Kansas City's World Cup: An Operations Review

Between 16 June and 11 July 2026, Kansas City hosted six matches of the FIFA World Cup at Arrowhead Stadium — four group games, a Round of 32 tie, and a quarter-final. Across that window and the fortnight of build-up around it, the region also ran a fan festival at the National WWI Museum, hosted three national team base camps, stood up a temporary consular centre in Overland Park, and operated an entire second public transport network that did not exist in May and does not exist now.

We are a quality management consultancy, not a sports business. But we are based here, and what Kansas City did across those weeks was one of the largest and most public exercises in operational planning this region has ever run — with an immovable deadline, a fixed scope, and every result measured in front of an audience.

Six weeks after the final whistle, enough has been published to look at it the way we would look at a factory. Which parts held, which parts missed, and what actually separated the two.

The short version: the things that were run as controlled processes performed extraordinarily well. The things that were run on forecasts missed badly. That distinction is the whole subject of ISO 9001 clause 8.1, and it is worth walking through carefully, because the same pattern shows up in every manufacturing plant that has ever taken a rush order.

What held up

The streetcar. KC Streetcar reported 838,826 passenger trips between 11 June and 11 July — a four-week record, including the five highest single-day ridership figures in its history, all on fan festival days. The peak, on 11 July, was around 56,000 people in a day. What is more interesting than the total is how it was delivered: nine to twelve streetcars running during festival hours instead of the usual complement, six to eight additional Streetcar Link buses added to absorb overflow, average waits held at five minutes or less, and expanded security at the highest-volume stops.

The airport. On Sunday 12 July, Kansas City International screened 24,266 outbound passengers — an all-time record for the airport, beating the previous high of 23,678 set in May 2025. Eight separate days during the tournament exceeded 20,000. Security wait times were held at two and a half to five minutes throughout. The mitigations were unglamorous: screening lanes opened earlier, additional staff scheduled, some concessions opened ahead of normal hours.

The joint operations centre. A south Kansas City firm, Secure Passage, ran a joint operations centre for 42 days, covering Kansas City and six other host cities with roughly 100 personnel, integrating camera, sensor, drone and robotics feeds across law enforcement, fire, EMS, communications and transportation agencies at federal, state, county and local level. Its president described the value in one sentence: information “didn’t have to come by radio, text, email.”

The temporary transit network. ConnectKC26 ran three motorcoach services for 33 days, 11 June to 13 July, on roughly 200 to 225 leased buses: a regional service with fifteen stops, a match-day stadium shuttle from four park-and-ride sites, and an airport service running every fifteen minutes. Preliminary figures published on 6 July — which exclude the quarter-final entirely — recorded more than 140,000 passenger trips.

Look at what those four have in common. Each had a defined process with a named owner. Each had an acceptance criterion that could be stated in advance and checked continuously — headway in minutes, screening throughput per hour, wait time at the checkpoint. Each had real-time measurement. And each had the authority to adjust inside the operating day: add a streetcar, open a lane, redeploy a bus.

That is not luck. That is clause 8.1.

What clause 8.1 actually asks for

ISO 9001’s operational planning and control clause is short, and most quality manuals reproduce it as a paragraph of boilerplate. Read as an operational instruction, it is a checklist. The clause opens by requiring you to plan, implement and control the processes needed to meet requirements and to implement the actions determined in Clause 6 — the risk and opportunity work — and then lists how:

Determine the requirements for the products and services.
Establish criteria for the processes and for acceptance of the outputs.
Determine the resources needed to achieve conformity to those requirements.
Implement control of the processes in accordance with the criteria.
Keep documented information to the extent necessary to have confidence the processes have been carried out as planned, and to demonstrate conformity.

Then two closing requirements that are easy to skim past: control planned changes, and review the consequences of unintended changes, taking action to mitigate any adverse effects — and ensure that outsourced processes are controlled.

That opening link to Clause 6 is worth pausing on, because it is the part most implementations sever. The standard does not treat risk assessment and operational planning as two exercises. It says the operational plan is where the risk actions get implemented. A risk register that lives in a separate document, reviewed annually, is not what the clause describes.

The streetcar operation satisfies every line of that. So does the airport. Neither was working from a projection of how many people would show up; both were working from a criterion — a maximum acceptable wait — with a measured actual, and a lever to pull when the actual moved.

What missed

Now the other half, and it is not a small half.

The headline number was never a plan. Kansas City’s public case for hosting rested on a projection of 650,000 visitors and more than $650 million in economic activity, repeated constantly from the moment the tournament was awarded. In April 2026, KCUR reported that it could not establish who produced the 650,000 figure or what methodology sat behind it. The number was defended by the head of the local sports commission, but its provenance was never published.

The accommodation forecast was wrong in a specific, measurable way. An American Hotel & Lodging Association survey published on 5 May 2026 found that 85 to 90 per cent of surveyed Kansas City hotels were reporting bookings below projections, against nearly 80 per cent below forecast nationally.

A caveat we are going to apply to ourselves, since the rest of this piece is about unowned numbers: that survey covered 205 operators and owners across all the US host markets, so the Kansas City subsample is not published and is probably small. The survey was also reported as ranking Kansas City worst among US host cities, and we are deliberately not repeating that ranking, because a ranking derived from an undisclosed subsample is exactly the kind of number this piece is arguing against. What the survey does support is a direction — bookings below forecast, in a market that had planned for the forecast — and that direction is corroborated independently by the rental data below.

The short-term rental data tells the same story more precisely. Analysis of AirDNA figures reported by KCUR found that during the tournament, rental revenue rose 88 per cent year on year and supply rose 48 per cent — but demand rose only 20 per cent, so occupancy actually fell. For the opening match, prices went up 103 per cent while occupancy rose 24 per cent. For one of the later group games, prices rose 78 per cent and occupancy fell 2 per cent.

And the benefit did not distribute the way the forecast implied. A Missouri Restaurant Association survey of more than thirty businesses representing 116 locations found that more than half reported slower business than expected, and nearly 60 per cent reported a negative overall economic impact. The association’s own summary was that if you were a barbecue restaurant or sat next to the fan festival you did very well, and if you were further out the traffic never materialised.

None of those were operational failures. Every one of them was a forecast failure — a number that was planned around rather than planned with. Nobody owned the 650,000. Nobody could adjust it mid-tournament. There was no acceptance criterion attached to it and no measurement loop feeding back into anything.

That is exactly the gap clause 8.1 is written to close, and it is the most common gap we find in a manufacturing quality system. A plant will have beautiful process controls on the line and a demand forecast that nobody has ever validated, owns, or revises. The controlled half performs. The forecast half determines whether the controlled half was pointed at the right thing.

The correction loop, working in public

There is one more episode worth taking seriously, because it is the system behaving correctly rather than failing.

At the opening match on 16 June, egress went badly. KCUR reported fans walking miles from Arrowhead because of slow traffic and bus delays, long queues at the stadium, and ride-share surge pricing timed to kickoff. It was, by any reading, a nonconformity against the plan.

And then it got better across the following five match days.

That is clause 8.1 meeting clause 9.1 and clause 10.2. The plan was wrong on day one. Monitoring caught it. The process changed. This is the part of a management system that people find hardest to believe is valuable until they watch it operate: the point of a control loop is not that the first plan is right. It is that the second one is better, and that the improvement is deliberate rather than accidental.

A regional planning study is now under way at the Mid-America Regional Council, combining a public survey with shuttle passenger counts, automatic passenger counter data from buses and the streetcar, vehicle telematics and navigation-app data, specifically to identify where the network bottlenecked. It has not published. When it does, it will be the closest thing this region has ever had to a documented management review of its own transport network under load.

The failure mode that should worry a manufacturer most

Here is the part of this story that maps most directly onto a factory, and it is not a happy one.

The Kansas City Area Transportation Authority received no dedicated state or federal funding for World Cup transit operations, despite years of asking. To cover operator overtime and fuel for the surge, it repurposed roughly $2.1 million from a 2018 federal grant that had been intended for new energy-efficient buses, plus $387,500 from another grant programme, and then drew on reserves. Its chief operating officer told the Missouri Independent in November 2025: “We’ve been pushing local, regional and the feds to no avail. So now we’re scrambling.”

The bill came due before the tournament had even finished being paid for. Nine bus routes were scheduled for elimination and frequency cuts on four more from 6 September — around a quarter of total service, affecting close to a tenth of riders. On 17 August 2026 the Kansas City council voted unanimously to approve $8.3 million to stop the cuts, preserving roughly 48 operator positions and buying time until the next budget round.

Before drawing a management lesson out of that, it is worth being clear about what was actually at stake while it was unresolved. A bus route is not a line item to the person who takes it to work. Several thousand people in this city spent months being told their commute might not exist in September, and the council’s vote is the reason it does. That is the real story, and it belongs to Kansas City rather than to us.

What we can add is the shape of it, because manufacturers will recognise the shape. A surge was delivered to specification, on time, in public — by consuming the resource base of the steady state. The surge succeeded. The baseline nearly failed.

Every manufacturer knows this shape. The hot job that eats the preventive maintenance window. The launch that consumes the quality team for a quarter, so the internal audit programme quietly stops running. The customer recovery plan staffed by pulling the two people who were doing supplier development. In each case the visible commitment is met and the invisible one is drawn down, and the consequence arrives one or two quarters later where nobody is looking for it.

Clause 8.1 asks you to determine the resources needed. Clause 6.1 asks you to determine the risks that need to be addressed. Between them sits the question almost nobody writes down: what are we taking those resources from, and what happens to that thing?

Four things worth taking from this

1. A forecast is not a plan. If a number drives your capacity decisions, someone must own it, its method must be written down, and it must be revised on a schedule. The 650,000 had none of those properties, and it is the single clearest lesson available here.

2. Acceptance criteria beat targets. “Wait time under five minutes” is a criterion you can measure hourly and act on. “A successful tournament” is not. In your own operation, the equivalent question is whether your process criteria are things an operator can check at the machine, or things a manager assesses at month end.

3. Plan the drawdown, not just the surge. Before you commit resources to a peak, write down where they are coming from and what degrades while they are gone. Then decide whether that degradation is acceptable — deliberately, in advance, in a document — rather than discovering it later.

4. Expect day one to be wrong, and build the loop that fixes it. The opening-match egress problem was corrected across subsequent matches because someone was measuring and someone had authority to change the plan. That is worth more than a perfect plan, because you are not going to get a perfect plan.

What we do not know yet

In the interest of not doing what the 650,000 did: several important numbers do not exist yet.

There is no published post-tournament economic impact study for Kansas City. As of late July, Visit KC said comprehensive visitation and economic data was expected in a few months. The regional transport study has not published. Total fan festival attendance has not been published. And the match attendance figures — 69,045 for four of the six matches — look like a configured capacity rather than a turnstile count, so we have described them as announced attendance rather than measured.

We will update this piece when the real numbers land. That, too, is the point.

If you are working through operational planning and control in your own quality management system — or trying to work out why a plan that looked complete on paper came apart under load — that is the kind of problem our ISO 9001 work is built around. We are based in the Kansas City metro, and we work with manufacturers across the region.

Sources for the figures above are the published reporting of KCUR, KSHB 41, the Missouri Independent and the organisations named. Where a number is a projection rather than a measurement, we have said so.