Creative Testing: DOOH Lessons Across 6 Verticals

Most advice about creative testing assumes one business, one audience, one channel: run variant A against variant B, keep the winner, repeat. That’s useful, but it sidesteps a harder problem, making creative work across genuinely different verticals at the same time, on the same channel.

We ran into that problem directly when Agora launched an LED-wall (DOOH) advertising offering. In its first month it took bookings across six verticals, food, music, pharmacy, beauty, printing, and influencers. Before going further, one honest distinction matters: that breadth of bookings is a signal about demand for the ad space, not a measurement of which creatives performed best. We’re not claiming a controlled creative experiment with measured winners and losers. What the launch gave us was something else, a clear view of how one channel’s hard constraints collide with six very different messages, and that collision is what this framework is built on.

The Core Tension: One Format, Many Verticals

A food spot, a music event, a pharmacy, a beauty brand, a print shop, and an influencer want to say completely different things. But on an LED wall they all face the same physical constraints: a glance of attention, viewing from a distance, no sound, no clicking, no second chance to explain.

That points to the first principle of cross-vertical creative testing: the constraints of the channel matter more than the specifics of the vertical. Whatever the message, every advertiser on that screen has to obey the same rules to be seen at all. The vertical determines what you say; the channel determines how it can be said. When you work in only one niche, that line is easy to blur. Lay six verticals against one channel and the separation becomes impossible to ignore.

Principle 1: Solve the Constraint Once, Then Vary the Message

Some creative variables are universal to the channel; others are local to the vertical.

The universal ones, legibility at a distance, a single dominant idea, strong contrast, no clutter, are channel prerequisites. They’re not profound discoveries; they’re the floor. But they either hold or they don’t, and they work the same for a pharmacy as for a music event. The local ones, the specific offer, the emotional tone, the call to action, are vertical choices.

The practical payoff: solve the channel constraints once, as a template, and you’ve handled most of the creative problem for every vertical at once. Ignore them in favor of a clever message and you get a beautiful ad nobody can read from across the street. Designing for many verticals at once is what makes that trade-off obvious, because a cluttered execution fails regardless of how good the underlying offer is.

Principle 2: Define “Winning” Per Goal, Before You Compare Anything

A naive cross-vertical comparison asks “which ad performed best?” as if all six chased the same outcome. They didn’t. Each vertical had a different intended goal, and the table below lists those objectives and the KPI each one implies, not results we measured from a single month on one screen.

Vertical	Intended goal of the creative	KPI that goal implies
Food	Trigger a nearby, immediate visit	Proximity-driven foot traffic
Music / events	Drive awareness before a fixed date	Reach within a tight time window
Pharmacy	Build constant, trusted familiarity	Sustained presence and recall
Beauty	Showcase a visual product	Brand impression / desirability
Printing (B2B)	Establish local credibility	Recognition, inbound enquiry
Influencer	Signal real-world visibility	Perceived status, audience growth

Score all six on one metric and five look like failures. That’s the trap, and it’s the same one that ruins single-business testing when an awareness asset gets judged by direct-response numbers. Running several goals side by side just makes the error harder to hide. Matching the KPI to the goal isn’t a sophisticated insight; it’s the prerequisite for any comparison being meaningful at all. Note, too, that most of these KPIs (recall, foot traffic, desirability) require their own measurement plan, you have to set up branded-search tracking, footfall counts, or a “how did you hear about us?” prompt before the run, or you’ll have goals with no way to read them.

Principle 3: Breadth of Demand Is Its Own Finding, But a Separate One

Here’s a finding the launch genuinely supports: six unlike verticals all chose to book the same format in its first month. That’s evidence the ad space has broad rather than niche appeal, it’s solving a problem common to very different businesses, namely the need for broad, local, glance-level awareness.

Keep this distinct from creative performance. “Six verticals bought space” tells you about product-market fit for the channel; it does not tell you whether any particular creative worked. Both are useful, but conflating them is how marketers talk themselves into believing a sales result proves a creative result. When you test something new, look deliberately for both signals, and label them honestly for what each one is.

Principle 4: A Shared Template Creates a More Controlled Test

Because the channel constraints are universal, the efficient way to handle many verticals is a shared creative system rather than a bespoke design per advertiser.

Build a template that already solves the hard channel problems, size, contrast, one-idea discipline, legible-from-distance type, and let each vertical drop its specific message into that proven frame. This isn’t a claim that templates always outperform bespoke creative; a brilliant custom ad can certainly beat a templated one. The point is testing control. When only the message changes and the frame is held constant, you can actually attribute a performance difference to the offer rather than the execution. Vary everything at once and you can attribute it to nothing.

A Simple Cross-Vertical Test Plan

To run this deliberately rather than learn it by accident, the plan is short:

1. Build the template first. Solve the channel’s hard constraints, legibility, one dominant idea, contrast, in a reusable frame before you write a single vertical’s message.
2. Lock the frame, vary one thing. Hold the template constant and change only the message element you’re testing, so any difference in response is attributable to the message, not the execution.
3. Write each vertical’s goal and KPI down up front. Decide before launch what “winning” means for each one, and confirm you can actually measure it, branded-search lift, footfall, enquiries, whatever the goal implies.
4. Set up measurement before the run, not after. A goal with no instrument behind it produces opinions, not findings.

This won’t give you a laboratory-grade experiment, especially on a channel as hard to attribute as out-of-home. But it gives you the next best thing: a structured way to separate what the channel demands from what each message is trying to do, so the lessons you take away are real ones.

Applying This Beyond Out-of-Home

You don’t need an LED wall to use these principles. Any time you run creative across different products, audiences, or business units:

• Separate channel rules from message choices, and standardize the channel rules into a template so every variant clears the bar.
• Define winning per goal, with a measurement plan set in advance, so awareness work isn’t punished for not closing sales.
• Treat the breadth of what responds as its own finding, distinct from how any single creative performed.
• Hold the frame constant so the test is clean, changing one variable at a time even when you’re tempted to redesign everything.

The lesson of the DOOH launch isn’t that any one ad was brilliant. It’s that disciplined separation, channel constraints solved once and reused, goals defined per vertical, demand signals kept distinct from creative signals, is what lets you test honestly across businesses that have nothing else in common.