How a Times Square takeover got built and approved.
Watch the session
The setup
The hardest thing about running a media practice inside an independent agency is that clients hire you to execute a decision they already made. The brief arrives pre-shaped. Buy this, at this weight, in this window. The idea, if there is one, came from somewhere else.
Liquid I.V. came to Zambezi with a version of that brief and one complaint attached. Grace Teng, who runs Scale by Zambezi, repeated it on stage.
"We want to do something for us. We spend, you know, quite a bit just in media overall, but we don't necessarily stand out."
That is the whole problem in two sentences. A brand with real money in market, invisible anyway. What Zambezi did about it became a 30-minute takeover of Times Square that Unilever went on to run in London and elsewhere.
This session was 25 minutes on how that was bought, built, staffed and sold internally. It is the most operational conversation of the week, and the mechanics are the point.
Who was on stage, and why they're worth listening to
Part One — How the work got in the door
Zambezi started as a creative-only shop. Scale launched around 2019 and was funded out of the business. Jean's reasoning was competitive rather than financial.
"Zambezi started as a creative-only agency, and I really saw from a value standpoint the value of having the media offering, and then also from the standpoint that a lot of indies really couldn't do that from a creative and a media perspective."
The Liquid I.V. relationship arrived the way most independent relationships do, through a person rather than a process. Jean named the mechanism precisely.
"Liquid I.V. originally came in through a personal relationship that Grace had on the media side, a very go-getter mid-level media director that Grace had worked with."
The structural detail matters more than the introduction. Unilever's buying agency handles performance at scale. What the client wanted, and could not get inside that arrangement, was strategy. Jean's read on why an indie could take the gap:
"I think the nice thing about being independents, as everyone knows here, is we can have a flexible way of how we structure and how we work. What wasn't really done within the Unilever aspect, it was like one agency kind of doing everything and very siloed."
So Zambezi came in through media strategy. Then it did the thing that turns a media assignment into a campaign: it delivered the plan it was hired for and brought something else besides.
"That was not the ask. But we brought it to them, and truthfully, I think they were kind of surprised on actually what to do."
Jean treated the unbriefed idea as a pitch, and she was unusually involved for a CEO who says she stays out of new business. The reason she gives is worth sitting with, because it is a screening criterion rather than an enthusiasm.
"When you see clients who will push and can just get things out of the way, because we know how difficult that can be when we have good work, but we don't have clients that know how to get it through their own system."
The idea itself came out of a consumer insight, not an inventory opportunity. Four o'clock is when people are most dehydrated. That produced the campaign line, and the placement followed from it. Jean's framing of the sequence is the transferable part.
"Really show a creative idea that was based in a media moment first and foremost."
Part Two — How it was built
Doug pushed Grace on the media math. 30 minutes of Times Square is a large, compressed, expensive unit of attention. What makes it pencil?
Grace answered with Scale's operating principle first.
"Our motto as Scale is we believe in ideas before inventory. And yes, spots, dots, all the math needs to make sense. But that being said, what is the idea? I think that's a really big reason why we continue to win, stand out with clients."
She did not put a number on the buy, and nobody made her. What she did give was the structural constraint, which is more useful to anyone contemplating the same move.
"When it comes to Times Square, you have to rent out 30 minutes."
Then the part that separates a stunt on a deck from a stunt in market.
"The 30 boards are actually owned by 10 different companies, so the creative from going to one board, two boards that's owned by one, to three boards down here, and making sure that it's all aligned was a big challenge. And the other thing is the boards are actually timed on different time zones."
30 screens, 10 landlords, mismatched clocks, and a creative idea that only works if the screens read as one canvas. That is the job. Grace's verdict: "really challenging but fun at the same time."
The date was chosen, not available. June 4, inside a summer brand window, timed to the first week of Love Island, which Zambezi had also bought an integration into. One activation, two reasons to be in that week.
The build then absorbed the client's entire internal wish list, on purpose. Before designing the activation, Scale toured the building.
"When we first started working with Liquid I.V., we actually met with almost all of their stakeholders internally when it came to PR, from their impact team, from media, from sampling, from social, from influencers. So we basically had 10-plus stakeholder meetings, and they were all like, with this idea, we wanted to incorporate it. So we actually did incorporate all those ideas within the stunt."
That is why the takeover has robots, sampling, a microsite, and eight to 10 influencers including Dude With Sign attached to it. Each of those elements is somebody's function inside the client, and each of those people became a reason the thing survived approval.
On the influencer layer, Grace made a point about sequencing that inverts how most plans are written.
"You're seeing a lot more that out of home is the one that's starting the social conversations. Things are happening in real life."
The creators were not sourced cold. Liquid I.V. and Unilever already had working relationships and a named shortlist; Scale narrowed it and scaled the output. Fewer negotiations, faster clearance, and a roster the client had already approved once.
As for whether it worked, the evidence offered was organizational rather than statistical. Unilever took the campaign global, activating in London and other markets. Jean noted that the CEO of Unilever Wellbeing was there in person, and that most of the Wellbeing group sits in New York, which is part of why Times Square was the right room to perform in.
The uncomfortable truth
The agency built something the client could not fully use.
Jean said it plainly, and it is the sharpest thing in the session.
"It took people by surprise that they couldn't take more advantage of it."
Liquid I.V. was not ready from a PR standpoint. The summer brand campaign had been scoped as a partnership exercise, so the surrounding machinery, the earned push, the retail tie-in, the regional pickup, had not been built to receive a moment of that size. The work outran the org.
Jean's explanation is a fact about the calendar of any large client.
"Moving into '26, '27, larger global orgs, their timelines are so pushed out. And again, I think the fact that we were able to bring something and get it in a few months showcases how great our clients were."
Read that twice. The compliment and the constraint are the same sentence. Zambezi's speed was the achievement and the problem. A holding-company client running on 18-month planning cycles cannot absorb a great idea delivered in three months, and the gap between those two clocks is where independent work goes to die.
There is a second discomfort, and the room raised it. A question from the floor asked whether entering through media and arriving with creative meant stepping on the incumbent creative agency's toes. Jean did not deny it. She reframed it as a matter of which door you are standing in.
"We were talking directly through the media lens."
That is a real answer and an incomplete one. Independents that hold both disciplines will keep running into this, and the session did not resolve it.
What you can use
Every session from Indies @ Cannes 2026 — plus every other Blueprint, the tools, and the surveys — opens up with one sign-in.
