Case Study 34.1: The Wedding That Was a Build

📜 Tier 3 — Illustrative. Fenwick Dray, the Hallowmere estate, and the Achterberg–Lindqvist wedding are constructed. The numbers are the author's, assembled to show a planner discovering the threshold at a wedding rather than at a festival.


The brief, as it looked

260 guests. A private estate with no permanent structures. Saturday in July.

A marquee, a dance floor, a stage, catering from a field kitchen, 260 chairs, 26 rounds, a generator, portable toilets, and a lighting scheme.

Fenwick Dray had run forty weddings and this was, in his words, "a big one."

He built a delivery schedule.

🕐 The delivery schedule, as written

THU Marquee company, 08:00
FRI 09:00 Furniture and rentals
FRI 09:00 Generator
FRI 10:00 Field kitchen
FRI 11:00 Dance floor
FRI 12:00 Staging
FRI 14:00 AV and lighting
FRI 16:00 Florals
SAT 08:00 Cake, final dressing

Nine rows. Every one of them a time and a supplier.

It looks like a plan and it is a list of arrivals.


What actually happened

🚨 Friday, hour by hour

09:00 Rentals truck and generator low-loader arrive together. The estate has one track in. The low-loader cannot pass the rentals truck
09:25 Rentals truck reverses 180 m up a farm track. Twenty-five minutes gone and a rut in the grass that will be in photographs
10:00 Field kitchen arrives and is told to wait. It waits until 11:40
11:00 Dance floor crew arrive and cannot lay a floor because the marquee's ground anchors are where the floor goes. The marquee company left on Thursday at 16:00
11:15 The dance floor crew go to lunch. Minimum call: four hours. They have been on site for fifteen minutes
12:00 Staging arrives. Cannot build — the dance floor is not down and the stage sits on it
14:00 AV arrives. Cannot rig — no stage. Cannot power anything — the generator is on a low-loader in a field, undelivered
15:30 Generator finally positioned. Cabling begins across a route nobody had drawn
16:00 Florals arrive into a marquee containing four crews, a forklift, and no floor
19:00 Dance floor down
21:15 Stage built
23:40 AV rigged. No focus, no programme, no sound check
SAT 07:00 AV crew return. Nine hours after wrapping — inside turnaround, technically breached, nobody says anything
SAT 13:00 Sound check completes, two hours before guests

The wedding was lovely. The couple never knew.


The arithmetic Fenwick did afterwards

💰 Run the Numbers: what the day cost

Dance floor crew: 4 hours minimum call, 15 minutes worked, returned 18:00 $1,240 of duplicated call
Field kitchen crew idle 10:00–11:40 $680
Staging crew idle 12:00–19:00seven hours $2,940
AV overtime, 19:00–23:40 $2,180
AV Saturday morning call (should have been Friday's tail) $1,560
Ground reinstatement — the rut $900
TOTAL $9,500

Against a $9,500 line that had not existed in the budget, on a wedding whose entire production budget was $34,000.

And the cost that was not money: no sound check until 13:00 on the day, no focus at all, and a crew who worked nine hours' turnaround because nobody was willing to be the person who said no.


What was missing

🚪 There was no dependency map, so there was no order — only preferences.

Every supplier had been asked when they wanted to arrive. Every one said morning.

Draw it and the whole day resolves itself:

``` Marquee ──► Ground anchors set ──► Dance floor ──► Stage ──► AV rig ──► Focus │ ▲ └──► Generator positioned ──► Power distro ─────────────────┘ │ └──► Field kitchen

Furniture ──────────────────► Dressing ──► Florals (NOT on critical path) ```

Alt-text: a dependency diagram for a marquee wedding build. The marquee leads to ground anchors, then dance floor, then stage, then AV rig, then focus. A parallel branch from the marquee runs to generator positioning and power distribution, which feeds both the AV rig and the field kitchen. Furniture, dressing, and florals form a separate chain that is not on the critical path.

Six sequential steps and a power branch that gates two of them.

The generator should have been the first vehicle on site and it was scheduled fourth. The florals should have been last and were scheduled sixth of nine. And the marquee company should not have left on Thursday, because their anchors were a dependency for a crew arriving Friday.


Rebuilt

🕐 The Friday that should have happened

Space
07:00 Generator in and positioned. Low-loader only vehicle on the track Track exclusive
07:45 Power distro laid, route pre-drawn Site
08:00 Marquee company return — anchors adjusted, sign-off Marquee
08:30 Dance floor Marquee
10:00 Staging Marquee
10:00 Field kitchen — different area, parallel Kitchen field
12:00 Crew meal. Fixed
12:30 AV rig Marquee
15:30 Focus and programme Marquee
16:30 Furniture and rentals — 260 chairs, 26 rounds Marquee
18:00 Sound check Marquee
19:00 Dressing Marquee
20:30 Florals — last, into a finished room Marquee
21:30 Off site

Same nine suppliers. Same money. Two hours earlier a finish, a sound check on Friday, and $9,500 that was never spent.

The only structural change is that vehicles were sequenced onto a single-track access and the order came from the dependency rather than from anybody's preference.

🎤 From the Field

Fenwick's own note, which he now reads to planners at trade evenings:

"I had run forty weddings and I had never once asked what had to happen before something else could happen. I asked when people wanted to come."

"Forty weddings, and it had never mattered, because at forty weddings the venue had a floor in it already."


Discussion Questions

DQ1. Fenwick had run forty weddings without needing a dependency map. Why did this one break, and what should have told him in advance?

Consider **Because a venue with a building in it hides every dependency.** The floor is down, the power is in, the room exists — **so the order genuinely does not matter much, and a delivery schedule is adequate.** **A bare field has no dependencies hidden. It has all of them, exposed, and they all have to be created in sequence.** **What should have told him: §34.1's diagnostics, four of which this event satisfies.** Load-in exceeds one working day. **More than three suppliers need the same access point** — in fact all nine need a single farm track. **The crew is larger than the guest-facing team.** **And every element requires another to be finished first.** **Any two of those means the method changes.** **He had four and was still writing a list of arrival times.**

DQ2. The dance floor crew were on site fifteen minutes and cost a four-hour call twice. Whose failure?

Consider **The planner's, entirely, and the mechanism is worth being precise about.** **The crew did nothing wrong** — they arrived when told, found they could not work, and took the meal break their agreement provides. **The marquee company did nothing wrong either**; nobody had told them their anchors were a dependency for somebody arriving the next day. **The failure is that no document existed in which those two facts appeared together.** **Which is Chapter 26's finding arriving at a build:** an instrument does not solve a problem, **it makes the problem findable** — and the walked-forward production meeting of §34.7 is exactly the instrument that would have found it, **because the marquee company and the dance floor company would have been in the same row.** **And note the compounding: the double call is $1,240, and it is the smallest number in the table.** **The seven idle hours of staging crew, downstream of the same cause, are $2,940.**

DQ3. The AV crew worked a nine-hour turnaround and nobody said anything. Unpack that.

Consider **Three things had to be true at once and all three are common.** **The breach was small** — nine hours against ten — **which makes it exactly the kind of thing nobody escalates.** **The alternative was visible and bad:** a wedding at 15:00 with no sound check. **Everybody could see the consequence of enforcing the rule and nobody could see the consequence of breaking it.** **And there was no named person whose job was to stop it.** §34.5a: **"everybody is responsible" means nobody stops anything**, and a crew chief who raises a nine-hour turnaround is volunteering to be the person who caused a problem. **What makes it serious rather than technical:** those people spent Saturday morning rigging and focusing lights, **at height, tired, on a build that had already gone wrong once.** **The turnaround rule exists for exactly that morning.** **And the honest note: it worked out. It usually does. That is why the rule gets broken and why it has to be somebody's explicit job to hold it.**

DQ4. The rebuilt Friday uses the same suppliers and the same money. Is that plausible, or has the case study made it too easy?

Consider **Mostly plausible, with two honest caveats the case study should carry.** **Plausible because the changes are sequencing, not resourcing.** Nothing was added — **the generator moved from fourth to first, the florals from sixth to last, and the marquee company was asked to return for thirty minutes.** **Caveat one: the marquee company returning on Friday morning is not free.** It is a call-out, it may be a minimum call, **and it had to be negotiated at contract rather than requested on Thursday evening.** Call it a few hundred dollars, agreed in month three. **Caveat two: the rebuilt schedule assumes the marquee was finished on Thursday and correct.** **If the marquee runs late — which marquees do — the entire Friday moves and there is no dark period in this build to absorb it.** §34.2's dark day is missing from the rebuilt version too, **and a genuinely good version of this schedule finds two hours somewhere for it.** **So: the $9,500 is real and the rebuild is slightly optimistic**, which is the normal relationship between a corrected plan and the world.

DQ5. "Forty weddings, and it had never mattered, because at forty weddings the venue had a floor in it already." What does that say about how planners acquire this skill?

Consider **That experience does not transfer across the threshold**, and that a planner's confidence scales with the wrong variable. **Forty weddings is real expertise** — vendor management, client work, timeline, triage, guest experience. **None of it contains a dependency map, because none of those forty events had a critical path that could not be reordered.** **And the dangerous property is that the skill's absence is invisible until it fails**, and it fails at scale, on somebody's wedding, expensively. **Which is an argument for the diagnostics in §34.1 being a routine check rather than an interesting idea.** **Seven yes/no questions, asked at contract, on every event** — because the answer changes what document you write, and the document has to exist before the first supplier is booked. **And a second argument, less comfortable:** this is a domain where **reading about it is genuinely a substitute for some of the experience**, because the failure is structural rather than judgmental. **Fenwick did not need forty more weddings. He needed forty minutes and a piece of paper.**