Polaris Hauled 200 Tons of Dirt Into Manhattan for 1,000 Rides
Polaris trucked 200 tons of dirt and 5 tons of concrete into a Manhattan lot in July 2026, building a 650-foot off-road track above the No. 7 subway line for Fanatics Fest. The event drew 200,000 people. The track was built for roughly 1,000 laps, which means somewhere around half a percent of attendees ever sat in one of the vehicles.
That gap is the whole story, and almost none of the coverage touched it.
The number the trade press skipped
The build stats got the headlines, and fair enough, they are genuinely absurd in a way that photographs well. Muse by Clio reported 200 tons of dirt hauled in across 25 truckloads, 150 feet of lumber for a timber-run section, 2,728 combined work-hours across build and teardown, and 12 distinct Polaris vehicles on site: nine off-road models, a snowmobile, a Slingshot, and a military vehicle. Additional coverage added the site specs, including 650 linear feet of track on a 0.7-acre lot with a 742-foot perimeter, 5 tons of concrete, seven days to build and four to tear down, and 734 horsepower sitting on the lot at once.
Buried in that same writeup is the number I keep circling back to. The track was expected to see about 1,000 laps over the run of the event. Roughly 123 miles of driving in total. Twenty-one loops of Central Park.
Meanwhile Event Marketer put Fanatics Fest 2026 at 200,000 attendees over four days at the Javits Center, with 500 athlete appearances and 150 panels competing for the same eyeballs. Be generous and assume every lap ran with a full vehicle, three passengers plus a professional driver. That gets you 3,000 riders at the absolute ceiling. Against 200,000 people, you land somewhere between half a percent and one and a half percent.
Run it as a labor ratio and it looks worse. 2,728 work-hours divided by 1,000 laps is about two and a half hours of human effort per ride. If the job was delivering rides, this was one of the least efficient machines anyone built in New York this year.
The job was not delivering rides.
The queue was the product
Most teams price experiential per attendee. Total spend divided by people who touched the thing, compare it against a CPM, feel good or feel bad. It is a tidy calculation and I think it quietly misprices almost every large activation I have looked at.
What Polaris built was closer to a film set with a working queue attached. The 1,000 riders were the talent. The other 199,000 got a 650-foot dirt spectacle sitting in the middle of Manhattan, visible from the street, loud, with vehicles launching off a jump above a subway line. That is a completely different product from a ride, and it scales to everybody walking past without adding a single work-hour.
Think about it the way you think about a commercial shoot. Nobody asks how many people were physically present at a shoot. You ask what came out of it and what the footage cost per usable asset. 2,728 work-hours for an eleven-day build, run, and strike is a normal production line item. 2,728 work-hours to hand out a thousand rides is indefensible. Same number, and the framing decides whether your CFO signs off.
So price the build against the asset list, not the foot traffic. Before anyone approves the spend, write down how many hero shots, b-roll clips, creator posts, and owned-channel pieces you expect the footprint to produce, then divide build cost by that number. From what I have seen, if you cannot name at least 40 to 60 distinct usable assets coming out of a six-figure activation, the thing is designed wrong, and no amount of on-site engagement rescues it later.
Polaris earned the gaming audience before it spent the money
The order of operations here matters more than the dirt does, and it is the part most brands get backwards.
Polaris did not show up at a gaming-adjacent festival cold. The RZR Pro R 4 became a playable vehicle in Call of Duty: Warzone in Season 01 of Black Ops 6, which launched November 14, 2024, alongside a program called Polaris Operator Academy that put the top 100 Call of Duty content creators on a closed course with professional drivers before they went off and posted about it. Engage Interactive, Activision, and Treyarch were all in that build. Holly Spaeth, Polaris VP of brand and powersports marketing, framed the Fanatics Fest activation as the physical half of something that already existed digitally: "We have been partnering with Call of Duty since 2024, and you really get a virtual experience of our products in their games."
So by July 2026 the audience already had a reference point. They had driven the vehicle. In a game, sure, but the shape and the sound and the badge were familiar. The dirt track was not introducing a product. It was letting people cash in something they already recognized.
Spaeth also cited a stat that 60 percent of gamers name outdoor recreation as one of their favorite activities, which is the kind of line I would normally read as post-hoc justification. Here the two-year in-game runway makes it look more like the thing the strategy was actually built on.
Most brands run this in reverse. They buy the activation first, scramble for an audience that has no prior relationship with the product, then wonder why the footage does not travel. The build should be the last step, not the first.
The 250 rental outfitters are the real funnel
This is the part I would steal outright.
A Polaris RZR is a five-figure purchase. Nobody standing in a dirt lot at a fan convention, two minutes after getting out of one, is buying one. The distance between the experience and the transaction is enormous, and that gap is where most experiential marketing quietly dies.
Polaris used the footprint to push Polaris Adventures, its rental network of more than 250 outfitters across the US. That is the next step. It costs a couple hundred dollars instead of thirty thousand, it exists in the attendee's own city, and it can be booked from a phone while standing in the queue.
The cheap next step has to exist before you approve the expensive build.
That is the most transferable thing in the activation and it costs nothing to copy. Before you sign the production quote, write down the exact next action an attendee takes. If it costs them more than roughly $100, or takes more than two clicks from where they are physically standing, you do not have a funnel. You have a nice afternoon. Plenty of brands built beautiful footprints at that event with no answer to "and then what," and those are the ones that get cut in next year's budget review.
Do not let anyone hand you an EMV number for this
The temptation with a build this photogenic is to justify it with earned media value. I would push back hard on that.
Michael Brito's breakdown of EMV's flaws is the reference I keep sending people. The same article gets valued at $10,000 by one vendor and $250,000 by another. It treats a New York Times front page and an influencer tweet as interchangeable units, ignores audience quality entirely, and inflates easily on duplicated impressions and generous CPM assumptions. Because there is no shared formula, benchmarking against another brand is meaningless by construction. A $20 million EMV figure with no matching movement in traffic or conversions is a number somebody chose, not a number anybody measured.
Even the practitioner guides have gotten more careful. Snapbar's 2026 measurement framework deliberately leaves out the dwell-time and earned-media figures that fill most guides, citing attribution concerns, calls the widely quoted 3:1 to 5:1 return "directional, not a measured average," and lands on the only benchmark that survives contact with reality: your own historical data.
For a build like this I would track four things instead. Cost per usable asset. Creator posts, measured on actual engagement rather than follower counts. Rental or trial bookings in the local market during the event and the 30 days after. And branded search volume in the DMA across the same window, which is slow and imperfect, but at least it is a signal you did not invent.
None of those are as satisfying as one big dollar figure. That is sort of the point.
Dirt is cheap, attention architecture is not
And to be fair, none of this is new. Brands have been building spectacle to generate footage for decades, and Škoda pulled something structurally similar at the Tour de France with a fraction of the logistics. What Polaris did well was the sequence: earn the audience inside the game, build the physical thing second, put a cheap next step at the exit.
The 200 tons of dirt is the least interesting decision in the whole project. Anyone with a budget can rent trucks. The two-year runway and the rental network took actual planning, and they are the parts nobody wrote about, probably because neither one photographs well.
If you are sitting on an experiential budget for Q4, I would spend the first two weeks on the exit path and the audience runway before anyone draws a footprint. Brand investment tends to get made too late and then rushed, and rushed experiential is where the money really goes missing. Build the boring parts first. The dirt will still be there when you need it.
Notice Me Senpai Editorial
