward from the deficit, then tell the client the smallest cut that closes it. Frequently it is less than half what a per-head calculation suggests, and every name you do not have to ask them to cut is a real gift.

5.4 Building the List

Before anyone can cut a name, the list must exist in a form where cutting is possible. Most couples' lists are not.

The typical couple's guest list is a single spreadsheet of names, sometimes with addresses, occasionally with a column marked "maybe." That structure makes cutting impossible, because there is no basis on which to compare two names.

The architecture

Every name gets five attributes.

⚡ Quick Reference: guest list architecture

Column	Values	Why it exists
Household	A household ID	Invitations, stationery cost, and RSVP tracking are per household, not per person
Owner	Partner A · Partner B · A's parents · B's parents	The only basis on which cuts can be allocated fairly
Tier	A · B · C	The cutting mechanism (§5.7)
Category	Immediate family · extended family · close friends · social friends · work · parents' friends · plus-ones · children	Attendance forecasting (§5.6) and cut policy
Travel	Local · regional · long-haul · international	Attendance forecasting; also drives accommodation and transport

Plus, later: RSVP status, meal choice, dietary requirement, table assignment.

The Owner column is the one that makes the chapter work. Without it, "we need to cut twenty-six" is an undifferentiated demand that lands on whoever is least willing to fight. With it, the conversation becomes arithmetic:

"You've got 186. We need 160. Your list is 78, Sam's is 61, your parents contributed 29 and Sam's parents 18. Proportionally that's about eleven from you, nine from Sam, four from your parents, and two from his."

That sentence is not pleasant, but it is fair, and it is fair in a way both people can check. Fairness is what makes the cut survivable. An unallocated cut becomes a negotiation about who loves whom more, and it is one of the more corrosive conversations available to two people planning a wedding.

The allocation question

Who decides how many names each side gets?

There are three conventions, and you should surface which one is operating rather than assuming.

Proportional to current list. As above. Simplest and usually the least contentious, because it asks everyone to give up the same share. Its weakness: it rewards whoever padded their list.

Equal thirds (or halves, or quarters). Each partner and each contributing parent gets an equal allocation from the start. Cleanest for a list being built from zero. Its weakness: it ignores genuine asymmetry — one partner may have four siblings and the other none.

Proportional to contribution. Whoever pays gets more names. This convention is real, it operates whether or not anyone says so, and it is the most likely to produce resentment. Surface it explicitly if a parent is paying — Chapter 3's decision-rights map should already have told you whether it is live.

📋 The Planner's Script: allocating the list

Context: Month two or three, before invitations are designed, ideally before the list has grown. This conversation is dramatically easier before a name is on paper than after.

Opening:

"Before the list gets long, I want to agree how many names each side gets. It sounds bureaucratic and it is genuinely the thing that stops this becoming unpleasant in about six months."

The frame:

"Your budget supports about 140 people. That's not a target — it's the number where the wedding you've described is the wedding you get. So let's decide now: how do we split 140?"

Offer the conventions:

"Couples usually do it one of three ways. Split it evenly between the two of you and then decide together how many your parents get. Or split it by how many people each of you genuinely needs there, which is fairer if one of you has a huge family. Or — and this happens whether people say it or not — weight it toward whoever's paying. I'd rather we picked one on purpose."

If a parent is paying:

"Your parents are putting in six thousand. Do they expect names? It's completely normal if they do, and it's much easier to give them a number now than to take names away later. What if we gave them a fixed allocation — say fifteen — and told them it's theirs to spend however they like?"

Why a fixed allocation works: it converts an open-ended entitlement into a bounded, ownable resource. A parent given "fifteen names, yours to choose" behaves completely differently from a parent given "let us know if there's anyone." The first is a gift; the second is an invitation to keep asking. This is the same mechanism as Chapter 3's contributor role menu.

Closing:

"I'll set up the list with a column for whose name each one is. Not to police it — so that if we have to cut later, it's arithmetic instead of an argument."

5.5 The Three Policies

Three questions produce more guest-list conflict than everything else combined. Decide all three explicitly, early, and in writing, and apply them without exception.

Plus-ones

⚡ Quick Reference: plus-one policies, ranked by defensibility

Policy	Rule	Defensible?	Notes
Married / engaged / cohabiting	Partners of these statuses are invited by name	★★★★★	The most defensible line. Objective, checkable, and hard to argue with
In a relationship you know about	Named partners invited	★★★☆☆	Fair in spirit; requires knowing, and people will be missed
Everyone gets one	Universal	★★★★★	Also very defensible, and expensive: 30 single guests = 30 extra at marginal cost
Wedding party only	Only attendants get a plus-one	★★★☆☆	Common; can feel arbitrary to a close friend who is not an attendant
Nobody gets one	No plus-ones at all	★★☆☆☆	Hardest to hold. Produces the most complaints, especially from guests who will know nobody
Case by case	Decided per guest	★☆☆☆☆	Avoid. Guarantees inconsistency, and inconsistency is what people actually object to

The rule that matters: any policy applied consistently is defensible. No policy applied inconsistently is. Guests almost never object to a rule; they object to discovering that someone comparable to them got something they did not.

The arithmetic: in a 140-guest wedding with 25 single guests, a universal plus-one policy adds 25 people. At $160 marginal, that is **$4,000 — plus, in this case, crossing the 150 restroom bracket and the shuttle, which adds another $1,120. Universal plus-ones for this couple cost **$5,120, which is 12% of their total budget. That is a legitimate choice and it must be a chosen one.

Children

Three positions, all defensible, one dangerous.

All welcome. Warm, and it changes the event: highchairs, kids' meals at $18–$30, a supervised space, noise during the ceremony, and an earlier finish for a third of the room. Plan for it properly or it will plan itself.

Immediate family only. The most common compromise. Defensible if defined precisely — "children of the couple's siblings" is checkable; "close family" is not.

Adults only. Entirely legitimate, will cost you two or three guests who cannot arrange care, and requires that you say it clearly and early. Do not bury it. Put it on the website, and if anyone asks, the answer is warm and unmovable.

The dangerous position is no position — where nobody decides, some people bring children and some do not, and the ones who arranged babysitting discover at the reception that they need not have. That produces genuine, lasting resentment, and it is entirely a planning failure.

📋 The Planner's Script: the children question, asked of the client

"Kids: all, none, or immediate family only? I'd like a rule rather than a feeling, because whatever we choose I'm going to have to apply it to everyone, and the complaints don't come from the rule — they come from exceptions.

And if it's 'immediate family only,' tell me exactly who that is. 'Nieces and nephews' is a rule. 'Close family' is a fight in month eight."

Coworkers

The category with the worst ratio of social cost to genuine desire, and the one most often included by default.

Ask directly: "If you left your job next month, which of these people would you still see?" It is a slightly brutal question and it is astonishingly effective. Most couples cut the work list by half within ninety seconds of being asked it.

Two special cases worth flagging: a boss whose non-invitation would have professional consequences (rare but real — a genuine cost to weigh), and a workplace where a partial invitation will be visible and awkward, in which case the honest options are all or none.

⚠️ Common Pitfall: the reciprocity trap

"They invited us to theirs, so we have to invite them."

This single belief accounts for an enormous share of over-long guest lists, and it is worth examining out loud with a client, because it usually will not survive examination.

A wedding invitation is not a debt. If a couple attended someone's wedding three years ago, sat at table eleven, and has seen them twice since, the relationship is not one that requires reciprocation at $160 a head plus, potentially, a step crossing.

The question that dissolves it: "If they didn't come, would you feel their absence on the day?" Not "would they be offended" — that is a different question about someone else's feelings. Would you miss them.

When the answer is no, and it usually is, you have found a name that costs $160 and provides nothing to either party except the avoidance of an imagined slight — one which, in the overwhelming majority of cases, never materializes.

5.6 Forecasting Attendance

Invited is not attending, and the gap between them is where a good planner earns their fee twice: once by not over-buying, and once by not under-buying.

⚡ Quick Reference: typical attendance rates by category

Category	Typical acceptance	Notes
Immediate family	95–100%	Effectively fixed
Wedding party	100%
Extended family, local	80–90%
Extended family, long-haul	50–70%	Falls steeply with distance and cost
Close friends, local	85–95%
Close friends, long-haul	60–80%	Younger cohorts travel more readily
Social friends, local	70–85%
Social friends, long-haul	35–55%	The most over-estimated category
Coworkers	60–80%	Local; drops sharply otherwise
Parents' friends	65–85%	Often higher than the couple expects
Children	85–95%	If invited, they come
Plus-ones (unnamed)	55–75%	Frequently declined or not brought
Overall, local wedding	75–85%
Overall, requiring travel for most	55–70%
Overall, international destination	35–55%	Chapter 33

These are ranges, and they are Tier 2 — widely reported trade patterns rather than measured data. They vary by culture, region, season, day of week, and the couple's own community. A wedding in a family's home town where three generations live within twenty miles will run at the top of every range. A wedding held on a Sunday in October, three hours from anywhere, will not.

Build your own numbers. After five weddings you will have a better forecast for your market than any table can give you. Record predicted versus actual for every category, every time.

The two-sided error

Getting this wrong is expensive in both directions, and the errors are not symmetrical.

Over-forecasting (planning for more than come) wastes money on a guaranteed count you cannot recover, empty seats, and food nobody eats. It is expensive and it is survivable.

Under-forecasting (planning for fewer than come) means insufficient seating, insufficient food, and a bar that runs out — and those are visible failures in front of everyone. Under-forecasting is much worse than over-forecasting, which is why every convention in this industry — the caterer's overage, the guarantee mechanism, the extra table — biases toward the safe side.

💰 Run the Numbers: forecasting the Reyes–Whitfield list

140 invited. Categorized:

Category	Invited	Rate	Expected
Immediate family (both sides)	22	98%	21.6
Wedding party	10	100%	10.0
Extended family, local (Reyes side, mostly within 40 mi)	31	88%	27.3
Extended family, long-haul (Whitfield side, Midwest)	14	60%	8.4
Close friends, local	24	90%	21.6
Close friends, long-haul	11	70%	7.7
Social friends	12	78%	9.4
Coworkers	8	70%	5.6
Parents' friends	8	80%	6.4
Total	140	~84%	117.9

Expected attendance: 118, from 140 invited.

This single calculation is worth thousands of dollars and it takes twenty minutes. At $160 marginal, the difference between planning for 140 and planning for 118 is **$3,520** — and it takes the wedding from 150-plus (restroom bracket, second shuttle, second bartender) down under 128, which saves another $1,400 in steps.

But do not spend the savings yet. §5.8 explains why the plan must sit between the forecast and the invitation count until the RSVPs are actually in — and why a planner who confidently books for 118 in month three is taking a risk that is not theirs to take.

🔄 Check Your Understanding

1. Why is under-forecasting worse than over-forecasting, and how does the industry's structure reflect that?
2. Which guest category is most commonly over-estimated, and why does that matter disproportionately?

Verify

1. Over-forecasting wastes money on unrecoverable guarantees and uneaten food — expensive but invisible. Under-forecasting produces insufficient seating, insufficient food, and a bar that runs out, all of which fail publicly in front of every guest. The industry biases toward safety accordingly: caterers cook 3–5% over, guarantees set a floor rather than a ceiling, and planners add a table.
2. Long-haul social friends, at 35–55%. It matters disproportionately because it is usually a large category and because couples' intuitions about it are formed by their closest relationships, where the rate genuinely is high. Applying a close-friend intuition to a social-friend list is the most common forecasting error.

5.7 The A/B List

The mechanism that lets a couple invite more people than they can afford, without either lying or over-committing.

The structure. The list is tiered. Tier A is invited immediately. Tier B is held, and invitations are released as declines arrive from Tier A. Tier C is the honest acknowledgment that some names will not be invited at all.

The mechanics. Send Tier A eight to ten weeks before the wedding, with an RSVP deadline four to five weeks out. As declines arrive, release Tier B invitations in batches — but only up to about three weeks before the wedding, after which an invitation arrives with visible lateness and does more harm than good.

The ethics. This makes some people uncomfortable, and the discomfort is worth taking seriously rather than dismissing.

The honest position: a B-list invitation sent six weeks out is indistinguishable from an A-list one, and nobody is harmed. A B-list invitation sent eleven days out is obviously a B-list invitation, and the recipient will know. The practice is fine; the timing is what makes it fine or not.

The stronger objection is that it requires the couple to rank their friends, which some people find genuinely distasteful. To which the answer is: they are going to rank them either way, because the budget forces it. The A/B list makes the ranking explicit and lets more people be invited than would otherwise be. Refusing to tier does not produce a wedding where everyone is invited; it produces one where the ranking happened silently, earlier, and with fewer people invited overall.

📋 The Planner's Script: introducing the A/B list

"You've got 186 names and a budget for 140. There are two ways to handle that. One is to cut 46 people now and invite 140. The other is to invite in two waves — send about 155 first, and as declines come in, send the rest. Historically about 15–20% decline, so a good number of your second wave will get an invitation.

I want to be straight about the one risk: if we send the second wave later than about three weeks out, it reads as an afterthought and it will be noticed. So we set a hard date, and anything not sent by then doesn't get sent. Are you comfortable with that?"

If they object to the ranking:

"I understand, and I'd say one thing. You're going to rank them anyway — the budget makes that decision whether or not you make it consciously. This way, more people get invited than would if we cut now, and everyone who gets an invitation gets a real one. The alternative isn't 'nobody gets ranked.' It's 'the ranking happens now and quietly, and thirty fewer people come.'"

⚠️ Common Pitfall: the tier that leaks

The one genuine failure mode. If a Tier B guest learns they were on Tier B, the damage is real and it is not recoverable.

Three rules, all of them absolute:

1. Never label the tiers in a shared document. Use an innocuous column heading and neutral values. If the couple's mother has view access to a spreadsheet with a column marked "B-LIST," this will end badly.
2. Never discuss tiers in a group chat, and tell the couple this explicitly.
3. Never send Tier B after the cutoff. The lateness is the tell.

And one more, learned the hard way: do not let the couple's wedding party know the tiers exist. Not because they would leak deliberately, but because a bridesmaid who knows will accidentally say "oh, are they coming?" in the wrong tone, in front of the wrong person, at the shower.

5.8 RSVPs and the Final Count

The RSVP deadline is a fiction. Plan around this rather than complaining about it. Roughly 15–25% of guests will not respond by the stated date, and a portion will never respond at all.

Set the stated deadline five weeks before the wedding when your caterer's guarantee is due at ten days to two weeks. That gap is your chase window, and you need all of it.

The chase. Two to three days after the deadline, the couple (not you — this must come from them) contacts every non-responder personally. Text or phone, not email. This recovers the large majority. A second pass a week later gets most of the rest. Anyone still silent at the guarantee deadline is counted as not attending, and the couple must be told this in advance so it is a policy rather than a surprise.

⚡ Quick Reference: the count timeline

When	What	Who
−10 weeks	Invitations sent (Tier A)	Planner coordinates, couple addresses
−5 weeks	Stated RSVP deadline
−5 to −3 weeks	Tier B released in batches as declines arrive	Planner tracks
−4.5 weeks	First chase: personal contact, every non-responder	The couple
−3 weeks	Hard cutoff for any Tier B invitation	Planner enforces
−3.5 weeks	Second chase	The couple
−2.5 weeks	Preliminary count to caterer and rentals	Planner
−10 to −14 days	Guaranteed count due — contractually binding	Planner
−7 days	Seating chart final	Planner
−3 days	Final headcount to venue, notify of any late changes	Planner

The guaranteed count is the one that costs money. It is the number you contractually agree to pay for regardless of who attends, and it is nearly always a floor and not a ceiling — you can usually add a few late, at a premium, and you can never subtract. Chapter 14 covers the mechanics and the negotiation.

How much to guarantee. Not your forecast. Not your invitation count. The convention that serves most weddings: guarantee the confirmed yes count, plus 2–3%, which covers the late additions and the guest who brings someone they should not have. A 118-confirmed wedding guarantees 121 or 122.

Below that and you risk a guest without a meal, which is a visible failure. Above it and you are paying $85 a head for empty chairs.

5.9 Practical Considerations

Raise the number in month two. Not month nine. In month two it is arithmetic; in month nine it is the kitchen. The single most valuable thing in this chapter is timing.

Build the step map before the first budget conversation. It takes an hour with the venue's constraints and a rental price list, and it makes you the only person in the room who can answer "what does one more cost?"

Get the Owner column populated before the list exceeds a hundred names. Retro-fitting ownership onto a 190-name list is an unpleasant afternoon.

Keep the forecast visible. Show the client both numbers — invited and expected — from the first budget onward, so that expected attendance is never a surprise revelation.

Common mistakes:

1. Treating the guest count as a constraint rather than a variable. The most expensive error in this chapter.
2. Using average cost per guest for a decision. It is wrong in both directions and it hides every step.
3. Cutting more names than necessary because nobody built a step map. This one is cruel as well as sloppy.
4. A "maybe" column instead of tiers. Maybe is not a decision; it is a decision deferred to a worse moment.
5. Case-by-case plus-ones. Guests object to inconsistency, not to rules.
6. No children policy, or one defined by an adjective.
7. Letting the couple believe the RSVP deadline will work.
8. Guaranteeing the invited count. You are paying for people who told you they were not coming.
9. Labeling the B-list in a shared document.

Things to do this week:

• Build a step map for a real venue in your market: table steps, server steps, bartender steps, restroom brackets, tent brackets, shuttle capacity.
• Build the guest list template with all five columns.
• Write your three policies — plus-ones, children, coworkers — as you would present them.
• Start recording predicted-versus-actual attendance by category. It is the only way to get local numbers.

5.10 Chapter Summary

Key concepts

• Guest count is the master variable. Thirty names produced the same saving as five substantial compromises to the wedding itself — and the smaller wedding was better. You cannot have a big wedding and a nice wedding on a fixed budget.
• Costs split three ways, not two: fully variable, step-function, and fixed. In a typical wedding, ~62% of the budget responds to guest count.
• Steps are where the money is. The 141st guest costs $160, the 145th costs $567, and the 151st costs $1,280 — same wedding, same week.
• Never use average cost per guest for a decision. Build a step map instead, and answer "what does one more cost?" in fifteen seconds.
• Every name needs five attributes: household, owner, tier, category, travel. The owner column is what makes cutting arithmetic rather than an argument about who loves whom more.
• Allocate the list explicitly — proportional, equal, or contribution-weighted. Give contributing parents a fixed number of names, which converts an open entitlement into a bounded, ownable resource.
• Three policies, decided early and applied without exception: plus-ones, children, coworkers. Any policy applied consistently is defensible; no policy applied inconsistently is.
• Attendance runs 75–85% for a local wedding, 55–70% when most guests travel. Long-haul social friends are the most over-estimated category.
• Under-forecasting is worse than over-forecasting, and the whole industry's conventions reflect it.
• The A/B list is ethical if the second wave lands more than three weeks out and the tiers never leak. Refusing to tier does not produce a wedding where everyone is invited — it produces one where fewer people are.
• The RSVP deadline is a fiction. Build a five-week gap and a two-pass chase, run by the couple, not by you.
• Guarantee the confirmed yes count plus 2–3% — never the forecast, never the invitation count.

Action items

• [ ] Build a step map for a real venue in your market
• [ ] Build the five-column guest list template
• [ ] Write your plus-one, children, and coworker policies
• [ ] Start a predicted-versus-actual attendance log

Decision framework — how many names must be cut?

1. Compute the deficit in dollars.
2. Build the step map for this specific event.
3. Working downward from the current count, identify the nearest step crossings.
4. Find the smallest cut that closes the deficit — frequently far fewer than a per-head calculation suggests.
5. Allocate the cut proportionally by owner.
6. Deliver it as arithmetic, with the number, in one conversation.

Spaced Review

Answer before opening.

1. (From Chapter 3) What is the difference between a payer and a decider, and what does payment actually confer?
2. (From Chapter 1) Why does a raw site frequently cost more all-in than a venue with twice the site fee?
3. (Bridging) Chapter 4's forced ranking and this chapter's step map are both tools for making a later decision fast. What is different about when each one is used, and why does the guest list need a tool that Chapter 4's ranking cannot provide?

What's Next

Part I ends here. You have an engagement, a scope, a client relationship, a vision, and a number. Everything Part II does is spend money against them.

In Chapter 6: Budget Architecture, we build the budget properly: benchmark allocations adjusted for the forced ranking, the separation of fixed from variable from step costs you have just learned, the service charge and tax arithmetic that ambushes half of all couples, a real contingency to replace the 2% placeholder that has been sitting in the Reyes–Whitfield budget since Chapter 1, and the presentation that makes a client accept a number rather than argue with it. It contains this book's fifth threshold concept: a budget is a priority statement rendered in dollars.

Before you go on, build the list. Chapter 6 cannot start without a number.

📐 Project Checkpoint: the Reyes–Whitfield list

Tier 3 — Illustrative. The Reyes–Whitfield wedding is a composite created for this book.

Alicia sends you a spreadsheet. It has 186 names on it and three columns: name, address, and a column called "?" containing 31 question marks.

1. Restructure it. Build the five-column architecture. Allocate all 186 names plausibly across four owners: Alicia (~78), Sam (~61), Alicia's parents (~29), Sam's parents (~18). Assign categories and travel status. Note that the Whitfield side is largely Midwestern and will mostly be long-haul.

2. Build the step map. Using the §5.2 step table and the Wildrye Farm constraints established since Chapter 1 (raw site, one restroom trailer, guest transport implied by the 90-minute distance), identify every threshold between 110 and 190 guests and what each costs. You will have to make reasonable assumptions; state them.

3. Forecast. Categorize and apply attendance rates. What is your expected attendance if all 186 are invited? What if 140 are?

4. Do the arithmetic that matters. They are at 186 with a budget supporting roughly 140.

• What is the smallest cut that gets them to a defensible position, given the steps?
• What does that cut save, precisely?
• How does the cut allocate proportionally across the four owners?

5. Handle the two live complications.

• Sam's parents contributed $6,000 and have 18 names. From Chapter 3 you know they pay and do not decide. Does contribution buy names? Write what you would advise Sam to say to them, and write the fixed-allocation version.
• Rosa Reyes. Alicia's grandmother is the highest must-be-pleased person in the engagement, and Mexican-American extended family lists are frequently large and non-negotiable in ways a planner should not attempt to negotiate. How does this constrain the cut? Where would you refuse to look for names, and how would you say so to Alicia?

6. Write the conversation. All of it — the number, the allocation, and the sentence you use when Alicia says "I can't cut my mother's list."

7. Set the three policies. Plus-ones, children, coworkers. Write each as you would put it on their wedding website, and compute what the plus-one policy costs at their step map.

File it in Reyes–Whitfield. Chapter 6 builds the budget from this number, and every number in Part II descends from it.

Chapter 5 Exercises → exercises.md

Chapter 5 Quiz → quiz.md

Case Study: Twenty-Six Names → case-study-01.md

Case Study: The Marginal Cost Curve → case-study-02.md

Key Takeaways → key-takeaways.md

Further Reading → further-reading.md